Professional Documents
Culture Documents
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Copyrights The Motorola products described in this document may include copyrighted Motorola computer programs stored in semiconductor memories or other media. Laws in the United States and other countries preserve for Motorola certain exclusive rights for copyright computer programs, including the exclusive right to copy or reproduce in any form the copyright computer program. Accordingly, any copyright Motorola computer programs contained in the Motorola products described in this document may not be copied or reproduced in any manner without the express written permission of Motorola. Furthermore, the purchase of Motorola products shall not be deemed to grant either directly or by implication, estoppel or otherwise, any license under the copyrights, patents or patent applications of Motorola, except for the rights that arise by operation of law in the sale of a product. Restrictions The software described in this document is the property of Motorola. It is furnished under a license agreement and may be used and/or disclosed only in accordance with the terms of the agreement. Software and documentation are copyright materials. Making unauthorized copies is prohibited by law. No part of the software or documentation may be reproduced, transmitted, transcribed, stored in a retrieval system, or translated into any language or computer language, in any form or by any means, without prior written permission of Motorola. Accuracy While reasonable efforts have been made to assure the accuracy of this document, Motorola assumes no liability resulting from any inaccuracies or omissions in this document, or from the use of the information obtained herein. Motorola reserves the right to make changes to any products described herein to improve reliability, function, or design, and reserves the right to revise this document and to make changes from time to time in content hereof with no obligation to notify any person of revisions or changes. Motorola does not assume any liability arising out of the application or use of any product or circuit described herein; neither does it convey license under its patent rights of others. Trademarks Motorola and the Motorola logo are registered trademarks of Motorola Inc. Intelligence Everywhere, M-Cell and Taskfinder are trademarks of Motorola Inc. All other brands and corporate names are trademarks of their respective owners.
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE
Contents
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1-3 1-4 1-4 1-4 1-4 1-6 1-6 1-6 1-6 1-6 1-6 1-8 1-8 1-8 1-8 1-8 1-8 1-10 1-10 1-10 1-12 1-12 1-12 1-14 1-14 1-14 1-14 1-14 1-14 1-14 1-16 1-16 1-16 1-16 1-16 1-16 1-16 1-18 1-18 1-18 1-18 1-20 1-20 1-22 1-22 1-24 1-24
i
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
Contents
DL-HSPDSCH-Info . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . HS-SCCH Info . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Downlink Scrambling Coding . . . . . . . . . . . . . . . . . . . . . . HS-SCCH Channelization Code info . . . . . . . . . . . . . . . . . . HS-SCCH Codes. . . . . . . . . . . . . . . . . . . . . . . . . . . . Channel Type Switching with HSDPA . . . . . . . . . . . . . . . . . . . . . . CELL_DCH ONLY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CELL_DCH & HSDPA . . . . . . . . . . . . . . . . . . . . . . . . . . . CELL_FACH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . HSDPA Control and User Plane Extensions . . . . . . . . . . . . . . . . . . Control Plane Extensions: . . . . . . . . . . . . . . . . . . . . . . . . . New NBAP, RRC & RNSAP IEs . . . . . . . . . . . . . . . . . . . . Radio Resource Sharing . . . . . . . . . . . . . . . . . . . . . . . . Admission / Congestion Control . . . . . . . . . . . . . . . . . . . . User Plane Extensions:. . . . . . . . . . . . . . . . . . . . . . . . . . . PDCP/RLC/MAC-d . . . . . . . . . . . . . . . . . . . . . . . . . . . Frame Protocol. . . . . . . . . . . . . . . . . . . . . . . . . . . . . HSDPA Packet Switched Protocol Stacks . . . . . . . . . . . . . . . . . . . . HSDPA Control Plane . . . . . . . . . . . . . . . . . . . . . . . . . . . Access Stratum Protocols . . . . . . . . . . . . . . . . . . . . . . . Non Access Stratum Protocols . . . . . . . . . . . . . . . . . . . . . MAC-hs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . HSDPA User Plane . . . . . . . . . . . . . . . . . . . . . . . . . . . . Configuration without MAC-c/sh . . . . . . . . . . . . . . . . . . . . MAC-d entity UE Side . . . . . . . . . . . . . . . . . . . . . . . . MAC-hs entity UE Side . . . . . . . . . . . . . . . . . . . . . . . . MAC-d entity UTRAN Side . . . . . . . . . . . . . . . . . . . . . . MAC-hs UTRAN Side . . . . . . . . . . . . . . . . . . . . . . . . HSDPA Timing Relations . . . . . . . . . . . . . . . . . . . . . . . . . . . . HSDPA Downlink Timing . . . . . . . . . . . . . . . . . . . . . . . . . . DPCH (Dedicated Physical Channel) . . . . . . . . . . . . . . . . . . HS-SCCH (High Speed Shared Control Channel). . . . . . . . . . . . HS-PDSCH (High Speed Physical Downlink Shared Transport Channel) HSDPA Uplink Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.5 slots delay of uplink HS-DPCCH . . . . . . . . . . . . . . . . . . Composition of the HS-SCCH . . . . . . . . . . . . . . . . . . . . . . . . . Modulation Type and Number of HS-PDSCHs . . . . . . . . . . . . . . . HARQ Info (Hybrid-ARQ-related Information) . . . . . . . . . . . . . . . . Hybrid ARQ Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . Type I Hybrid ARQ Low UE Complexity . . . . . . . . . . . . . . . . . . Type II Hybrid ARQ High UE Complexity . . . . . . . . . . . . . . . . . Type III Hybrid ARQ medium UE Complexity . . . . . . . . . . . . . . . DCH / HS-DSCH Comparison . . . . . . . . . . . . . . . . . . . . . . . . . Variable Spreading Factor . . . . . . . . . . . . . . . . . . . . . . . . . Fast Power Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . AMC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Multicode Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fast Layer 1 ARQ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DPCH / HS-PDSCH Comparison contd. . . . . . . . . . . . . . . . . . . . 16-QAM Modulation / Interleaving / Channel Coding Scheme . . . . . . . . TrCH Multiplexing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Static rate matching - DCH versus HS-DSCH . . . . . . . . . . . . . . . . Transport block concatenation - DCH versus HS-DSCH. . . . . . . . . . . Soft(er) Handover . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Advantages and Disadvantages of HSDPA . . . . . . . . . . . . . . . . . . . Advantages of HSDPA . . . . . . . . . . . . . . . . . . . . . . . . . . . Reduced Latency for retransmission . . . . . . . . . . . . . . . . . . Higher DL Throughput and Peak Throughput Rates up to 14 Mbit/s . . . Support of Non-Real Time and Streaming Applications . . . . . . . . . High Spectrum and Code Efficiency due to 16-QAM . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1-26 1-26 1-26 1-26 1-26 1-28 1-28 1-28 1-28 1-30 1-30 1-30 1-30 1-30 1-30 1-30 1-30 1-32 1-32 1-32 1-32 1-32 1-34 1-34 1-34 1-34 1-34 1-34 1-36 1-36 1-36 1-36 1-36 1-38 1-38 1-40 1-40 1-40 1-42 1-42 1-42 1-42 1-44 1-44 1-44 1-44 1-44 1-44 1-46 1-46 1-46 1-46 1-46 1-46 1-48 1-48 1-48 1-48 1-48 1-48
ii
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Contents
Disadvantages of HSDPA . . . . . . . . . . . . . . . . . . . . . Higher Processing power in UE. . . . . . . . . . . . . . . . . More Memory in UE due to HARQ . . . . . . . . . . . . . . . Advanced Receiver for 16-QAM . . . . . . . . . . . . . . . . Higher Vulnerability using 16-QAM . . . . . . . . . . . . . . . Software and possibly Hardware Upgrade . . . . . . . . . . . No Macro-Diversity for HS-PDSCH, HS-SCCH . . . . . . . . . Power / Interference Increase in uplink and downlink . . . . . . Signalling Overhead caused by HS-SCCH (dl) and DPCCH (ul) . Future Enhancements of HSDPA . . . . . . . . . . . . . . . . . . Beamforming . . . . . . . . . . . . . . . . . . . . . . . . . . Transmit Diversity . . . . . . . . . . . . . . . . . . . . . . . MIMO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Preview to HSUPA . . . . . . . . . . . . . . . . . . . . . . . . . Total available Transmission Power. . . . . . . . . . . . . . . Much larger Dynamic of the UL Power Control . . . . . . . . . UL does not suffer from Channelization Code Shortage . . . . . Fast Power Control cannot be abandoned in UL . . . . . . . . HSUPA - Areas of Improvement . . . . . . . . . . . . . . . . . . NodeB controlled scheduling . . . . . . . . . . . . . . . . . . Hybrid ARQ in Uplink . . . . . . . . . . . . . . . . . . . . . . Shorter TTI . . . . . . . . . . . . . . . . . . . . . . . . . . HSUPA - Less Beneficial Areas . . . . . . . . . . . . . . . . . . . Higher Order Modulation . . . . . . . . . . . . . . . . . . . . Fast DCH Setup . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . .
1-50 1-50 1-50 1-50 1-50 1-50 1-52 1-52 1-52 1-54 1-54 1-54 1-54 1-56 1-56 1-56 1-56 1-56 1-58 1-58 1-58 1-58 1-60 1-60 1-60
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
iii
Contents
Channel Quality Indication . . . . . . . . . . . . . . . . . . . . . . . Uplink HS-DPCCH Code Allocation . . . . . . . . . . . . . . . . . . . . . Multicode Operation in Uplink. . . . . . . . . . . . . . . . . . . . . . Uplink HS-DPCCH Coding Chain . . . . . . . . . . . . . . . . . . . . . . Channel Coding for HS-DPCCH . . . . . . . . . . . . . . . . . . . . Channel Coding for HARQ-Ack . . . . . . . . . . . . . . . . . . . Channel Coding for CQI . . . . . . . . . . . . . . . . . . . . . . Uplink HS-DPCCH Spreading . . . . . . . . . . . . . . . . . . . . . . . Spreading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DPCCH / DPDCH / HS-DPCCH . . . . . . . . . . . . . . . . . . . . Gain Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Uplink Transmission Delay T(0) . . . . . . . . . . . . . . . . . . . . . . . Uplink / Downlink Timing at UE . . . . . . . . . . . . . . . . . . . . . Uplink Fast Power Control Reason for T(0) . . . . . . . . . . . . . . HS-SCCH / HS-DSCH and HS-DPCCH Timing (1) . . . . . . . . . . . . . HS-SCCH / HS-DSCH and HS-DPCCH Timing (2) . . . . . . . . . . . . . Uplink HS-DPCCH . . . . . . . . . . . . . . . . . . . . . . . . . . . T(TX_diff) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Delta Ack / Nack and Ack-Nack Repetition Factor . . . . . . . . . . . . . . DeltaAck . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DeltaNack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ACK-NACK-repetitionFactor . . . . . . . . . . . . . . . . . . . . . . Delta Ack / Nack and CQI Power . . . . . . . . . . . . . . . . . . . . . . HS-DPCCH ACK / NACK and DTX Recognition. . . . . . . . . . . . . . . False Alarm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CQI Description and Task. . . . . . . . . . . . . . . . . . . . . . . . . . TFRC Info . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Out of Range Signaling . . . . . . . . . . . . . . . . . . . . . . . . . Power Reduction Factor 1 . . . . . . . . . . . . . . . . . . . . . . . Measurement Power Offset 2. . . . . . . . . . . . . . . . . . . . . . CQI Reporting Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . HS-DPCCH Measurement Feed Back - CQI . . . . . . . . . . . . . . . . Measurement Power Offset 3 . . . . . . . . . . . . . . . . . . . . . . CQI Feedback cycle, k . . . . . . . . . . . . . . . . . . . . . . . . . CQI-RepetitionFactor . . . . . . . . . . . . . . . . . . . . . . . . . . DeltaCQI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CQI Reporting Principle . . . . . . . . . . . . . . . . . . . . . . . . . . UE measures the downlink channel quality . . . . . . . . . . . . . . . CQI Algorithm in UE . . . . . . . . . . . . . . . . . . . . . . . . . . NodeB TFRC decision . . . . . . . . . . . . . . . . . . . . . . . . . CQI Mapping Table UE Category . . . . . . . . . . . . . . . . . . . . . UE Category . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CQI value . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Transport Block Size . . . . . . . . . . . . . . . . . . . . . . . . . . Reference Power Adjustment. . . . . . . . . . . . . . . . . . . . . . N(IR) and X(RV) . . . . . . . . . . . . . . . . . . . . . . . . . . . . Example for a CQI Report . . . . . . . . . . . . . . . . . . . . . . . . . CQI and Ack/Nack Considerations . . . . . . . . . . . . . . . . . . . . . HS-DPCCH Power Control during Soft Handover . . . . . . . . . . . . CQI Repetition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Intra NodeB Softer HO . . . . . . . . . . . . . . . . . . . . . . . . . CQI Reporting Feedback Cycle and Repetition . . . . . . . . . . . . . . Ack/Nack Repetition Factor . . . . . . . . . . . . . . . . . . . . . . . Trade off between Ack/Nack/CQI Repetition and long Feedback Cycles . HSDPA during Compressed Mode Operation . . . . . . . . . . . . . . . . Effects of Using 16-QAM . . . . . . . . . . . . . . . . . . . . . . . . . . Higher Throughput Rates. . . . . . . . . . . . . . . . . . . . . . . . Increased Spectrum Efficiency . . . . . . . . . . . . . . . . . . . . . Higher Interference Vulnerability . . . . . . . . . . . . . . . . . . . . Smaller Decision Space . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2-28 2-30 2-30 2-32 2-32 2-32 2-32 2-34 2-34 2-34 2-34 2-36 2-36 2-36 2-38 2-40 2-40 2-40 2-42 2-42 2-42 2-42 2-44 2-46 2-46 2-48 2-48 2-48 2-48 2-48 2-48 2-50 2-50 2-50 2-50 2-50 2-52 2-52 2-54 2-56 2-58 2-58 2-58 2-58 2-58 2-58 2-60 2-62 2-62 2-62 2-62 2-64 2-64 2-64 2-66 2-68 2-68 2-68 2-68 2-68
iv
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Contents
16-QAM Constellation Rearrangement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-QAM Constellation Rearrangement for b = 2 and b = 3 . . . . . . . . . . . . . . . . . . . . Bit Re-arrangement b = 2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bit Re-arrangement b = 3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Power Sharing between HSDPA and DCH . . . . . . . . . . . . . . . . . . . . . . . . . . . . Node B Transmit Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Maximum Transmission Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . UE Transmit Power. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Adaptive Modulation and Coding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Main Benefits of AMC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Availability of higher throughput and therefore increased average throughput per cell. . . . Reduced interference variation due to modulation and coding based link adaptation instead of variations in transmit power. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Utilization of short term fading in the way that a user might be always served in a constructive fade. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Boosted effectiveness in combination with time and code multiplexing scheduling techniques.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Adaptive Modulation and Coding - Principle. . . . . . . . . . . . . . . . . . . . . . . . . . . . AMC Scheduling Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Layer 1 Processing Chain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Physical Layer Processing Chain of AMC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methods of selecting the suitable MCS . . . . . . . . . . . . . . . . . . . . . . . . . . . . AMC Processing Chain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . AMC Liabilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . UEs measurement inaccuracy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Delay in CQI Reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Limited Granularity for Data Rate Selection . . . . . . . . . . . . . . . . . . . . . . . . . . NodeB Processing and Scheduling Delay. . . . . . . . . . . . . . . . . . . . . . . . . . . HSDPA HARQ Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chase Combining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Partial IR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Full IR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . HARQ Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chase Combining Performance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Operation of Partial IR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Operation of Full IR. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Scheduling Strategies Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . AMC Scheduling Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The predicted or estimated channel quality . . . . . . . . . . . . . . . . . . . . . . . . The current load of the cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The traffic priority class (e.g. streaming class or background class) . . . . . . . . . . . . QoS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Comparison of Basic Packet Scheduler Methods . . . . . . . . . . . . . . . . . . . . . . . . . Scheduling Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Serve Order . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Allocation Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . FT (Fair Throughput) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . P-FR (Proportional Fair Resources) . . . . . . . . . . . . . . . . . . . . . . . . . . . M-C/I (Maximum C/I) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Packet Scheduler in NodeB. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fairness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Opportunistic Scheduling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Minimum Throughput Guarantees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Effect of TCP/IP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Influence of Coverage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Packet Scheduling Strategies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fair Throughput Scheduling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fair Time Scheduling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Long Term C/I Scheduling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Proportional Fair Resource Scheduling . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2-70 2-72 2-72 2-72 2-74 2-74 2-74 2-74 2-76 2-76 2-76 2-76 2-76 2-76 2-78 2-78 2-78 2-80 2-80 2-80 2-82 2-82 2-82 2-82 2-82 2-84 2-84 2-84 2-84 2-86 2-86 2-88 2-90 2-92 2-92 2-92 2-92 2-92 2-92 2-94 2-94 2-94 2-94 2-94 2-94 2-94 2-96 2-96 2-96 2-96 2-96 2-96 2-98 2-98 2-98 2-98 2-98
v
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
Contents
Multi-User Selection Diversity Dynamic Scheduling . . . . . . . . . . . . . . . . Dynamic Scheduling according to best C/I . . . . . . . . . . . . . . . . . . . Scheduler Resources include: . . . . . . . . . . . . . . . . . . . . . . . . . MAC-hs Scheduler should not allocate any HSDPA Resources to a User which:
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
vi
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Contents
Example for Initial Transmission Self-decodable with 1st RM . Example for Initial Transmission-Self-decodable with 1st RM Example for Retransmission Non-self-decodable . . . . . . . Full IR . . . . . . . . . . . . . . . . . . . . . . . . . . . Partial IR . . . . . . . . . . . . . . . . . . . . . . . . . . Example for Chase Combining . . . . . . . . . . . . . . . . . Overview over transmission capacity considerations . . . . 2nd RM Parameter Calculation Self-Decodable . . . . . . . . Turbo Coding with Transparent 1st RM stage . . . . . . . . X(i), e(plus) and e(minus) Calculation. . . . . . . . . . . . Initial Setting of e and Redundancy Version r . . . . . . . . 2nd RM Bit Position Calculation Self-Decodable . . . . . . . Systematic RM . . . . . . . . . . . . . . . . . . . . . . . Parity_1 RM . . . . . . . . . . . . . . . . . . . . . . . . Parity_2 RM . . . . . . . . . . . . . . . . . . . . . . . . 2nd RM Self-Decodable Transmission followed by Full IR (1) . 2nd RM Self-Decodable Transmission followed by Full IR (2) . 2nd RM Parameter Calculation Non-Self-Decodable . . . . . 2nd RM Bit Position Calculation Non-Self-Decodable . . . . . Rate Matching Exercise . . . . . . . . . . . . . . . . . . . . Comparison between Full and Partial IR . . . . . . . . . . . . R = 3/4 . . . . . . . . . . . . . . . . . . . . . . . . . . . R = 1/2 . . . . . . . . . . . . . . . . . . . . . . . . . . . HARQ Performance Enhancement Techniques . . . . . . . . . HS-PDSCH Interleaving . . . . . . . . . . . . . . . . . . . . DTX Indication Bits . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
3-38 3-38 3-40 3-40 3-40 3-42 3-42 3-44 3-44 3-44 3-44 3-46 3-46 3-46 3-46 3-48 3-50 3-52 3-54 3-56 3-58 3-58 3-58 3-60 3-62 3-62
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
vii
Contents
HS-SCCH Power Offset . . . . . . . . . . . . NBAP HS-DSCH Info Response . . . . . . . . . . . . HS-DSCH Initial Capacity Allocation . . . . . . . . HS-SCCH Specific Information Response . . . . . HARQ Memory Partitioning Information . . . . . . . HS-PDSCH and HS-SCCH Power and Code Information HSDPA Cell Configuration - AUDIT . . . . . . . . . . . HSDPA Data Transfer Procedure (1) . . . . . . . . . . HSDPA Data Transfer Procedure (2) . . . . . . . . . . RRC Protocol Enhancement New IEs with HSDPA . . MP Mandatory Present . . . . . . . . . . . . . . MD Mandatory Default . . . . . . . . . . . . . . CV Conditional Value . . . . . . . . . . . . . . . OP - Optional. . . . . . . . . . . . . . . . . . . . Added or Reconfigured MAC-d Flow . . . . . . . . . . MAC-d Flow Identity . . . . . . . . . . . . . . . . MAC-hs Queue Id . . . . . . . . . . . . . . . . . T1 Reorder Release Timer . . . . . . . . . . . . . MAC-d PDU Size Info . . . . . . . . . . . . . . . Uplink DPCH Power Control Info and Others . . . . . . Uplink DPCH Power Control Info . . . . . . . . . . MAC-hs Reset . . . . . . . . . . . . . . . . . . . Serving HS-DSCH Radio Link Indicator . . . . . . . MAC-hs Protocol . . . . . . . . . . . . . . . . . . . . MAC-hs - co-incident CRNC and SRNC . . . . . . HS-DSCH RNTI . . . . . . . . . . . . . . . . . . MAC-hs Tasks . . . . . . . . . . . . . . . . . . . . . Iub Flow Control . . . . . . . . . . . . . . . . . . Packet Buffering and Scheduling . . . . . . . . . . TFRC Selection . . . . . . . . . . . . . . . . . . AMC . . . . . . . . . . . . . . . . . . . . . . . . HARQ . . . . . . . . . . . . . . . . . . . . . . . HS-DSCH MAC PDU . . . . . . . . . . . . . . . . . . General . . . . . . . . . . . . . . . . . . . . . . MAC-hs SDU. . . . . . . . . . . . . . . . . . . . MAC-hs Header of MAC-hs PDU . . . . . . . . . . MAC-hs Entity UTRAN Side . . . . . . . . . . . . . . Scheduling/Priority Handling . . . . . . . . . . . . HARQ . . . . . . . . . . . . . . . . . . . . . . . TFRC Selection . . . . . . . . . . . . . . . . . . MAC-d / Flow Control. . . . . . . . . . . . . . . . MAC-hs Entity UE Side . . . . . . . . . . . . . . . . . HARQ . . . . . . . . . . . . . . . . . . . . . . . Reordering Queue distribution . . . . . . . . . . . Reordering . . . . . . . . . . . . . . . . . . . . . Disassembly . . . . . . . . . . . . . . . . . . . . MAC-d . . . . . . . . . . . . . . . . . . . . . . . HS-DSCH MAC Architecture UTRAN Side . . . . . . Overall Architecture . . . . . . . . . . . . . . . . MAC-c/sh (optional) . . . . . . . . . . . . . . . . MAC-d Flow in UTRAN . . . . . . . . . . . . . . . MAC-d flow: 0 . . . . . . . . . . . . . . . . . MAC-d flow: 1 . . . . . . . . . . . . . . . . . MAC-d flow: 2 . . . . . . . . . . . . . . . . . HS-DSCH Architecture UE Side . . . . . . . . . . . Overall Architecture . . . . . . . . . . . . . . . . MAC-d Flow in UE . . . . . . . . . . . . . . . . . Reordering / Disassembly . . . . . . . . . . . MAC-d flow: 0, 1, 2. . . . . . . . . . . . . . . HS-DSCH Scheduler Functions. . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4-14 4-16 4-16 4-16 4-16 4-18 4-20 4-22 4-24 4-26 4-26 4-26 4-26 4-26 4-28 4-28 4-28 4-28 4-28 4-30 4-30 4-30 4-30 4-32 4-32 4-32 4-34 4-34 4-34 4-34 4-34 4-34 4-36 4-36 4-36 4-36 4-38 4-38 4-38 4-38 4-38 4-40 4-40 4-40 4-40 4-40 4-40 4-42 4-42 4-42 4-42 4-42 4-42 4-42 4-44 4-44 4-44 4-44 4-44 4-46
viii
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Contents
UE Scheduling . . . . . . . . . . . . . . . . . . . . . . Serving Priority Queues . . . . . . . . . . . . . . . . . Updating MAC-hs Header . . . . . . . . . . . . . . . . Scheduling new Transmission and retransmission . . . . Determining RV . . . . . . . . . . . . . . . . . . . . . NodeB HARQ . . . . . . . . . . . . . . . . . . . . . . . . HARQ Entity . . . . . . . . . . . . . . . . . . . . . . . HARQ Process . . . . . . . . . . . . . . . . . . . . . . Error Handling in HARQ . . . . . . . . . . . . . . . . . HS-DSCH Bit Rate Measurement. . . . . . . . . . . . . . . UE MAC-hs Operation . . . . . . . . . . . . . . . . . . . . UE HARQ . . . . . . . . . . . . . . . . . . . . . . . . HARQ Entity . . . . . . . . . . . . . . . . . . . . . HARQ Process . . . . . . . . . . . . . . . . . . . . Transmitter and Receiver Stall Avoidance. . . . . . . . . . . Timer-Based Scheme . . . . . . . . . . . . . . . . . . Window Bases Scheme . . . . . . . . . . . . . . . . . HARQ Activity Scheme . . . . . . . . . . . . . . . . . . UE Reordering Entity . . . . . . . . . . . . . . . . . . . . . Parameters. . . . . . . . . . . . . . . . . . . . . . . . State Variables . . . . . . . . . . . . . . . . . . . . . . Timer T1 . . . . . . . . . . . . . . . . . . . . . . . . . Receiver Window. . . . . . . . . . . . . . . . . . . . . Timer and / or Window Based Stall Avoidance . . . . . . . . Window Based Scheme . . . . . . . . . . . . . . . . . Timer Bases Scheme. . . . . . . . . . . . . . . . . . . UE MAC-hs Other Functions . . . . . . . . . . . . . . . . . Disassemble Entity . . . . . . . . . . . . . . . . . . . . MAC-hs Reset . . . . . . . . . . . . . . . . . . . . . . HS-PDSCH and DL TrCH Information. . . . . . . . . . . . . Activation Time . . . . . . . . . . . . . . . . . . . . . . Downlink Transport Channel Type . . . . . . . . . . . . HS-DSCH Data Frame on Iub/Iur (HS-DSCH FP) . . . . . . . CmCH-PI (Common Transport Channel Priority Indicator). MAC-d PDU Length . . . . . . . . . . . . . . . . . . . NumOfPDU . . . . . . . . . . . . . . . . . . . . . . . User Buffer Size . . . . . . . . . . . . . . . . . . . . . MAC-d PDU . . . . . . . . . . . . . . . . . . . . . . . Control Frame Format for HS-DSCH FP . . . . . . . . . . . HS-DSCH Capacity Request Control Frame . . . . . . . HS-DSCH Capacity Allocation Control Frame . . . . . . . . . Maximum MAC-d PDU Length . . . . . . . . . . . . . . HS-DSCH Credits . . . . . . . . . . . . . . . . . . . . HS-DSCH Interval . . . . . . . . . . . . . . . . . . . . HS-DSCH Repetition Period . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4-46 4-46 4-46 4-46 4-46 4-48 4-48 4-48 4-48 4-50 4-52 4-52 4-52 4-52 4-54 4-54 4-54 4-54 4-56 4-56 4-56 4-56 4-56 4-58 4-58 4-58 4-60 4-60 4-60 4-62 4-62 4-62 4-64 4-64 4-64 4-64 4-64 4-64 4-66 4-66 4-68 4-68 4-68 4-68 4-68
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
ix
Contents
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
Version 1 Rev 0
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
HSDPA Overview
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-1
Version 1 Rev 0
HSDPA Overview
1-2
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Objectives
Version 1 Rev 0
Objectives
On completion of this chapter the Student will be able to: Describe the important changes and characteristics of HSDPA State the new channels of HSDPA and how they operate in principle Describe the extended UTRAN protocol stack with HSDPA State the advantages and disadvantages of HSDPA and future enhancements Describe the concept of HSUPA
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-3
Version 1 Rev 0
Feature Study
The HSDPA feature in 3GPP Release 5 is the result of a study carried out in the Release 4 time frame. This study considered a number of techniques in order to provide instantaneous high speed data in the downlink. Some of the considerations and goals taken into account in the evaluation of the different techniques were: To focus on the streaming, interactive and background services: services which require a constant and/high throughput or low error rate. To prioritise urban environments and then indoor deploymens (but not limited to these environments and supporting full mobility). To enable compatibility with advanced antenna and receiver techniques: transmit and receive diversity methods are used and might be enhanced To take into account User Equipment processing time and memory requirements: UEs limitations are taken into account by the network To minimize changes on existing techniques and architectures: modest changes to NodeB hardware and UTRAN software
1-4
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-5
Version 1 Rev 0
HSDPA Targets
HSDPA Targets
Higher Data Rates for Streaming-, Interactive- and Background Services
HSDPA is a feature based on a downlink shared channel that allows user net-data rates of up to 10 Mbit/s. It is designed to support services that require instantaneous high rates in the downlink and lower rates on the uplink. This feature also decreases the level of retransmissions (at the radio link and hence higher layers), in turn allowing the reduction of delivery time. Examples of end-user services targeted by HSDPA are internet browsing and video on demand.
1-6
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
HSDPA Targets
Version 1 Rev 0
HSDPA Targets
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-7
Version 1 Rev 0
HSDPA Characteristics
HSDPA Characteristics
High Speed Downlink Packet Access comes with certain characteristics which distinguishes it clearly from Rel. 99 UTRAN.
Modulation Types
QPSK is already known from Rel. 99 UTRAN. Besides QPSK, HSDPA incorporates the 16-QAM modulation to increase the peak data rates for users served under favorable radio conditions. Support for QPSK is mandatory, though the support for 16-QAM is optional for the network and UE. 16-QAM (Quadrature Amplitude Modulation) was newly introduced in Rel .5 with HSDPA. It is a so called higher order modulation which basically doubles the data rate in good radio conditions. Thus it increases the spectrum efficiency of WCDMA.
Hybrid ARQ
HARQ functionality combines retransmission with the original transmissions. There a two different ways for HARQ to operate. Either identical retransmission of the data block are sent or retransmission are not identical and differ in data and parity bits compared to the original transmission. The first method is known as chase combining and, the latter as incremental redundancy . HARQ operates on an N-channel Stop and Wait principle.
1-8
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
HSDPA Characteristics
Version 1 Rev 0
HSDPA Characteristics
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-9
Version 1 Rev 0
QPSK
Each symbol corresponds to 2 consecutive input bits. The four symbols are represented by different phase shifts in the I/Q plane.
16QAM
Each symbol corresponds to four consecutive input bits. Thus the data rate can be doubled with 16-QAM compared to QPSK. The 16 symbols are represented in the I/Q plane by different phase shifts and amplitudes. In 16-QAM modulation the symbol value is determined by phase and amplitude. Compared to that, in QPSK the phase is only modulated and variation in amplitude have only minor influence on the decision space in the I/Q diagram. However with 16-QAM the decision space is heavily influenced by amplitude variations, thus higher constraints are put on the transmitter linearity. Note, a more accurate phase estimate is necessary with 16-QAM since constellation points have smaller differences in phase domain compared to QPSK. Note: The number of constellation points in the I/Q-diagram can be calculated with 2m, where m represents the number of bits or chips per modulation symbol. QPSK modulation has four constellation points in the I/Q-diagram: 2^m = 4 m = 2. 16-QAM modulation has 16 constellation points in the I/Q-diagram: 2^m = 16 m = 4
1-10
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-11
Version 1 Rev 0
HSDPA Rel. 5
In HSDPA the spreading factor for the user plane is fixed to 16. Thus up to 15 physical channels can be allocated at maximum per UE. QPSK: The downlink slot format # 0 allows for 320 bits/slot and per physical channel. This results in 960 bits/ 2 ms subframe. The maximum physical data rate considering 15 spreading codes @ sf16 is: 7.2 Mbit/s. 16-QAM The downlink slot format #1 allows for 640 bits/slot and per physical channel. This results in 1920 bits/ 2 ms subframe. The maximum physical data rate considering 15 spreading codes @ sf16 is: 14.4 Mbit/s.
1-12
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
Uplink
1 bits/symbol x [3.84 Mcps / (4 chips/symbol)] x (6 OVSF's) = 5.76 Mbits/S
HSDPA Rel. 5
QPSK
2 bits/symbol x [3.84 Mcps / (16 chips/symbol)] x (15 OVSF's) = 7.2 Mbits/S
16-QAM
4 bits/symbol x [3.84 Mcps / (16 chips/symbol)] x (15 OVSF's) = 14.4 Mbits/S
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-13
Version 1 Rev 0
1-14
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-15
Version 1 Rev 0
Transport Channel:
HS-DSCH (High Speed Downlink Shared Channel) The HS-DSCH is the actual transport resource carrying the packet data of the user applications. As it also follows the shortened TTI of 2 ms, it allows for short round trip delay in the operation between NodeB and UE. The 2 ms TTI is short when compared to 10, 20, 40 or 80 ms TTIs supported by Rel. 99 and Rel. 4 transport channels. HS-DSCH describes the physical layer processing by MAC-hs of a HSDPA transport block. Dynamic part: TB size = TBS size {1 to 200 000 bits with 8 bit granularity}; modulation scheme {QPSK, 16-QAM}; redundancy / constellation version {1 8}. Static part: TTI {2 ms for FDD}; type of channel coding {turbo coding}; mother code rate {1/3}, CRC size {24 bits} No semi-static attributes are defined for HS-DSCH.
Physical Channels
High Speed Shared Control Channel (HS-SCCH) The HS-SCCH has a fixed spreading factor of value 128 and is configured only in the downlink direction. It also adopts the shortened TTI of 2 ms. In theory, up to 127 HS-SCCHs can be configured in a cell. However, the UE is required only to be able to listen to up to four HS-SCCH in parallel. The HS-SCCH allows the efficient sharing of one or more HS-PDSCHs among different users. Nevertheless every UE needs to be informed on the DCCH via RRC messages about the specific HS-SCCH-set that it shall monitor in order to receive data via the HS-PDSCHs. High Speed Physical Downlink Shared Channel (HS-PDSCH) The HS-PDSCH has a fixed spreading factor of value 16. Thus, it provides for multicode operation using up to 15 channelization codes in parallel. Of course the UE must support the use of up to 15 channelization codes which depends on its category. The HS-PDSCH adopts the shortened TTI of 2 ms. Uplink Dedicated Control Channel Associated with HS-DSCH Transmission (Uplink HS-DPCCH). The HS-DPCCH has a fixed spreading factor of value 256 and is only configured in uplink direction. The HS-DPCCH also follows the shortened TTI of 2 ms. Its purpose is to provide feedback information about the downlink receive quality and whether the packet data received by the UE are error-free or need to be retransmitted. Thus the NodeB is quickly notified of unsuccessful transmissions and/or changing radio conditions in downlink direction.
1-16
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
Transport Channels
Physical Channels
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-17
Version 1 Rev 0
MAC-hs
MAC-hs located in NodeB receives user packets from the SRNC. MAC-hs is responsible for transmission and in case of erroneous reception also for retransmission of user packets. The retransmission of user packets is commanded by NodeBs MAC-hs which represents the fundamental change in HSDPA compared to Rel. 99 or Rel. 4. In legacy UMTS releases retransmission are always performed between the RLC peers in UE and SRNC. Of course in downlink the SRNC retransmits the user packets if the UE indicates a Nack. The UE sends feedback information about the downlink channel quality in regular intervals to the NodeB.
New Channels
Two UEs receiving user packets via the HS-DSCH. The physical channels in uplink and downlink are only necessary to firstly signal which UE shall decodea specificHS-DSCH. This is indicated by the UE-id in the HS-SCCH. Secondly after decoding of the user packet, the respective UE has to signal the successful or unsuccessful transmission to the MAC-hs entity in NodeB via the so called Ack/Nack description. HSDPA employs the uplink HS-DPCCH to signal the downlink reception quality and the Ack/Nack description. The downlink reception quality informs the NodeB about the current radio condition which serves as a vital input for NodeBs MAC-hs. The MAC-hs is therefore able to derive the proper modulation scheme and code rate for transmission and retransmission. This process is denoted as AMC in HSDPA.
UE Scheduling
The NodeB also contains a scheduling/priority handling function which determines whether a new transmission or retransmission shall be performed. The green and blue line represent the changing downlink channel quality reported by UE in the uplink. One scheduling method which could be implemented in NodeB is to serve each UE according to the reported downlink channel quality and therefore always exploit best radio conditions. This method maximizes user throughput as it allows to use 16-QAM and an aggressive code rate if C/I is high. Another option is to serve each user proportionally fair despite unfavorable downlink radio conditions. This benefits especially UE 2 which indicates for several TTIs a bad radio quality, but this second method assures at least a minimum guaranteed throughput for UE 2.
1-18
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-19
Version 1 Rev 0
3.
4.
5.
1-20
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
HSDPA and DPCH Operation HSDPA Setup HSDPA Setup via the DPCH
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-21
Version 1 Rev 0
Uplink
The maximum number of DCHs and the maximum bit rate are dependent on UE radio access capabilities. This combination is required in case HS-DSCH(s) are configured. The maximum number of DCHs and the maximum channel bit rate are dependent on UE radio access capabilities This combination is required by a UE in CELL_DCH state to be able to read the SFN of a neighbouring cell and support "SFN-CFN observed time difference" and "SFN-SFN observed time difference" measurements while HS-DSCH(s) are configured.
Downlink
DPCCH + one or more DPDCH + one or more HS-SCCHs + zero, one or more HS-PDSCHs PCCPCH (neighbour cell) + DPCCH + one or more DPDCH + one or more HS-SCCHs + zero, one or more HS-PDSCHs
One HS-DSCH coded into a single CCTrCH + one or more DCH coded into a single CCTrCH BCH (neighbour cell) + one or more DCHs + one HS-DSCH
1-22
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
HSDPA and DPCH Operation HSDPA Setup HSDPA Transmission and Retransmission
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-23
Version 1 Rev 0
H-RNTI
The H-RNTI (High Speed Radio Network Transaction Identifier) is firstly used to identify the UE which shall receive data from a HS-DSCH. Therefore the UEs H-RNTI is implicitly encoded in the HS-SCCH. Implicitly encoded means that part 1 of HS-SCCH uses H-RNTI as input for UE specific masking. For part 2 of HS-SCCH, the H-RNTI is used for UE specific CRC attachment. Thus the UE is able to distinguish which out of a maximum of four HS-SCCHs contains decoding information for the following HS-PDSCHs in the sub-sequent HS-DSCH subframe. This means on the other hand, that the actual HS-DSCH TB does not contain any UE-id. Thus the HS-DSCH resource allocation in HSDPA is entirely done without higher layer involvement. The reasons for this are twofold. Leaving the resource assignment at layer 1 speeds up the entire process (avoiding higher layer processing delays) and avoids to transmit UE specific RNTIs and sequence numbers. The timely relationship between HS-SCCH and HS-DSCH avoids the need of attaching an UE-id within HS-DSCH transport block to identify the recipient. For each HS-DSCH TTI of 2 ms, each related HS-SCCH occurring 2 slots ahead of HS-DSCH carries HS-DSCH related downlink signaling for one UE only. The H-RNTI encoded in the HS-SCCH identifies therefore an UE having a HS-PDSCH assignment within a cell. Therefore the UE needs at first to decode the HS-SCCH (part 1) before it can attempt to demodulate the allocated HS-PDSCHs. The HS-SCCH signaling message is divided into two parts, with part 1 containing the time critical information on channelization code set and modulation scheme. Part 2 of HS-SCCH consists of transport block size and HARQ-related (indicating new transmission or retransmission, HARQ process id, redundancy and constellation version). A 16-bit UE specific CRC is computed over part 1 and part 2 and attached to part 2. The UE ID is only implicitly included in the CRC. In order to allow the UE quickly to determine whether HS-PDSCHs are going to be allocated, the part 1 alone of HS-SCCH is scrambled with the H-RNTI as well. If more than one UE shall be served within a subframe, more than one HS-SCCH needs to be transmitted as one HS-SCCH can only allocate HS-DSCH resources for one UE only. However code multiplexing may be limited between 2-3 users per subframe as it becomes less efficient and processing is quite high for NodeB.
1-24
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-25
Version 1 Rev 0
DL-HSPDSCH-Info
DL-HSPDSCH-Info
The downlink HS-PDSCH information consists of HS-SCCH Info and measurement feedback info. (Measurement feedback info will explain at a later stage in chapter 3.)
HS-SCCH Info
Downlink Scrambling Coding This information element informs the UE about the DL scrambling code to be applied for HS-PDSCHs and HS-SCCHs. Default is same scrambling code as for the P-CPICH. If dl-Scrambling Code IE is present, then secondary scrambling code is used in the cell under which HSDPA operates. HS-SCCH Channelization Code info This information element contains the maximum number of HS-SCCH codes that can be assigned to a UE. In Rel. 5 this parameter is set to four. HS-SCCH Codes This information element tells the UE the specific channelization codes used for the HS-SCCH set. As a HS-SCCH channel has a fixed spreading factor of 128, the channelization code ranges from 0127. The channelization code C(ch,128,0) cannot be used as the related branch is already occupied by the P-CPICH and P-CCPCH. The same is valid for C(ch,16,0). And also if secondary scrambling codes are used. NOTE HSDPA may operate on a different frequency than legacy UMTS traffic so almost the entire resources of the second carrier could be dedicated to HSDPA.
1-26
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
DL-HSPDSCH-Info
Version 1 Rev 0
DL-HSPDSCH-Info
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-27
Version 1 Rev 0
CELL_DCH ONLY
This RRC state is identical to the CELL_DCH in Rel. 99. Here we would like to highlight that whenever the UE moves from a HSDPA capable cell to a non-capable cell, the radio bearer(s) have to be reconfigured to e.g. a ordinary 384 kbit/s bearer known since Rel. 99.
CELL_FACH
The advantage of being in CELL_FACH state is twofold. Firstly the UE does not have to monitor the HS-SCCH code-set anymore which extends its battery capacity. Secondly the UTRAN saves channelization code resources as the UE does not longer occupy DCHs. Through traffic volume measurements the UE reports its RLC transmit buffer occupancy and a rise in uplink or downlink (RLC transmit buffer in SRNC) above a certain threshold triggers the reconfiguration of UE into CELL_DCH and possibly into HSDPA standby mode.
1-28
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-29
Version 1 Rev 0
1-30
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-31
Version 1 Rev 0
1-32
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
HSDPA Packet Switched Protocol Stacks HSDPA Control Plane Protocol Stack
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-33
Version 1 Rev 0
1-34
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-35
Version 1 Rev 0
1-36
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
P-CPICH
HS-PDSCH
HS-PDSCH
HS-PDSCH
HS-PDSCH
HS-PDSCH
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-37
Version 1 Rev 0
1-38
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
n-th DPCH: DPDCH/D + DPCCH/D (SF = 4.....512) HS-PDSCH m HS-PDSCH HS-PDSCH HS-PDSCH HS-PDSCH
m x 256 Chips
HS-PDSCH
T0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-39
Version 1 Rev 0
1-40
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-41
Version 1 Rev 0
1-42
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
Type I
Type II
Type III
L1 Soft Combining
Link Adaptation
Incremental Redundancy
Full IR
Partial IR
Chase Combining
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-43
Version 1 Rev 0
AMC
The MAC-hs in NodeB is applying adaptive modulation and coding only for the HS-DSCH. Note, the detailed function of AMC is explained in chapter 2.
Multicode Operation
Both HS-DSCH and DCH may use more than one channelization code, but for HS-DSCH the UE can go for the maximum of either 5, 10 or 15 parallel channelization codes with SF = 16 depending on the UEs category. Therefore HSDPA uses the code multiplexing extensively compared to DCH operation.
1-44
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-45
Version 1 Rev 0
TrCH Multiplexing
The HS-DSCH is a transport channel with only one TB (transport block) per TTI per UE, so there is no need for static rate matching, transport block concatenation or balancing of quality. This complex matter shall be explained here in detail now:
Soft(er) Handover
For HS-DSCH there is no macro diversity possible as it is a common transport channel. Of course DCH supports SHO.
1-46
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-47
Version 1 Rev 0
Advantages of HSDPA
Reduced Latency for retransmission With MAC-hs installed in NodeB, the uplink feedback channel and the shortened TTI of 2 ms, reception of erroneous packet data results in a faster retransmission. On top of this, no transmission is wasted due to the use of incremental redundancy or chase combining. Higher DL Throughput and Peak Throughput Rates up to 14 Mbit/s As HSDPA makes extensive use of code multiplexing in conjunction with higher order modulation, the downlink throughput can be increased beyond 10 Mbit/s. If the radio conditions are favorable, AMC allows a very aggressive code rate of 1:1, thus offering peak data rates up to 14 Mbit/s. Support of Non-Real Time and Streaming Applications Through a sophisticated packet scheduler implementation in the NodeB it is possible to manage stringent QoS requirements of streaming applications and of course less time critical services as well. The packet scheduler is able to adjust the code rate, modulation type, number of channelization codes and the transmission power within every subframe. Thus the NodeB is highly flexible in assuring various QoS requirements of subscribers services. High Spectrum and Code Efficiency due to 16-QAM Spectrum efficiency increases when the modulation scheme changes from QPSK to 16-QAM as literally more kbit/s respectively Mbit/s can be transmitted in the 5 MHz bandwidth. Code efficiency increases when the UE is able to dynamically make use of the available channelization codes. In downlink, the WCDMA power control dynamics is in the order of about 20 dB. It is limited by the intra-cell and inter-cell interference compared to the uplink power control dynamics of about 70 dB. This means that for a particular user, very close to the NodeB, the DL power control cannot reduce the power maximally, and on the other hand reducing power beyond 20 dB dynamics may only have marginal impact on the downlink capacity. HSDPA now utilizes this available DL capacity though link adaptation and through AMC. The higher Ec/Io signal quality (when user is close to NodeB) is exploited by selecting a very aggressive code rate and 16-QAM. Vice versa, if radio conditions become worse, HSDPA switches down to the more robust QPSK, applies lower code rate (meaning more protection) but may use increasing number of channelization codes to support the users QoS demands (e.g. guaranteed throughput rate).
1-48
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-49
Version 1 Rev 0
Disadvantages of HSDPA
Disadvantages of HSDPA
Higher Processing power in UE
The higher throughput rates in the downlink require that the UEs processors and DSPs are designed accordingly. UEs battery consumption will increase, too. This imposes serious challenges on the UE vendors.
1-50
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Disadvantages of HSDPA
Version 1 Rev 0
Disadvantages of HSDPA
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-51
Version 1 Rev 0
Disadvantages of HSDPA
Disadvantages of HSDPA
No Macro-Diversity for HS-PDSCH, HS-SCCH
As HS-PDSCH and HS-SCCH are shared channels, no soft handover is possible and thus the performance decreases particularly at cell edge. Nevertheless mobility procedures like intra-NodeB and inter NodeB cell changes are supported which will mitigate this disadvantage.
1-52
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Disadvantages of HSDPA
Version 1 Rev 0
Disadvantages of HSDPA
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-53
Version 1 Rev 0
Beamforming
Beamforming makes use of adaptive antennas and can therefore provide a better C/I to UEs in the downlink. At the same time beamforming allows re-use of scarce downlink channelization codes as the individual UEs are separated in space and possibly through different downlink scrambling codes, thus making use of secondary scrambling codes. The signals toward different UEs from the same cell are typically transmitted under the same primary scrambling code and separated by means of orthogonal channelization codes. However, some of the beams may be transmitted under a secondary scrambling code with its associated channelization code tree, thereby increasing the resources in the cell. Note that the loss of the reduced orthogonality between primary and secondary scrambling code can be partly mitigated in the case of beamforming by splitting the cell into multiple scrambling code regions, so the spatial isolation between beams using different scrambling codes helps to compensate the lack of orthogonality.
Transmit Diversity
The downlink capacity could be improved by using receive antenna diversity in the UE. However for small and cheap mobiles it is not feasible to use two antennas and receiver chains. Therefore, the WCDMA standard already supports the use of base station transmit diversity in Rel. 99. There are two modes: open loop (TSTD and STTD) and closed loop mode (mode1 with phase adjustment only and mode 2 with phase and amplitude adjustment). The open loop mode simply transmits the coded information from two antennas, but on the diversity antenna the bits are time reversed and complex conjugated. The STTD method provides two kinds of diversity. The physical separation of the antennas provides space diversity and the time difference derived from a bit-reversing process provides for time diversity, thus the decoding in the receiver becomes more reliable. The closed loop mode can only be applied to the downlink channel, if there is an associated uplink channel. Thus this mode can only be used with dedicated channels (DPCH, PDSCH or HS-PDSCH with an associated uplink DPCCH).
MIMO
With MIMO (Multiple Input Multiple Output) at the transmitter, x independent data streams are transmitted out of the x antennas on the same frequency band. At the receiver, each antenna receives all of the transmitted sub-streams superimposed, not separately. If multipath scattering is sufficient, these x data streams have different spatial signatures to each of the e.g. p receive antennas and they are separable, the signals arrive with different phases. When a transmitter has x antennas and the receiver has p antennas, the link speed increases linearly with min (x,p) given the same power and bandwidth budget.
1-54
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-55
Version 1 Rev 0
Preview to HSUPA
Preview to HSUPA
The aim of the HSUPA is to enhance the uplink DCH operation and performance, using several techniques in order to support services like video-clips, multimedia, e-mail, telematics, gaming and video-streaming. A study showed that various techniques such as Node-B controlled scheduling, shorter TTI and a hybrid ARQ layer in the Node-B can enhance the uplink packet transfer performance significantly compared to Release-99/Rel-4/Rel-5. The targeted improvements are a higher air interface capacity in the uplink and a better end user experience by increasing both the cell throughput and the coverage of higher bit rates in uplink. There are fundamental differences between uplink and downlink data transmission, making it impossible to simply introduce the same HSDPA solutions for the uplink.
1-56
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Preview to HSUPA
Version 1 Rev 0
Preview to HSUPA
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-57
Version 1 Rev 0
Shorter TTI
Introducing a shorter TTI in the uplink reduces the HARQ retransmission delay. The motivation for the 2 ms TTI introduced in downlink HSDPA was due to the lack of power control being employed for the HS-PDSCH transmission. Such reasoning is not applicable in the uplink as disabling uplink power control would reduce the system capacity. The challenge for the 2 ms TTI stems from the resulting uplink range with limiting power resources in the UE. With an equal amount of data per TTI, it is possible to transmit less energy during 2 ms than e.g. compared to 10 ms. Additionally, the interleaving gain is reduced at 2 ms TTI. Thus, even if shorter TTIs were used due to the advantage for the HARQ, for the cell edge operation, the use of 10 ms TTI might be still be necessary in order to make the uplink more robust at low Ec/No conditions.
1-58
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-59
Version 1 Rev 0
1-60
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
1-61
Version 1 Rev 0
1-62
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
HSDPA Principles
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-1
Version 1 Rev 0
HSDPA Principles
2-2
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Objectives
Version 1 Rev 0
Objectives
After this Lecture the Student will be able to: Describe the coding chains of HS-DSCH, HS-SCCH and HS-DPCCH State the timing relations between DPCH and HS-DSCH and HS-DPCCH Describe the implications of using 16-QAM State the advantages and liabilities of AMC in HSDPA Describe difference between chase combining, partial IR and full IR
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-3
Version 1 Rev 0
Part 1
Part 1 consists of the time critical HS-PDSCHs channelization code set and modulation scheme. Both parameters are vital for the physical decoding of HS-PDSCHs. This function is indicated by the arrow pointing from part 1 towards the HS-PDSCH subframe. In order to distinguish different UEs the H-RNTI is encoded into the part1 bits. The 16 bit value for H-RNTI is actually not present inside the part 1, but it is just masked with the part 1 bits. This method allows the layer 1 alone to determine if the HS-SCCH subframe contains a HS-DSCH resource assignment without layer 3 involvement. It can be assumed that invoking layer 3 for determining resource allocation would consume more time and processing. Therefore this method using physical layer is very efficient.
Part 2
Part 2 consists of several parameters. The H-RNTI is encoded together in part 2 and part 1 bits. Transport-block size information is used to derive the actual transport block size. The transport block size varies from 137 to 27,952 bits. Hybrid ARQ process information is signaled by RRC to maintain several transmission instances at the same time. Redundancy version: On each retransmission of a block, the set of bits that are actually transmitted may be changed for example, the set of parity bits chosen for transmission on the second attempt may be different to the set chosen on the first attempt. These different versions of the coding format for the data block are called redundancy versions. In order to attempt a decoding, the receiver needs to know the redundancy version used when each transmission attempt is made (in order to perform the HARQ combining process). Constellation version: only used in case of 16-QAM. An additional parameter which can be varied between retransmission of a block is the bit-to-constellation mapping (also referred to as Constellation Re-arrangement for details please refer to slide 16-QAM Constellation Rearrangement). New data indicator allows the UE to distinguish between initial transmission or retransmission within a HARQ process. Encoded UE Identity comprises the H-RNTI which is attached in the CRC checksum.
2-4
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-5
Version 1 Rev 0
2-6
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
80 Part 1 Bits
H-RNTI Encoding
40 Part 2 Bits
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-7
Version 1 Rev 0
2-8
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
H-RNTI Encoding
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-9
Version 1 Rev 0
2-10
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-11
Version 1 Rev 0
HARQ Process ID
The part 2 of HS-SCCH informs the individual UE what HARQ process is currently being served. The HARQ process id identifies an ongoing TB transmission which remains active in both NodeB and UE until the TB has been successfully decoded by UE and the NodeB received an Ack for it. As long as the NDI bit (new data indicator) of a certain HARQ process does not toggle, every (re-) transmission is targeted for the same HARQ process. As Layer 1 soft-combining takes place per HARQ process, the UEs receiver is informed that a combined decoding out of original transmission and retransmission(s) should be performed. The fixed timing relation between downlink HS-DSCH TB transmission and sub-sequent uplink acknowledgment on HS-DPCCH makes the use of sequence numbering obsolete. The effect of this is twofold. Firstly no higher layer in UE needs to be involved to determine whether the sequence number is correct or incorrect which keeps the entire process very hardware oriented and therefore very fast (avoiding processing delays of higher layers). Secondly the sequence number does not even have to be transmitted reducing unnecessary overhead signaling which could be prone to errors too.
2-12
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
les
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-13
Version 1 Rev 0
2-14
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-15
Version 1 Rev 0
Preconditions:
UE-1 is on 16-QAM and thus needs to obtain beside the RV also the CV for bit rearrangement (e.g. in case of retransmission). UE-1 might be very close to the NodeB and is exposed to a good C/I respectively high Ec/No in downlink. UE-2 has less favorable channel condition and is therefore on QPSK.
HS-SCCH Signaling
The HS-SCCH indicates the UE-identity and the modulation type used for HS-DSCH transmission in the subsequent HS-DSCH subframe. Although the HARQ process info, RV and CV and NDI (new data indicator) are signaled in the second part of HS-SCCH, we drew them directly above each HS-DSCH subframe. This should enhance the understanding of the picture. Note: NDI does not get incremented for retransmission..
2-16
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-17
Version 1 Rev 0
HS-SCCH-set Decoding
The graphic demonstrates that both UEs have to decode their assigned HS-SCCH-set first, before they can attempt to decode the HS-PDSCHs. For simplicity reasons, both UEs have the same HS-SCCH-set assigned. A HS-SCCH-set consists of up to a maximum of four HS-SCCH channelization codes C(ch,128,1) C(ch,128,127). As depicted, only one of the four HS-SCCHs contains valid information per UE. This is indicated by the appropriate colour coding for each UE. All the information necessary for demodulating the related HS-PDSCH subframe which follows always 2 slots later is transmitted to the UEs within part 1 of each HS-SCCH. It can be seen that every HS-SCCH is logically divided into two parts. The second part contains the necessary information on how to decode the HS-DSCH. So part 1 and part 2 serve different purpose. Part 1 allows the demodulation of the HS-PDSCH subframe and part 2 is responsible for layer 2 decoding of the HS-DSCH. Note, that part 2 is fully covered in chapter 2 Note: The HS-SCCH is intended for the very UE once it recognizes its UE-id inside part1 of the HS-SCCH subframe. If the UE detected consistent control information intended for this UE in the immediately preceding subframe, it is sufficient to only monitor the same HS-SCCH used in the immediate succeeding subframe. In the graph this is indicated by e.g. user 1 in slot 0 to slot 6 where user 1 gets two consecutive valid HS-SCCHs. Therefore UE 1 only needs to decode the same HS-SCCH from slot 3 onwards.
HS-DSCH Demodulation
If a UE detects that one of the monitored HS-SCCHs contains its UE-id and consistent control information intended for this UE, the UE shall start receiving the HS-PDSCHs. Consistent control information hereby means that modulation scheme and HS-PDSCH channelization code-set info are valid according to the UEs capability. The UE has about one slot duration time after receiving part 1 to prepare for HS-PDSCHs reception. As already mentioned, the UE indicates via the category parameter if it supports up to 5, 10 or 15 HS-PDSCH channelization codes in parallel. The color coding used in the figure for the HS-SCCH and their related HS-DSCH shows that HSDPA allows for time multiplexing and code multiplexing of the HS-PDSCHs. Time multiplexing means that user 1 and user 2 get the HS-PDSCHs assigned one after the other in different subframes. Code multiplexing or multicode operation means that several user, here user 1 and user 2, use different HS-PDSCHs within the same subframe. The various HS-PDSCHs are separated by different channelization codes.
2-18
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-19
Version 1 Rev 0
2-20
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
2-21
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
A HS-PDSCH is transmitted either with QPSK (mandatory to be supported) or 16-QAM modulation. The symbol rate of HS-PDSCH is obtained by following calculation: 38400 chips / (16 chips / symbol) = 2400 symbols / (10 ms) = 240 ksps. Within a HS-DSCH subframe all the allocated HS-PDSCHs per user must use the same modulation type, e.g. 16-QAM. However for a different user in the same HS-DSCH subframe the remaining HS-PDSCHs can be transmitted using a different modulation type, e.g. QPSK. This is code multiplexing. The potential drawback is that the processing requirement (and delay) for the NodeB becomes quite high when doing extensive multi code operation. In QPSK, each symbol represents 2 bits, whereas in 16-QAM each symbol corresponds to 4 bits. The maximum gross data rate with the maximum number of channelization codes for HS-PDSCH can be calculated like this: QPSK: 480 kbit/s x 15 physical channels = 7.2 Mbit/s 16-QAM: 960 kbit/s x 15 physical channels = 14.400 Mbit/s Taking the maximum transport block size which can be signaled via HS-SCCH HSDPA: Practical maximum transport channel throughput rate: 27952 bits / 2 ms = 13.976 Mbit/s NOTE All relevant signalling information is transmitted in the associated HS-SCCH i.e. the HS-PDSCH does not carry any signalling information on how to decode the HS-DSCH TB.
2-22
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-23
Version 1 Rev 0
Example:
The picture shows 8 consecutive codes reserved for HS-DSCH, starting at C(ch,16,8) and ending with C(ch16,15). Additionally, HS-SCCH codes with SF = 128 need to be allocated for the HS-SCCH-set. The HS-SCCH-set can comprise up to 4 HS-SCCHs thus giving the C-RNC a wider flexibility in sharing the OVSF tree between Rel. 5 and legacy release UEs. The HS-DSCH channelization-code-set information signalled over HS-SCCH is mapped as follows: The OVSF codes shall be allocated in such a way that they are positioned in sequence in the code tree. That is, for p multicodes at offset y the following codes are allocated: C(ch,16,y) C(ch, 16,y+p-1) For the example presented here, the parameter y and p have to be set like this: y = 8 and p = 8.
2-24
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-25
Version 1 Rev 0
2-26
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
Subframe 0
Subframe 1
Subframe 2
Subframe 3
Subframe 4
Subframe 5
Subframe 6
Subframe 0
Subframe 1
Subframe 2
Subframe 3
Subframe 4
Subframe 5
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-27
Version 1 Rev 0
HARQ-Ack
The Ack/Nack transmission reflects the results of the CRC check after the packet decoding and also possible combining with former retransmission.
2-28
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-29
Version 1 Rev 0
2-30
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-31
Version 1 Rev 0
2-32
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-33
Version 1 Rev 0
Spreading
Spreading is applied to the physical channels. It consists of two operations. The first is the channelization operation, which transforms every data symbol into a number of chips, thus increasing the bandwidth of the signal. The number of chips per data symbol is called the spreading factor. The second operation is the scrambling operation, where a scrambling code is applied to the spread signal.
Gain Factors
After channelization, the real-valued spread signals are weighted by gain factors, (c) for DPCCH, 2(d) for all DPDCHs and 2(hs) for HS-DPCCH (if one is active). Note: A maximum of one CCTrCH is allowed for one UE on the Uplink. It can be either one CCTrCH of dedicated type or common type.
2-34
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-35
Version 1 Rev 0
2. 3.
2-36
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-37
Version 1 Rev 0
2-38
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-39
Version 1 Rev 0
Uplink HS-DPCCH
A HS-DPCCH sub-frame starts m x 256 chips after the start of an UL DPCH frame. The UL DPCH corresponds to the DL DPCH frame from the HS-DSCH serving cell. The value m is calculated as: m = (T(TX_diff)/256 ) + 101
T(TX_diff)
T(TX_diff) is the difference in chips (T(TX_diff) = 0, 256, ....., 38144), between a) the transmit timing of the start of the related HS-PDSCH subframe (see HSDPA Downlink Timing in chapter 1) b) the transmit timing of the start of the downlink DPCH frame from the HS-DSCH serving cell that contains the beginning of the HS-PDSCH subframe Therefore T(TX_diff) is the resulting time difference between firstly the time relation of HS-SCCH and DL DPCH and secondly the time relation between HS-SCCH and the beginning of the HS-PDSCH subframe. Note: At any one time, m therefore takes one of a set of five possible values according to the transmission timing of HS-DSCH sub-frame timings relative to the DPCH frame boundary. The UE and Node B shall only update the set of values of m in connection to UTRAN reconfiguration of downlink timing. Example calculation for the second HS-DSCH subframe and its related uplink HS-DPCCH: T(TX_diff) = [t(start of HS-DSCH-subframe) t(DPCH,n) versus start of HS-SCCH)] T(TX_diff) = [(5120+7680) 2560] = 10240 m = (T(TX_diff) / 256) + 101 m = (10240 / 256) + 101= 141 x 256 chips = 36096 chips Rule:101 x 256 chips + 1024 chips + T(TX_diff) = 19200 chips + 7680 chips+ T(TX_diff) Note: The multiple of 256 chips offset between HS-DPCCH and uplink DPCCH ensures orthogonality between the two physical channels.
2-40
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-41
Version 1 Rev 0
DeltaNack
Integer value ranging from (0..8) Determines power increase of HS-DPCCH slot carrying Nack. The power increase is relative to the associated uplink DPCCH.
ACK-NACK-repetitionFactor
Integer value ranging from (1..4) The UE shall transmit the Ack/Nack information received from UEs MAC-hs in the slot allocated to the HARQ-ACK in the corresponding HS-DPCCH sub-frame as shown in slide HS-SCCH / HS-DSCH and HS-DPCCH Timing. When ACK-NACK-repetitionFactor is greater than one, the UE shall: repeat the transmission of the Ack/Nack information over the next (ACK-NACK-repetitionFactor-1) consecutive HS-DPCCH sub-frames, in the slots allocated to the HARQ-ACK and not attempt to receive nor decode transport blocks from the HS-PDSCH in HS-DSCH sub-frames corresponding to HS-DPCCH sub-frames in which the Ack/Nack information transmission is repeated. If consistent control information is not detected (i.e. UE is not able to decode its H-RNTI) in the downlink HS-SCCH subframe on any of the HS-SCCHs in the HS-SCCH set, neither Ack, nor Nack shall be transmitted in the corresponding HS-DPCCH subframe.
2-42
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-43
Version 1 Rev 0
where 4(c) value is signaled by higher-layer, e.g. through RRC or calculated based on TFCI selection. With the exception of the start and end of compressed frames, any DPCCH power change shall not modify the power ratio between the DPCCH and the HS-DPCCH.
2-44
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-45
Version 1 Rev 0
False Alarm
In very rare cases another UE decodes successful part 1 and part 2 and transmits either Ack/Nack in the uplink although no resources were allocated to this UE at all via HS-SCCH. So the NodeB should ignore that Ack/Nack as it knows that no resource allocation was sent to that UE via HS-SCCH. Note: Every UE has its own uplink scrambling code thus ACK/NACK and/or CQI repetitions can overlap among different UEs in uplink. Both procedures are independent from each and therefore may work asynchronous to each other (not every CQI repetition has a preceding ACK/NACK repetition).
2-46
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-47
Version 1 Rev 0
2-48
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-49
Version 1 Rev 0
CQI-RepetitionFactor
Integer value ranging from (14) Determines how often the CQI is repeatedly sent from UE towards NodeB to indicate the downlink radio channel quality.
DeltaCQI
Integer value ranging from (08). Determines power increase of HS-DPCCH slot carrying CQI. The power increase is relative to the associated uplink DPCCH.
2-50
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-51
Version 1 Rev 0
2-52
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-53
Version 1 Rev 0
2-54
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-55
Version 1 Rev 0
2-56
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-57
Version 1 Rev 0
CQI value
The CQI value 0 is reserved for the state do not bother to transmit while the other states represent the transmission capacity that the UE can receive at the current time. The CQI value consists of 5 bits that carry quality related information.
2-58
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-59
Version 1 Rev 0
2-60
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-61
Version 1 Rev 0
CQI Repetition
As a partial solution to the problem outlined above, it is recommended to have UTRAN-controlled repetition factors for the Ack/Nack and CQI fields in order to decrease the power offsets. The repetition factors for these two fields are set independently of each other and range from 1 through 5. The repetition takes place across consecutive sub-frames taking the value of m and the value of k into account. For more details please see page HS-DPCCH Measurement Feed Back CQI. Note: When the ACK/NACK repetition factor is greater than one, then the UE will not be scheduled for any new or retransmitted data in the corresponding successive HS-DSCH TTIs.
2-62
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-63
Version 1 Rev 0
2-64
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-65
Version 1 Rev 0
2-66
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-67
Version 1 Rev 0
2-68
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-69
Version 1 Rev 0
2-70
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-71
Version 1 Rev 0
Bit Re-arrangement b = 3
Output bit sequence = i2q2i1q1 Swapping i1 with i2 and q1 with q2 and XOR with 0011
2-72
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-73
Version 1 Rev 0
UE Transmit Power
The HS-DPCCH needs some part of the uplink transmission power, which has an impact on the link budget for uplink. With active HS-DPCCH the uplink power increases in the network resulting in a higher uplink interference value. At the same time the UEs battery capacity decreases. Note: On the HS-DPCCH there is some possibility to set an optimum between deltaAck/Nack and the ack-Nack-repetition-factor. The same is possible between deltaCQI and CQI-repetition-factor. A higher repetition factor requires less power increase for CQI and Ack/Nack reporting. However there is a trade off between consecutive HS-DSCH allocation (also depends on the UE category) and minimizing uplink interference. As long as the Ack/Nack is repeated, no new HS-DSCH can be allocated. A high repetition factor for CQI reporting may delay further HS-DSCH allocation as well as AMC in NodeB is not able to decode the reported CQI value as the UE transmitted with too low uplink transmission power. So there is some room for optimizing the system performance.
2-74
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-75
Version 1 Rev 0
2-76
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-77
Version 1 Rev 0
2-78
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-79
Version 1 Rev 0
2-80
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-81
Version 1 Rev 0
AMC Liabilities
AMC Liabilities
The implementation of AMC in UMTS also offers several challenges.
2-82
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
AMC Liabilities
Version 1 Rev 0
AMC Liabilities
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-83
Version 1 Rev 0
Chase Combining
Retransmission(s) of the same packet as that of the first attempt occur. The decoder combines multiple received copies of the coded packet weighted by their SNR prior to decoding. This method provides time diversity gain and is very simple to implement. Time diversity gain is simply the fact that the fast fading and interference changes between first transmission and multiple retransmission thus not all (re-)transmissions are affected by the same (bad) radio channel conditions.
Partial IR
The retransmission takes place with a partially different packet from the first one. Each packet transmitted in the partial IR scheme is self-decodable because it has the systematic bits of turbo codes. Instead of sending simple repeats of the entire coded packet, additional redundant information is incrementally transmitted. Note: The systematic bits of turbo encoded bits shall not be punctured, the other bits (parity bits p1 and parity bits p2) may be punctured.
Full IR
The retransmission of a fully different packet from the first one occur. The retransmission of packets are not self-decodable, thus they may contain only the parity bits of the turbo code output. This means also that the first transmission of a packet must at least contain the systematic bits of a packet. IR usually yields better performance compared to chase combining. However, it requires more implementation complexity and may not result in good performance unless the link adaptation errors are very small. Chase combining yields reasonable performance with lower implementation complexity and cost. Note: HARQ is controlled by MAC-hs scheduler.
2-84
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-85
Version 1 Rev 0
HARQ Operation
HARQ Operation
Tthe following sections describe the principle of HARQ for RLC-AM. All three figures (chase combining, partial IR and full IR) always show at the end the status of the demodulator. A tick means that the demodulator could reliably predict whether the received bits are determined as logical 0 or 1. This information is used as input for the turbo decoder which decodes the HS-DSCH TB. Only the CRC check after turbo decoding ensures if the HS-DSCH TB was successfully received.In these sections, the aim is to demonstrate the differences in transmission and retransmission by using chase combining, partial and full IR The receiver comprises a demodulator and a channel decoder. The demodulator generates soft decision bits for each received bit. The soft decision bits represent the likelihood of the real bit value. Thus the demodulator tries to find out all possible combinations of bit values for unknown bits targeting the maximum probability of the bit sequence. The demodulator stores, therefore after each packet has been demodulated, the highest probability values for each bit in soft decision bits. When the channel decoder performs the decoding and the CRC check indicates a block error, retransmission is requested from the sender. The HS-DSCH is turbo encoded with a rate of 1/3. The turbo coder delivers 3 output bit streams. Systematic bits Parity P1 bits Parity P2 bits The systematic bits are always needed to decode the original transport block. Transmission or retransmission containing the systematic bits are therefore also called self-decodable transmissions. The parity bits add redundancy to the encoded block and are not sufficient to decode the transport block without systematic bits.
2-86
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
HARQ Operation
Version 1 Rev 0
HARQ Operation
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-87
Version 1 Rev 0
HARQ Operation
HARQ Operation
Operation of Partial IR
With partial IR, every transmission and retransmission is still self-decodable however the puncturing scheme varies between retransmission. Thus the retransmission becomes more effective compared to chase combing where the puncturing scheme stays the same. As multiple retransmissions vary in their punctured bits, the demodulator can benefit from several instances of the same packet reducing the code rate. After each retransmission the demodulator can re-calculate the maximum likelihood by either taking the soft decision values of the transmission and retransmission into account or their added sum weighted by their individual SNRs. The buffer requirements are still modest compared to full IR Note: IR is another Hybrid ARQ technique wherein instead of sending simple repeats of the entire coded packet; additional redundant information is incrementally transmitted if the decoding fails on the first attempt. Partial IR is called Hybrid ARQ type-III as each retransmission is restricted to be self-decodable. In the IR schemes the receiver buffers coded symbols, which introduce new information to the HSDPA TTI transmitted first. Therefore the amount of data to be buffered increases with consecutive retransmission. Prior to decoding these symbols are soft-valued, i.e. each symbol is represented by two or more bits.
2-88
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
HARQ Operation
Version 1 Rev 0
HARQ Operation
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-89
Version 1 Rev 0
HARQ Operation
HARQ Operation
Operation of Full IR
Using full IR, the first transmission of a packet must be self-decodable thus the systematic bits need to be inside. However with every retransmission instance of the same packet only the parity bits may be transmitted. Every retransmission therefore reduces the code rate and increases the decoding probability. However the likelihood calculation makes it necessary to buffer each transmission instance. The soft decision values cannot be added up as the retransmissions containing the parity bits only are not self decodable. The full IR requires the biggest memory size and thus may not be useable for maximum data rate transmissions. Note: Full IR is called Hybrid ARQ type II as each retransmission is not self-decodable. In case of HARQ type II or type III with multiple redundancy versions, additional redundancy bits are sent during each retransmission yielding potentially more coding gain than simple type III with single redundancy version.
2-90
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
HARQ Operation
Version 1 Rev 0
HARQ Operation
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-91
Version 1 Rev 0
2-92
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-93
Version 1 Rev 0
Serve Order
The order in which users are served; e.g. random order (round robin) or according to channel quality (C/I or throughput based). More advanced mechanisms require higher computational processing in NodeB.
Allocation Method
The criterion for allocating resources, e.g. same data amount or same power / code / time resources for all queued users per allocation interval (e.g. TTI). FT (Fair Throughput) Allocation method: Resources are allocated according to equal data amount for all users up to a maximum allocation. The users are served in random order. In theory, all users currently active will therefore experience the same delay and throughput. P-FR (Proportional Fair Resources) Allocation method: The same resources (power / code / time) are allocated for all users. Due to the fast scheduling frequency, the P-FR PS is able to track the fast fading. P-FR utilizes the CQI as well as link quality estimation algorithms, which are located in the NodeB. The proportional scheduling method results in all users getting approximately an equal probability of becoming served even though they may experience very different channel quality M-C/I (Maximum C/I) Allocation method: The same resources (power / code / time) are allocated per TTI. Like P-FR, the M-C/I scheduling rate is fast and therefore it is able to track the fast fading). The M-C/I scheduling method is the most drastic one, which serves the best user during the current TTI. For example, the user who can sustain the highest throughput. An available option with the HSDPA concept is to make very fast scheduling, which tracks the fast fading variations. Ultimately, users are only scheduled when they are experiencing constructive fading. This improves both, the user throughput and cell throughput for time-shared channels. However if QoS requirements dominate the scheduling strategy, the differentiation between the above-mentioned PS strategies become less significant as the PS strives to fulfill the users requirements instead of treating either everyone fair or special if when considering the channel quality. Event the capacity of the most aggressive schedulers, e.g. M-C/I reduces when QoS must be ensured. Actually they become more fair.
2-94
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-95
Version 1 Rev 0
Fairness
The key to HSDPA operation and performance lies in the PS entity. Fast PS methods track the user channel quality for each TTI and makes scheduling decisions on the fly. The important issue is that of inherent fairness among users. The FT (fair throughput) approach ensures that all simultaneously queued users receive the same average throughput, which means that users in bad conditions are given relatively more HS-DSCH resources. On the other end, the M-CI (maximum C/I) method only gives resources to the users in the most favorable conditions thereby enhancing cell throughput at the expense of inter-user fairness. In between, the FR (fair resource method gives equal resources to all users so that all users will achieve a throughput corresponding to the channel quality.
Opportunistic Scheduling
The proportional methods P-FT (proportional fair throughput) and P-FR (proportional fair resource) attempt to give all users the same probability of being scheduled by using a relative instantaneous channel quality; e.g. defined as the ratio between current and average user Es/No. Ideally, this method will schedule users only during constructive fades thereby raising both the overall cell throughput and the user data rates. By weighting the relative instantaneous channel quality with buffer statistics, the fairness of the proportional methods can be increased (e.g. P-FT type scheduling).
Effect of TCP/IP
Avoiding delay jitter is essential to achieve high throughput for TCP. In particular, delay spikes can cause spurious timeouts. Such timeouts force TCP into slow-starts, which reduce TCPs congestion window sizes drastically. For HS-DSCH, jitter can occur due to varying interference as well properties of the PS influence the jitter.
Influence of Coverage
As mentioned already in chapter 1, there is no soft handover possible on the HS-DSCH. Therefore users at the cell edge do not benefit from soft handover gain when transmitting on the HS-DSCH. In order to transmit to these users at satisfactorily throughput levels, HS-DSCH must be transmitted with relatively higher power compared to users closer to the center of the cell. As the overhead associated with control signaling on HS-SCCH and HS-DPCCH is also highest at cell edge, it may be desirable to limit the coverage of the HS-DSCH. Note that the cost in carried load can be very high when supporting the full cell area as the effective throughput declines with increasing coverage. So less revenue per effective throughput can be achieved at large cell sizes.
2-96
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-97
Version 1 Rev 0
2-98
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-99
Version 1 Rev 0
MAC-hs Scheduler should not allocate any HSDPA Resources to a User which:
Has no data for transmission Has no free capacity in any reordering buffer Has no ready-to-transmit HARQ process Has not completed the minimum inter-TTI interval after the last HSDPA transmission (see chapter 3 page HSDPA Category and Reference Combinations) Has reported a too low/bad CQI or CQI is out of range Has not reported any CQI or is physically out of sync in UL
2-100
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
2-101
Version 1 Rev 0
2-102
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-1
Version 1 Rev 0
3-2
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Objectives
Version 1 Rev 0
Objectives
Describe the Turbo Coding and Decoding Principle State the reason for Hybrid ARQ in HSDPA Describe the HARQ Functionality State the different purpose of Rate Matching Describe different performance of Partial IR and Full IR
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-3
Version 1 Rev 0
Turbo Coding
Turbo Coding
Turbo Coder Principle
In the beginning of this chapter we would like to introduce only the basics of turbo coding and decoding. This is necessary to ease the understanding of the Hybrid ARQ functionality which follows later on. First of all the fundamental turbo code encoder is built using two identical RSCs (recursive systematic convolutional coder) with parallel concatenation. An RSC encoder is typically r = 1/2 and is termed a component encoder. The two component encoders are separated by an interleaver. Only one of the systematic outputs from the two component encoders is used, because the systematic output from the other component encoder is just a permuted version of the chosen systematic output. The figure shows the fundamental turbo code encoder with an r = 1/3. One input bit corresponds to three output bits N (Sys, P_1 and P_2). The first RSC encoder outputs the systematic bits denoted as Sys and recursive convolutional sequence of parity bits denoted as P_1. The second RSC encoder discards its systematic bit sequence and only outputs its recursive convolutional sequence of parity bits denoted as P_2. Both encoder components are parallel concatenated and the same input data bits are fed in, therefore these codes are also called PCCC (parallel concatenated convolutional code).
RSC Encoder
The RSC encoder is obtained from the non-recursive non-systematic (conventional) convolutional encoder by feeding back one of its encoded outputs to its input.
Interleaver
For turbo codes, an interleaver is used between the two component encoders. The interleaver is used to provide randomness to the input sequences. The interleaver has the task to de-correlate the input sequence between the two RSCs. It generates out of Sys input bits a pseudo random re-ordered sequence of Sys bits. The de-correlation function of the interleaver is very important as only when the two input sequences Sys and Sys are most de-correlated, a powerful decoder can be realized. Note: In order to generate higher code rates, e.g. 1/2, puncturing can be used. Puncturing deletes bits according to some rule or algorithm and therefore reduces the number of output bits. The percentage of code-word to redundancy information gets reduced. However in turbo coding only the parity bits are punctured, never the systematic bits. Example: In order to generate from a 1/3 turbo coder a 1/2 bit sequence, the following scheme could be used: all systematic bits are transmitted, but the even numbered parity bits of P_1 and the odd numbered bits of P_2 are removed from the output stream N 1/(1+0.5 + 0.5) = 1/2.
3-4
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Turbo Coding
Version 1 Rev 0
Turbo Coding
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-5
Version 1 Rev 0
3-6
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-7
Version 1 Rev 0
3-8
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-9
Version 1 Rev 0
QPSK
For all transmissions of a transport block, the transport block size is derived from the TFRI value as specified in the figure. However there is an exception. In the case of retransmission, where the Node-B selects a combination for which no mapping exists between the original transport block size, the selected combination of channelization code set and the modulation type, then the transport block size index value signaled to the UE shall be set to 111111, i.e. k(i) = 63. The parameter k(i) corresponds to TFRI signaled on the HS-SCCH to the UE. The parameter k(0,i) is obtained from the table corresponding to the modulation type and the number of codes which are also signaled on the HS-SCCH to the UE. Let k(t) be the sum of the two values: k(t) = k(i) + (k0,i) The transport block size L[k(t)] can be calculated in the following way: IF k(t) < 40: L[k(t)] = 125 + 12 x k(t) ELSE L[k(t) = 296 x [(2085 / 2048) ^ k(t)]
3-10
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-11
Version 1 Rev 0
3-12
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-13
Version 1 Rev 0
DCH
All Rel. 99 transport channels terminate at the RNC. Hence, the retransmission procedure for packet data is located in the SRNC, which also handles the connection for the particular core network. In RLC-AM the SRNC starts for every sent RLC-PDU a RLC retransmission polling timer. If the RLC status PDU positively or negatively acknowledging the RLC PDU in time a retransmission is sent. How many RLC retransmissions are performed for the same PDU depend on the bearer configuration, e.g. MaxDat parameter could be used to limit the RLC retransmission.
HS-DSCH
With the introduction of HSDPA additional intelligence in form of MAC-hs is installed in the NodeB. This way, retransmission can be controlled directly by the NodeB, leading to faster retransmission and thus shorter delay with packet data operation when retransmission is needed. The lower figure presents the difference. The RLC-PDU is sent via MAC-hs from NodeB towards UE. MAC-hs in NodeB monitors the successful transmission based on the fixed uplink timing of HS-DPCCH. If either the HS-DPCCH decoding fails or a Nack is received, then MAC-hs in NodeB schedules retransmission autonomously. In RLC-AM mode however the RLC of SRNC starts like in legacy UMTS releases a retransmission polling timer. If the RLC status PDU is not received in time to ack the RLC-PDU, the timeout of the RLC polling timer triggers the retransmission of the RLC-PDU. For MAC-hs it looks like an entire new transmission. With HSDPA the Iub interface between NodeB and SRNC (MAC-d) requires a flow control mechanism to ensure that NodeB buffers are used properly and that there is no data loss of RLC-PDUs due to NodeB buffer overflow. The MAC-hs in NodeB is responsible for scheduling retransmission. The RLC-PDU sent by SRNC to NodeB is embedded by MAC-hs into a so called MAC-hs-PDU. However, the SRNC still retains the Rel. 99 /Rel. 4 functionalities of RLC, such as taking care of retransmission (only in case of RLC-AM) in case of the HS-DSCH transmission from NodeB to UE fails after exceeding the maximum number of physical layer retransmission or a MAC-hs PDU release timer T1 expires. The key functionality of the MAC-hs is to handle the HARQ functionality and scheduling as well as priority handling. Ciphering is done in any case in the RLC layer to ensure that the ciphering mask stays identical for each retransmission. This enables physical layer combining of retransmission. Note: The type of scheduling to be carried out in NodeB is not defined in 3GPP standardization. Only some parameters, such as discard timer or scheduling priority indication can be used by RNC to control the handling of an individual user. In chapter 4 we explain these parameters and their functions in full detail.
3-14
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-15
Version 1 Rev 0
Chase combining: Every transmission and retransmission of HS-DSCH data blocks must be always self-decodable. Therefore s = 1 and r = 0 for all HS-DSCH data blocks. As there is only one value allowed for r, chase combining contains always the same redundant parity bits. Partial IR: Also when partial IR is performed, every transmission and retransmission must be always self-decodable. Therefore s = 1 but different redundant bits of the parity information are added. The different parity bits are selected by the parameter r. Full IR: Compared to partial IR, full IR is only used for retransmission. The initial transmission must be self-decodable in order to transmit the systematic bits. Bit Rearrangement only for 16-QAM: The positions I2 and q2 are less reliable therefore in retransmission the bits are rearranged in order to have equal error probability among transmission and retransmission.
3-16
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-17
Version 1 Rev 0
HS-DSCH Category
New capability parameter introduced in Rel. 5. Category 11 and 12 only support QPSK, all other categories support both QPSK and 16-QAM.
Maximum Number of HS-DSCH Transport Channel Bits received within HS-DSCH TTI
This parameter defines the maximum transport channel bits which can be received by the UE within 2 ms interval.
3-18
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-19
Version 1 Rev 0
Chase Combining
QPSK: no problem the total transmission capacity is only 4800 bits with 5 HS-PDSCHs so the UEs soft-buffer cannot be overloaded 16-QAM: no problem the total transmission capacity with 9600 bits is still below the maximum HARQ process memory 16-QAM code rate R: 7298 / 9600 bits = 0.76 ~
IR Retransmission
QPSK: no problem the total transmission capacity is only 4800 bits with 5 HS-PDSCHs so the UEs soft-buffer cannot be overloaded 16-QAM: issue the total number of bits of 21978 exceeds UEs soft-buffer capacity puncturing of parity P1 and P2 bits down to 9474 bits Alternative for 16-QAM: (16800 bits 12) / 3 24 = 5572 bits With TB size 8 5572 bits UEs total soft-buffer capacity gets never exceeded when Partial IR or Full IR are used for retransmission. If the amount of soft-buffer is not an issue, then IR is superior compared to chase combining. However, e.g. for UE category 10, the available soft-buffer capacity with 6 HARQ processes is only 28800. If the biggest TB size of 27952 is chosen, then only chase combining makes sense as no further parity bits can be retransmitted when UEs soft-buffer is only marginal bigger than UEs max. physical transmit capacity.
3-20
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-21
Version 1 Rev 0
HARQ Information
HARQ Information
The UE configures the MAC-hs entity with the number of HARQ processes indicated in IE "Number of Processes". The UE assigns to each of these HARQ processes IDs going from 0 to "Number of Processes" 1 HARQ Memory Partitioning Information: The first instance of the parameter corresponds to HARQ process with identifier 0, the second instance to HARQ process with identifier 1, and so on.
3-22
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
HARQ Information
Version 1 Rev 0
HARQ Information
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-23
Version 1 Rev 0
UE Side
HARQ entity is responsible for generating Ack/Nacks of the data in the corresponding HARQ process and schedule the positive or negative acknowledgement for transmission relative to the reception of data ( HARQ control N). The detailed configuration of HARQ protocol is provided by RRC. In order to combine multiple retransmissions of the same MAC-hs PDU belonging to a certain HARQ process, the UE need to have a buffer per HARQ process. Note: After channel decoding, the CRC check determines if the HS-DSCH transport block is correct or erroneous. Reordering Queue distribution function routes the MAC-hs PDUs to the correct reordering buffer based on the queue ID. Reordering entity & Disassemble: The reordering is done according to the received TSN in the MAC-hs PDU. MAC-hs PDUs with consecutive TSNs are delivered to the disassembly function upon reception. MAC-hs PDUs are not delivered to the disassembly function if MAC-hs PDUs with lower TSN are missing. There is one reordering entity for each queue ID configured at the UE.
3-24
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-25
Version 1 Rev 0
Bit Separation
This unit separates the bits stream in systematic bits (x(1), (x(2),..., x(k), x(k+1), z(k+2) x(k+1), z(k+2)), parity 1 bits (z(1), z(2), ..., z(k), z(k+1), x(k+3) z(k+1), x(k+3)) and parity 2 bits (z(1), z(2), ..., z(k), x(k+2), z(k+3) x(k+2), z(k+3)).
1st RM
The first rate-matching stage is identical to the Rel.99 rate-matching functionality except that the number of output bits does not match to the number of physical-channel bits available in the HS-DSCH TTI. Instead the first rate matching block is used to adjust the number of available encoded bits N(TTI) to the virtual UEs buffer size. Note: The first rate matching block is always transparent for the systematic bits therefore only parity-1 and parity-2 bits are punctured. If N(TTI) UE soft-buffer capability (N(IR) > 3N), the first rate matching stage is also transparent for parity-1 and parity-2 bit streams.
2nd RM
The second rate matching stage matches the number of bits after first rate matching to the number of physical channel bits available in the HS-DSCH TTI. The second rate matching also uses the Rel. 99 rate matching algorithm. However, the rate matching only considers bits that have not been punctured by the first rate matching stage and the rate matching parameters used in a particular transmission are controlled by the RV parameter.
RV
The HARQ functionality is controlled by the parameter RV, i.e. the exact set of bits at the output of the physical-layer Hybrid-ARQ functionality depends on the number of input bits, the number of available physical channel bits (number of HS-PDSCHs and modulation type), and the RV parameter. Note: The number of RVs allowed for HS-DSCH is limited to 8. Parameter RV decides about the different redundancy versions for IR.
Bit Collection
The HARQ bit collection is achieved using a rectangular interleaver. The size of the interleaver is determined by the modulation type and the number of encoded and rate matched bits. The purpose of the bit collector is to transmit the systematic bits on the more reliable positions of a 16-QAM symbol. A perfect mapping exists only for code rate 1/2. Otherwise it is just best effort.
3-26
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-27
Version 1 Rev 0
3-28
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-29
Version 1 Rev 0
3-30
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-31
Version 1 Rev 0
First RM Stage
First RM Stage
HARQ first stage rate matching for the HS-DSCH transport channel is done with the general method as described in the picture with the following specific parameters: The maximum number of soft channel bits available in the virtual IR buffer is N(IR) which is signaled via RRC for each HARQ process. The number of coded bits in a TTI before rate matching is N(TTI) this is deduced from information signaled via RRC as well and parameters signaled on the HS-SCCH for each TTI HARQ processing and physical layer storage occurs independently for each HARQ process currently active. If N(IR) is greater than or equal to N(TTI) (i.e. all coded bits of the corresponding TTI can be stored) the first rate matching stage is transparent. Note that no repetition is performed in 1st RM stage. The 2nd RM stage matches the encoded bits of the TB to the available channel capacity N(data). This function is explained later on. If N(IR) is smaller than N(TTI) the parity bit streams are punctured by setting the rate matching parameter 3N(TTI) = N(IR) N(TTI). NOTE The negative value is expected when the rate matching implements puncturing. Bits selected for puncturing which appear as x(m) in the algorithm depicted above shall be discarded and not counted in the totals for the streams through the virtual IR buffer.
Puncturing Algorithm
The variables e(ini), X(i), e(plus) and e(minus) must be initialized accordingly. X(i) represents the number of parity bits before 1st RM. As X(i) = N(TTI)/3 and N(TTI) is always a multiple of 3 the result is always an integer. e(ini) is the start value in the beginning of the algorithm which sets the variable e =X(i) at the beginning only. e(minus) decrements e in order to get e 8 0. Only when e 8 0 the corresponding bit can be deleted. x(m) represents the bit position of the bit being deleted once e 8 0. The parameter m starts with first bit position and is incremented after every loop cycle. e(plus) increments e in order to push e above zero again after a bit has been determined by the loop for puncturing 3N(TTI) determines the number of bits to be punctured in total. As parity-1 and parity-2 have the same amount of bits, 3N(TTI)/2 bits must be deleted for each parity stream. At the end of the puncturing process, the algorithm delivers the amount of N(TTI)/3 - 3N(TTI)/2 bits for parity-1 and parity-2
3-32
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
First RM Stage
Version 1 Rev 0
First RM Stage
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-33
Version 1 Rev 0
3-34
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-35
Version 1 Rev 0
RM Pattern Determination
RM Pattern Determination
The figure shows the algorithm determining which bit positions are to be punctured or repeated. The variables X(i), e(ini), e(plus) and e(minus) are calculated earlier. X(i) is the number of bits in each sequence {systematic bits}, {parity-1 bits} and {parity-2 bits} m is the loop variable which runs from 1 to X(i) and determines the bit number which is to be deleted or punctured.
Bit Repetition
A similar repetition rate in all bit streams is achieved by setting the number of transmitted systematic bits to: N(t,sys) = N(sys) x N(data)/[N(sys) + 2 x N(p1)] The number of parity bits in a transmission is: N(t,p1) = [N(data) N(t,sys)] / 2 N(t,p2) = [N(data) N(t,sys)] / 2 Example: N(sys) = 30; N(p1) = N(p2) = 30; N(data) = 100 N(t,sys) = 30 x 100/[30 + 2 x 30] = 30 x 100 / 90 = 30 x 10 / 9 = 34 N(t,p1) = [100 34] / 2 = 66./2 = 33 N(t,p2) = [100 34] / 2 = 66./2 = 33 So 4 Bits of the systematic bits are repeated as well 3 bits each of both parity bits are repeated.
3-36
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
RM Pattern Determination
Version 1 Rev 0
RM Pattern Determination
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-37
Version 1 Rev 0
3-38
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-39
Version 1 Rev 0
Full IR
The parameter s = 0 indicates non-self-decodable transmission and parameter r = 1 indicates which parity bits will be transmitted according to the 2nd RM algorithm. In the original transmission 22% of the parity bits were already transmitted. Calculation: (696 + 697) / (3097 + 3096) = 22% In the retransmission the remaining 78% parity bits can be transmitted. Calculation: 4800 / (3097 + 3096) = 78% The significant sign of full IR is that the systematic bits are not retransmitted as shown in the picture. Below we would like to explain shortly how the retransmission would like if partial IR is used.
Partial IR
In this case also the retransmission contains all of the systematic bits thus always only 22% of the parity bits can be transmitted. Calculation: (696 + 697) / (3097 + 3096) = 22% Thus it takes much longer until all parity bits are transmitted. Theoretically it would take 5 transmissions in order to transmit all of the parity bits. Each retransmission would contain 3407 systematic bits. The advantage of full IR becomes obvious now. Higher data rates get more effectively retransmitted however the disadvantage is that the systematic bits are only contained in the original transmission. So the UE must be able to successfully decode the TB out of the corrupted original transmission and the parity bits.
3-40
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-41
Version 1 Rev 0
3-42
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-43
Version 1 Rev 0
3-44
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-45
Version 1 Rev 0
Systematic RM
As e(minus) is always 0, no bits are punctured e never becomes less or equal to 0. The variable e remains constant and positive; e = e(ini) for every bit position.
Parity_1 RM
Here the variable e gets decremented for every m by e(minus). We always indicate the result of the subtraction per m. Once e becomes less or equal 0, the very bit position indicated by m needs to be punctured. Thus we mark that row with red color. According to the RM loop algorithm, whenever e becomes less or equal to 0, e has to be incremented by e(plus) in order to continue.
Parity_2 RM
The bit position calculation for parity-2 bits is basically the same as for parity-1. Only the loop variables e(ini), e(plus) and e(minus) differ in order to get different results compared to parity_1 RM. Thus different bit positions are punctured. Note: For retransmission with s = 1 and always the same redundancy version, e.g. r = 0, this scheme represents HARQ III (chase combining).
3-46
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-47
Version 1 Rev 0
3-48
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-49
Version 1 Rev 0
3-50
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-51
Version 1 Rev 0
3-52
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-53
Version 1 Rev 0
3-54
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-55
Version 1 Rev 0
3-56
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-57
Version 1 Rev 0
R = 3/4
For high initial code rates it is obvious that it takes partial IR several retransmissions until all parity bits have been sent once. R(eff) is slowly decreasing and only after six transmissions all bits have been sent (3/4, 3/5, 3/6, 3/7, 3/8, 3/9) Full IR allows decreasing the effective code rate more rapidly. As depicted in the figure, all bits have been sent once already after three transmissions.
R = 1/2
For R = 1/2, however, there is no longer a decoding gain advantage for full IR, since partial IR allows also to send all available bits within two transmissions. Detailed investigations have shown, that for HSDPA full IR is favorable for initial code rates R > 1/2, while partial IR should be used for lower initial code rates, e.g. R8 1/2.
3-58
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-59
Version 1 Rev 0
3-60
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-61
Version 1 Rev 0
HS-PDSCH Interleaving
HS-PDSCH Interleaving
The interleaving for HSDPA is performed as shown in the figure, separately for each physical channel. For QPSK the interleaver is the same as Rel-99 2nd interleaver. The interleaver is of fixed size: R2=32 rows and C2=30 columns. For 16QAM, there are two identical interleavers of the same fixed size R2C2 = 3230. The output bits from the physical channel segmentation are divided between the interleavers: all odd numbered bits to interleaver one ( upper Interleaver) and all even numbered bits to interleaver two (lower interleaver). Note: The outputs of the interleavers will result in mapping to 16QAM symbols such that the output of first interleaver is mapped to the more reliable positions (i1 and q1) whereas the output of the second interleaver is mapped to the less reliable positions (i2 and q2).
3-62
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
HS-PDSCH Interleaving
Version 1 Rev 0
HS-PDSCH Interleaving
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
3-63
Version 1 Rev 0
HS-PDSCH Interleaving
3-64
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-1
Version 1 Rev 0
4-2
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Objectives
Version 1 Rev 0
Objectives
After this Lecture the Student will be able to: State the interworking between MAC-hs in NodeB and MAC-d in RNC Describe the various functions and components of MAC-hs State the new HS-DSCH FP Control and Data Frame Describe the protocol extensions of NBAP State the basic HSDPA Data Transfer
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-3
Version 1 Rev 0
HS-DSCH Characteristics
The HS-DSCH transport channel has the following characteristics: An HS-DSCH transport channel is processed and decoded from one CCTrCH; There is only one CCTrCH of HS-DSCH type per UE; The CCTrCH can be mapped to one or several physical channels; There is only one HS-DSCH per CCTrCH; Existence in downlink only; Possibility to use beam forming; Possibility of applying link adaptation techniques other than power control; Possibility to be broadcast in the entire cell; Always associated with a DPCH and one or more shared physical control channels (HS-SCCHs).
4-4
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-5
Version 1 Rev 0
4-6
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-7
Version 1 Rev 0
4-8
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-9
Version 1 Rev 0
4-10
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-11
Version 1 Rev 0
4-12
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-13
Version 1 Rev 0
UE Capabilities Information
HS-SCCH Power Offset The HS-SCCH Power Offset IE indicates the Power offset relative to the pilot bits on the DL DPCCH.
4-14
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-15
Version 1 Rev 0
4-16
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-17
Version 1 Rev 0
4-18
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-19
Version 1 Rev 0
4-20
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-21
Version 1 Rev 0
4-22
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-23
Version 1 Rev 0
4-24
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-25
Version 1 Rev 0
MP Mandatory Present
A value for that information is always needed, and no information is provided about a particular default value. If ever the transfer syntax allows absence (e.g., due to extension), then absence leads to an error diagnosis.
MD Mandatory Default
A value for that information is always needed, and a particular default value is mentioned (in the semantical information column). This opens the possibility for the transfer syntax to use absence or a special pattern to encode the default value. Example: Activation Time If there is no value for Activation Time then the mandatory default value is now.
CV Conditional Value
The need for a value for that information depends on the value of some other IE or IEs, and/or on the message flow (e.g., channel, SAP). The need is specified by means of a condition, the result of which may be that the information is mandatory present, mandatory with default value, not needed or optional. If one of the results of the condition is that the information is mandatory present, the transfer syntax must allow for the presence of the information. If in this case the information is absent an error is diagnosed. If one of the results of the condition is that the information is mandatory with default value, and a particular default value is mentioned (in the Semantical information column), the transfer syntax may use absence or a special pattern to encode the default value. If one of the results of the condition is that the information is not needed, the transfer syntax must allow encoding the absence. If in this case the information is present, it will be ignored. In specific cases however, an error may be diagnosed instead. If one of the results of the condition is that the information is optional, the transfer syntax must allow for the presence of the information. In this case, neither absence nor presence of the information leads to an error diagnosis.
OP - Optional
The presence or absence is significant and modifies the behavior of the receiver. However whether the information is present or not does not lead to an error diagnosis.
4-26
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-27
Version 1 Rev 0
MAC-hs Queue Id
The MAC-hs queue ID is unique across all MAC-d flows.
4-28
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-29
Version 1 Rev 0
MAC-hs Reset
After a MAC-hs reset; among other things (which are explained on page UE MAC-hs Other Functions); indicate to all AM RLC entities mapped on HS-DSCH to generate a status report. Note: The UE shall generate a status report, even if the STATUS prohibit or the EPC (estimated PDU counter) mechanism is configured to prohibit the Receiver from sending a status report containing any of the SUFIs LIST, BITMAP, RLIST or ACK if an indicator is received from MAC requesting generation of a status report following MAC-hs reset.
4-30
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-31
Version 1 Rev 0
MAC-hs Protocol
MAC-hs Protocol
The MAC protocol in the RNC will decide on which transport channel the MAC-d PDU will be transmitted. For the HS-DSCH transport, the MAC is further split into the NodeB MAC called the MAC-hs (MAC high speed), which enables fast radio resource allocation. The MAC-hs takes care of the transport block scheduling, HARQ channel allocation and transport format selection. The forward signaling on the HS-SCCH serves the MAC-hs at the receiver and contains, beside terminal identification, parameters for TFRC (number and indexes of the channelization codes and modulation), HARQ process identifier, TB size, redundancy version and new data indicator. The reverse signaling on the HS-DPCCH contains the CQI-report and the HARQ acknowledgements. The presented figure is only informative which aims to provide an overall picture. Note two MAC protocol configurations are possible on the UTRAN side. As already mentioned in chapter 1, configuration with or without MAC-c/sh is implementation dependent (vendor specific).
HS-DSCH RNTI
The HS-DSCH RNTI (=H-RNTI) is used for the UE specific CRC attachment in HS-SCCH. HS-DSCH RNTI is allocated by controlling RNC upon UE establishing a HS-DSCH channel. HS-DSCH RNTI shall be unique within the cell carrying the HS-DSCH.
4-32
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
MAC-hs Protocol
Version 1 Rev 0
MAC-hs Protocol
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-33
Version 1 Rev 0
MAC-hs Tasks
MAC-hs Tasks
The MAC-hs is responsible for handling the data transmitted on the HS-PDSCH. Furthermore it is responsible for the management of the physical resources allocated to HS-PDSCH. MAC-hs receives configuration parameters from the RRC layer via the MAC-Control SAP. There is priority handling per MAC-d PDU in the MAC-hs. MAC-hs has various tasks as outlined below:
TFRC Selection
TFRC (transport format and resource combination) is the selection of an appropriate transport format and resource combination for the data to be transmitted on HS-DSCH. The TFRI is the indicator for the TFRC selection. The TFRI is explicitly signaled on HS-SC.CH which indicates the TB size, channelization code set and modulation scheme. Compared to that the MAC-d PDU size is implicitly signaled via MAC-hs PDU header.
AMC
As explained in chapter 2 and 3 already AMC function provides for the initial MCS selection based on CQI reporting, DL PC, Ack/Nack ratio and UE capability.
HARQ
The HARQ protocol is based on an asynchronous downlink (N-SAW scheme) and synchronous uplink scheme (CQI reporting). The ARQ combining scheme is based on Incremental redundancy. Chase combining is considered to be a particular case of Incremental Redundancy. The UE soft memory capability is defined according to the needs for chase combining. The soft memory is partitioned across the HARQ processes in a semi-static fashion through upper layer signaling. The UTRAN should take into account the UE soft memory capability when configuring the different transport formats (including possibly multiple redundancy versions for the same effective code rate) and when selecting transport formats for transmission and retransmission. Ack / Nack and CQI Reporting is performed by UE using HS-DPCCH.
4-34
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
MAC-hs Tasks
Version 1 Rev 0
MAC-hs Tasks
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-35
Version 1 Rev 0
MAC-hs SDU
For HS-DSCH the MAC-d PDU format equals the MAC PDU format for non HS-DSCH case. The MAC-d PDU header for DTCH and DCCH mapped on HS-DSCH includes the C/T field if multiplexing on MAC is applied. In case of HS-DSCH a MAC PDU consists of one MAC-hs header and one or more MAC-hs SDUs where each MAC-hs SDU equals a MAC-d PDU. A maximum of one MAC-hs PDU can be transmitted in a TTI per UE. The MAC-hs header is of variable size. The MAC-hs SDUs in one TTI belongs to the same reordering queue. There need to be per PDP context / QoS profile one (re-)ordering queue configured in NodeB and UE in order to allow the individual user bearer(s) (RLC instances) being served by HSDPA according to their QoS needs.
4-36
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-37
Version 1 Rev 0
Scheduling/Priority Handling
This function manages HS-DSCH resources between HARQ entities and data flows according to their priority. Based on status reports from HS-DPCCH either new transmission or retransmission is determined. Further it determines the Queue ID and TSN for each new MAC-hs PDU being serviced. To maintain proper transmission priority, a new transmission can be initiated instead of a pending retransmission at any time to support the priority handling. The TSN is unique to each priority class within a HS-DSCH, and is incremented for each new data block. This means every queue has its own priority class value determined by the PDP context / QoS profile. The priority class per priority queue is assigned through NBAP. The PS may change the queues priority dynamically in order to manage various PDP contexts one user and also among different users. The queues priority determines when it is going to be serviced, means MAC-hs PDUs are transmitted on HS-DSCH. Note: It is not permitted to schedule new transmissions, including retransmission originating in the RLC layer, within the same TTI, along with retransmission originating from the HARQ layer.
HARQ
One HARQ entity handles the hybrid ARQ functionality for one user. One HARQ entity is capable of supporting multiple instances (HARQ process) of stop and wait HARQ protocols. There is one HARQ process per HS-DSCH per TTI.
TFRC Selection
Selection of an appropriate transport format and resource for the data to be transmitted on HS-DSCH. Note: the TFRC selection can be compared with the TFC selection on legacy UMTS transport channels.
4-38
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-39
Version 1 Rev 0
HARQ
The HARQ entity is responsible for handling the MAC functions relating to the HARQ protocol. The HARQ functional entity handles all the tasks that are required for hybrid ARQ. It is responsible for generating Acks or Nacks. The detailed configuration of the hybrid ARQ protocol is provided by RRC over the MAC-Control SAP. During HSDPA data reception, the UE obtains the HARQ (HARQ process ID, NDI bit and X(RV)) and TFRI information from the HS-SCCH. HARQ and TFRI info provide the necessary input parameter for the HARQ protocol.
Reordering
The reordering entity reorders received MAC-hs PDUs according to their TSN. MAC-hs PDUs with consecutive TSNs are delivered to the disassembly function upon reception. MAC-hs PDUs are not delivered to the disassembly function if MAC-hs PDUs with lower TSN are still missing. There is one reordering entity for each Queue ID configured at the UE.
Disassembly
The disassembly entity is responsible for the disassembly of MAC-hs PDUs. When a MAC-hs PDU is disassembled the MAC-hs header is removed, the MAC-d PDUs are extracted and any present padding bits are removed. Then the MAC-d PDUs are delivered to higher layer.
MAC-d
The MAC-d entity has a connection to the MAC-hs entity. This connection is used to receive data from the HS-DSCH transport channel which is handled by MAC-hs (downlink). There is one MAC-d entity in the UE.
4-40
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-41
Version 1 Rev 0
Overall Architecture
The MAC-hs is added to the MAC architecture of Rel. 99. The MAC-hs is located in the NodeB. If an HS-DSCH is assigned to the UE, the MAC-hs SDUs, i.e. the MAC-d PDUs to be transmitted are transferred from MAC-c/sh to the MAC-hs via the Iur/Iub interface in case of configuration with MAC-c/sh, or from the MAC-d via Iur/Iub in case of configuration without MAC-c/sh. The HS-DSCH FP is used to transfer the MAC-d PDUs from SRNC towards MAC-hs in NodeB by using the HS-DSCH Data Frame.
MAC-c/sh (optional)
The data for the HS-DSCH is subject to flow control between the SRNC and the DRNC/CRNC. Therefore a new flow control function is included in DRNC/CRNC to support the data transfer between MAC-d and MAC-hs.
4-42
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-43
Version 1 Rev 0
Overall Architecture
The figure shows the overall MAC architecture up to RLC level. The data received on the HS-DSCH is mapped to the MAC-hs. The MAC-hs is configured via RRC signaling to set the parameters in MAC-hs such as allowed transport format combinations for the HS-DSCH. The associated downlink signaling on HS-SCCH carries information for support of HS-DSCH while the associated uplink signaling on HS-DPCCH carries feedback information.
MAC-d Flow in UE
The figure shows the MAC-d flow mapping of one UE. The MAC-d entity is modified with the addition of a link to the MAC-hs entity. Note that the links to MAC-hs and MAC-c/sh cannot be configured simultaneously in one UE. This means that the UE is not able to support HS-DSCH and additional downlink shared or common TrCHs beside DCH. The mapping between C/T MUX entity in MAC-d and the reordering buffer in MAC-hs is configured by RRC signaling. One reordering buffer maps to one C/T MUX entity but many reordering buffer can map to the same C/T MUX entity. Note: De-/Ciphering is only performed for RLC-TM in MAC-d, for RLC-UM and RLC-AM de-/ciphering is performed in RLC. This means that RLC-TM is not supported on HS-DSCH as no frame number exists for HS-DSCH to count the 2 ms subframe continuously. Reordering / Disassembly The reordering entity organizes received HS-DSCH data blocks (MAC-hs PDUs) according to their TSN. MAC-hs PDUs with consecutive TSNs are delivered to higher layers upon reception. One MAC-hs PDU contains only MAC-d PDUs with the same priority and from the same MAC-d flow. Different MAC-d PDU sizes can be supported in a given MAC-hs PDU due to the size index parameter SID. MAC-d flow: 0, 1, 2 One or more queues can be associated with the same MAC-d flow. The MAC-hs queue ID is unique across all MAC-d flows. Thus the queue ID within MAC-hs PDU header identifies the associated MAC-d flow. If more than one DTCH is mapped onto the same MAC-d flow, the C/T field is necessary. A MAC-d flow can be associated with more than one queue. In such a case, the queues must have different IDs.
4-44
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-45
Version 1 Rev 0
UE Scheduling
HS-DSCH scheduler schedules all UEs within a cell.
Determining RV
The scheduler determines a suitable redundancy version for each transmitted and retransmitted MAC-hs PDU and indicates the redundancy version to lower layer ( HS-SCCH).
4-46
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-47
Version 1 Rev 0
NodeB HARQ
NodeB HARQ
HARQ Entity
There is one HARQ entity per UE in UTRAN. The HARQ entity sets the Queue ID in transmitted MAC-hs PDUs to the value indicated by the UTRAN scheduler; sets the transmission sequence number (TSN) in transmitted MAC-hs PDUs to the value indicated by the UTRAN scheduler; sets the HARQ process identifier in transmitted MAC-hs PDUs; determines a suitable HARQ process to service the MAC-hs PDU and set the HARQ process identifier accordingly.
HARQ Process
The HARQ process sets the NDI (new data indicator) in transmitted MAC-hs PDUs. The HARQ process sets the New Data Indicator to the value "0" for the first MAC-hs PDU transmitted by a HARQ process does not increment the NDI for retransmission of a MAC-hs PDU; increments the NDI with one for each transmitted MAC-hs PDU containing new data; processes received status messages; delivers received status messages e.g. Ack/Nack or DTX to the scheduler which prepares for new or retransmission<.
4-48
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
NodeB HARQ
Version 1 Rev 0
NodeB HARQ
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-49
Version 1 Rev 0
4-50
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-51
Version 1 Rev 0
UE MAC-hs Operation
UE MAC-hs Operation
The UE operation in support of HS-DSCH is split among the following functional units: HARQ entity, HARQ process, reordering entity and disassembly entity
UE HARQ
HARQ Entity There is one HARQ entity at the UE which processes the HARQ process identifiers received on the HS-SCCH transmissions associated with MAC-hs PDUs received on the HS-DSCH. A number of parallel HARQ processes are used in the UE to support the HARQ entity. The number of HARQ processes is configured RRC and limited to maximum of eight: Each received MAC-hs PDU shall be allocated to the HARQ process indicated by the HARQ process identifier of the MAC-hs PDU. HARQ Process Discard MAC-hs PDUs The UE may, if the MAC-hs PDU is received within 5 sub-frames from the reception of the previous MAC-hs PDU intended for this HARQ process; discard the MAC-hs PDU. If the NDI (one bit) has been incremented compared to the value in the previous received transmission in the very HARQ process or it is the first received transmission in the HARQ process, the UE shall replace the data being currently in the soft buffer. If the NDI is identical to the value used in the previous received transmission in the HARQ process, combine the received data with the data currently in the soft buffer for this HARQ process. Note: the HARQ process processes the Queue ID in the received MAC-hs PDU. The UE shall arrange the received MAC-hs PDUs in queues based on the Queue ID:
4-52
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
UE MAC-hs Operation
Version 1 Rev 0
UE MAC-hs Operation
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-53
Version 1 Rev 0
Timer-Based Scheme
Each time a missing packet stalls the delivery of packets to higher layers at the receiver, a timer T1 is started. If no other missing packets are thereafter detected then, once the long timer expires, the missing packet is assumed to have been lost, and all packets stalled by this missing packet are then delivered to higher layers. This mechanism requires the maintenance of one timer T1 per re-ordering queue (or a maximum of eight long timers for the eight re-ordering queues defined in W-CDMA Release 5).
4-54
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-55
Version 1 Rev 0
UE Reordering Entity
UE Reordering Entity
Parameters
TRANSMIT_WINDOW_SIZE is the size of the transmitter window. This is a parameter in the Node B and the value of the parameter is configured through NBAP signaling. The NodeB stops transmitting new MAC-hs PDUs if their TSN is out of the transmitters window size because there are too many unacknowledged MAC-hs PDUs in the transmit buffer. Note that this handling is performed per priority queue. RECEIVE_WINDOW_SIZE is the size of the receiver window. This is a parameter in the UE and the value of the parameter is configured through RRC signaling.
State Variables
All state variables are non-negative integers. MAC-hs PDUs are numbered by modulo integer TSN cycling through the field 0 to 63. All arithmetic operations on next_expected_TSN, RcvWindow_UpperEdge, T1_TSN and TSN_flush are affected by the 64 modulus. When performing arithmetic comparisons of state variables or TSN values a 64 modulus base is used. This modulus base is subtracted (within the appropriate field) from all the values involved and then an absolute comparison is performed. next_expected_TSN is the Transmission sequence number (TSN) following the TSN of the last in-sequence MAC-hs PDU received. It shall be updated upon the delivery to the disassembly entity of the MAC-hs PDU with TSN equal to next_expected_TSN. The initial value of next_expected_TSN =0. RcvWindow_UpperEdge represents the TSN, which is at the upper edge of the receiver window. After the first MAC-hs PDU has been received successfully, it also corresponds to the MAC-hs PDU with the highest TSN of all received MAC-hs PDUs. The initial RcvWindow_UpperEdge equals 63. RcvWindow_UpperEdge is updated based on the reception of new MAC-hs PDU. T1_TSN: The TSN of the latest MAC-hs PDU that cannot be delivered to the disassembly entity, when the timer T1 is started.
Timer T1
Re-ordering release timer T1 controls the stall avoidance in the UE reordering buffer. The value of T1 is configured by upper layers. ( through RRC).
Receiver Window
Receiver window: defines TSNs of those MAC-hs PDUs that can be received in the receiver without causing an advancement of the receiver window. The size of the receiver window equals RECEIVE_WINDOW_SIZE and spans TSNs going from RcvWindow_UpperEdge RECEIVE_WINDOW_SIZE + 1 to RcvWindow_UpperEdge included.
4-56
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
UE Reordering Entity
Version 1 Rev 0
UE Reordering Entity
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-57
Version 1 Rev 0
4-58
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-59
Version 1 Rev 0
MAC-hs Reset
If a reset of the MAC-hs entity is requested by upper layers, the UE shall: flush soft buffer for all configured HARQ processes; stop all active re-ordering release timer (T1) and set all timer T1 to their initial value; start TSN with value 0 for the next transmission on every configured HARQ process; initialize the variables RcvWindow_UpperEdge and next_expected_TSN to their initial values; disassemble all MAC-hs PDUs in the re-ordering buffer and deliver all MAC-d PDUs to the MAC-d entity; flush the re-ordering buffer. and then: indicate to all AM RLC entities mapped on HS-DSCH to generate a status report. MAC-hs reset is usually performed in case of the serving hs cell is changed from one NodeB to another when the HARQ soft information is lost. This could also be applied for cell changes also within the same NodeB if there are other implementation restrictions in the NodeB causing soft information to be lost at cell change.
4-60
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-61
Version 1 Rev 0
Activation Time
If the UE receives an RRC message where the IE "Activation time" has a value other (e.g. CFN) than the MD value "Now", the UE shall: in the case of HS-DSCH, let the CCTrCH including the associated DCH be the "reference CCTrCH", if the frame boundary immediately before the frame with the CFN value indicated by the IE "Activation Time" is at the TTI boundary common to all the transport channels that are multiplexed onto the reference CCTrCH: select that frame boundary as the activation time T. else: select the next TTI boundary, which is common to all the transport channels that are multiplexed onto the reference CCTrCH, after the frame with the CFN value indicated by the IE "Activation Time", as the activation time T. For an HS-DSCH related reconfiguration caused by the received RRC message: select the HS-SCCH subframe boundary immediately before the first HS-SCCH subframe, which entirely falls within the 10 ms frame following T; start using at that HS-SCCH subframe boundary the new HS-DSCH configuration in the received RRC message, replacing any old HS-DSCH configuration. Note: In FDD an "HS-DSCH related reconfiguration" includes, in particular, reconfigurations that need to be time-aligned with the 2ms subframe of the HS-SCCH, HS-PDSCH and/or HS-DPCCH. For example, start and stop of HS-SCCH reception and serving HS-DSCH cell change. If the UE receives a RRC message where the IE "Activation time" has the default value "Now", the UE shall: choose an activation time T as soon as possible after the reception of the message, respecting the performance requirements in 3GTS 25.331 (13.5.2) at the activation time T: perform the actions for the information elements in the received message as specified elsewhere
4-62
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-63
Version 1 Rev 0
NumOfPDU
Indicates the number of MAC-d PDUs in the payload. Value range: {1-255}; Field Length: 8 bits.
MAC-d PDU
According to NumOfPDU the amount of MAC-d PDU is contained in the payload of HS-DSCH DATA FRAME.
4-64
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-65
Version 1 Rev 0
4-66
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-67
Version 1 Rev 0
HS-DSCH Credits
The HS-DSCH Credits IE indicates the number of MAC-d PDUs that a CRNC may transmit during one HS-DSCH Interval granted in the HS-DSCH CAPACITY ALLOCATION Control Frame. Value range: {0-2047, where 0=stop transmission, 2047=unlimited}; Field length: 11 bits.
HS-DSCH Interval
The value of this field indicates the time interval during which the HS-DSCH Credits granted in the HS-DSCH CAPACITY ALLOCATION Control Frame may be used. The first interval starts immediately after reception of the HS-DSCH CAPACITY ALLOCATION Control Frame, subsequent intervals start immediately after the previous interval has elapsed. This value is only applied to the HS-DSCH transport channel. Value range: {0-2550 ms}. Value 0 shall be interpreted that none of the credits shall be used; Granularity: 10ms; Field Length: 8 bits.
4-68
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
4-69
Version 1 Rev 0
4-70
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
HSDPA Mobility
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
5-1
Version 1 Rev 0
HSDPA Mobility
5-2
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Objectives
Version 1 Rev 0
Objectives
After this Lecture the Student will be able to: Describe the serving HS-DSCH cell change State the various HSDPA mobility procedures Describe the measurement event 1D State intra and inter NodeB serving HS-DSCH cell change
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
5-3
Version 1 Rev 0
5-4
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
5-5
Version 1 Rev 0
HS-DSCH Handover
As can be seen in the figure, the UE moves from cell 1 towards cell 2. At a point in time, the P-CPICH_Ec/Io of cell 2 is better compared to cell 2s P-CPICH_Ec/Io. After a safe guard time of TTT (time to trigger), the UE reports event 1D to SRNC indicating cell 2s scrambling code and P-CPICH_Ec/Io and / or P-CPICH_RSCP value in order to inform SRNC about a better cell. TTT prevents in conjunction with H(1d) too frequent measurement reports and consequently too rapid cell changes. For example, the cell 2 must exceed cell 1s P-CPICH_Ec/Io quality level by H(1d) for a minimum time of TTT. Here we assume that CIO is set to 0 and thus not delaying or speeding up event 1D measurement reports. Once the SRNC has received the event 1D measurement report it may execute the HS-DSCH cell change procedure. Thus the HS-DSCH cell change is network controlled. The SRNC specifies with the activation time when the UE has to leave the old cell and change to the new cell. As this cell change involves RRC, NBAP and ALCAP signaling procedures in conjunction with an activation time, there is a certain delay until the serving HS-DSCH cell change is finished.
5-6
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
5-7
Version 1 Rev 0
MAC-hs Preservation
MAC-hs preservation means that in case of intra NodeB HS-DSCH serving cell change all the PDUs for the UE are moved from the MAC-hs in the source cell to the MAC-hs in the target cell. This means that the status of the HARQ manager is also preserved without triggering any higher layer retransmission such as RLC retransmission during intra NodeB HS-DSCH to HS-DSCH handover. If the NodeB does not support MAC-hs preservation, then the handling of the not completed PDUs is the same as in the inter NodeB case. Note: In the case of RLC-AM, the polling function can be used to obtain the status of the data transmission to the UE at the RLC level. So the SRNC can immediately forward all not yet acknowledged RLC PDUs to the new cell that were deleted in the source cell prior to the HS-DSCH cell change. In the case of RLC-UM the PDUs in the old cell waiting for transmission might be transferred to the target cell in case of intra NodeB HS-DSCH cell change but this depends on vendor implementation. During intra NodeB HS-DSCH to HS-DSCH handover, it is likely, that the UEs DPCH is in soft handover. It could be possible that the HS-DPCCH is also in soft handover condition, so both cells of the NodeB allocate Rake fingers for demodulation and receive a combined HS-DPCCH signal. This could improve uplink coverage of the HS-DPCCH and no power control problems are expected compared to the Inter NodeB soft handover.
5-8
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
5-9
Version 1 Rev 0
3. 4. 5.
6.
5-10
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
5-11
Version 1 Rev 0
5-12
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
5-13
Version 1 Rev 0
4. 5.
6. 7.
8. 9.
5-14
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
5-15
Version 1 Rev 0
5-16
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
MOTOROLA LTD.2002
Version 1 Rev 0
MOTOROLA LTD.2002
HSDPA System Architecture and Protocols FOR TRAINING PURPOSES ONLY - THIS MANUAL WILL NOT BE UPDATED
5-17