2 NOKIA Training Center / RDM Objectives Objectives: This section explains the DCR8 Drop Call Rate formula and why it is better that the DCR 3G. It also explains how the new DCR BREAKDOWN OBSERVATION can be used as an effective tool to troubleshoot drop calls. There is also a brief description about AMR drops and RLT with AMR.
References: Michael Ledesma 3 NOKIA Training Center / RDM Drop Call Rate Formula The common formula is DCR3G and it is composed of the following counters : tch_abis_fail_call tch_abis_fail_old tch_lapd_fail tch_bts_fail tch_a_if_fail_call tch_a_if_fail_old tch_act_fail_call tch_tr_fail tch_tr_fail_old tch_user_act tch_bcsu_reset tch_netw_act tch_radio_fail tch_rf_old_ho The new drop call counter 1202 counts customer perceived drops. These are TCH failures that happens between ASSIGNMENT COMPLETE and DISCONNECT. This new counter has a smaller supervision period than the DCR3G counters thus lesser drop call counts but measures the exact user experience The succeeding slides shows the difference between DCR3G and the new formula DCR8E and why is it better to use 4 NOKIA Training Center / RDM 1202 (DCR8E) Counter Supervision Time DCR3G Start Point : Channel Activation Channel Activation command from the BSC. Voice Call has not yet been successfully established End Point : RF Channel Release Ack RF Channel Rel Ack received from the BTS 1202 (DCR8E) Start Point : Assignment Complete This implies that the voice or data call has been successfully established. End Point : Disconnect BSS and MS are aware of the disconnection. Call release continues.
Supervision Period of DCR3G Supervision Period of DCR8E 5 NOKIA Training Center / RDM Advantages The 1202 counter is much closer to customer experience when it comes to drop calls The counter has been verified and mapped together with the DCR BREAKDOWN OBSERVATION measurement (S11). This will give you the ability to drill down on the DCR investigation. Better drop call details in S11.5 with the enhanced DCR BREAKDOWN OBSERVATION This drop call counter was developed to match the drop call counters of other GSM vendors The counter is already available in S10.5 and S11 as 57044 and 1202 respectively
6 NOKIA Training Center / RDM Drop Call Breakdown Observation in S11 Detail Information per drop call
This observation can be activated in the BSC with no constraints of overload. It can be activated and available across all the network 24/7.
In S11, this observation provides:
TRX Information Phase In and Cause In Phase Out and Cause Out
The information above is already available and helps to better troubleshoot network incidents. This observation currently is not integrated in OSS, therefore it is required to FTP the BSC to access this information. Length : 25 Type : 29 Tracing Status : 0 Tracing Reference (hex) : 0000 Header Version : 13 Compatibility Version : 1 BTS ID : 20 Release time : 2005-05-13 10:48:50,63 Phase out : 2 Cause out : 368 TRX ID : 2 Phase In : 3 Cause in : 43
Drop Call Observation Example One drop call event, c1202 = 1 7 NOKIA Training Center / RDM Example of DCR Breakdown Observation Length : 25 Type : 29 Tracing Status : 0 Tracing Reference (hex) : 0000 Header Version : 13 Compatibility Version : 1 BTS ID : 20 Release time : 2005-05-13 10:48:50,63 Phase out : 15 Cause out : 317 TRX ID : 2 Phase In : 3 Cause in : 43
In S11 you can have the time, Phase out, Cause Out, Phase in, Cause in and TRX ID of the drop call Using the Call Related DX Causes in BSC document you can narrow down the cause of the drop call This gives you better information on finding a solution to improve your drop call rate. Drop occurred during conversation The Abis interface receives a CONNECTION FAILURE message and its cause is radio link failure Counter 1013 from the traffic table is pegged 8 NOKIA Training Center / RDM S11.5 DCR Breakdown Observation Enhancements The purpose of Breakdown of counter 1202 is to get more accurate data on dropped calls for BSC counter 'Drop After TCH Assignment. With the data provided by new observation, TCH drops happening after TCH is assigned can be analysed in TRX level with the actual reason for the drop. When the observation is activated in BSC, it will collect data from the whole BTSs under BSC. SUDDEN DROP ADJ CELL PLMN ID 1 ADJ CELL LAC 1 ADJ CELL ID 1 RX LEVEL ADJ 1 . . . ADJ CELL PLMN ID 6 ADJ CELL LAC 6 ADJ CELL ID 6 RX LEVEL ADJ 6
BTS ID RELEASE TIME PHASE OUT CAUSE OUT PHASE IN CAUSE IN TIMING ADVANCE DL RX LEVEL UL RX LEVEL DL RX QUALITY UL RX QUALITY DL FER UL FER DL LAST USED BIT RATE UL LAST USED BIT RATE Drop call Detail Information per Drop 9 NOKIA Training Center / RDM 5 Minute DCR BD OBS S11.5 Exercise 10 NOKIA Training Center / RDM Activity Attached is a DCR BD OBS file in txt format Using the Call Related DX Causes in BSC document. Try to figure out what caused the call to drop.
11 NOKIA Training Center / RDM AMR and Drop Calls The introduction of the AMR codec gives us the MS the ability to carry a conversation in C/I conditions where EFR cannot. This improvement in the speech codices bring it closer to the robust performance of the SACCH and at the worst conditions even outperform the SACCH. This can result in more radio related drops as the customer can carry the conversation but the signaling is lost (measurement reports) It is imperative then that we consider how we use AMR and the parameterization that it requires to operate without sacrificing the user perceived gains and still maintaining reasonable KPIs. More details about AMR will be discussed in the succeeding chapters. 12 NOKIA Training Center / RDM AMR and Radio Link Time Out Recommendations: EFR 20 SACCH periods AMR FR 32 to 44 SACCH periods AMR HR 20 to 32 SACCH periods AMR HR is not as robust as AMR FR. It is generally used in good C/I conditions. At RXQUAL 4, AMR HR is already giving poor voice performance that an extremely long RLT will artificially improve your DCR as the subscriber will terminate the call due to bad voice quality.