Chapter 4

Routing Protocols

Overview

Routing in WSNs is challenging due to distinguish from

other wireless networks like mobile ad hoc networks or
cellular networks.

First, it is not possible to build a global addressing scheme for a

large number of sensor nodes. Thus, traditional IP-based protoc
ols may not be applied to WSNs. In WSNs, sometimes getting t
he data is more important than knowing the IDs of which nodes
sent the data.

Second, in contrast to typical communication networks, almost

all applications of sensor networks require the flow of sensed d
ata from multiple sources to a particular BS.

Overview (cont.)

Third, sensor nodes are tightly constrained in terms of energ

y, processing, and storage capacities. Thus, they require care
fully resource management.
Fourth, in most application scenarios, nodes in WSNs are ge
nerally stationary after deployment except for, may be, a few
mobile nodes.
Fifth, sensor networks are application specific, i.e., design re
quirements of a sensor network change with application.
Sixth, position awareness of sensor nodes is important since
data collection is normally based on the location.
Finally, data collected by many sensors in WSNs is typically
based on common phenomena, hence there is a high probabil
ity that this data has some redundancy.

Overview (cont.)
The task of finding and maintaining routes in WSNs is
nontrivial since energy restrictions and sudden change
s in node status (e.g., failure) cause frequent and unpr
edictable topological changes.
To minimize energy consumption, routing techniques
proposed for WSNs employ some well-known routing
strategies, e.g., data aggregation and in-network proce
ssing, clustering, different node role assignment, and d
ata-centric methods were employed.

Outline
4.1 Routing Challenges and Design Issues in WSNs
4.2 Flat Routing
4.3 Hierarchical Routing
4.4 Location Based Routing
4.5 QoS Based Routing
4.6 Data Aggregation and Convergecast
4.7 Data Centric Networking
4.8 ZigBee
4.9 Conclusions

Chapter 4.1
Routing Challenges and Design Issues in
WSNs

Overview
The design of routing protocols in WSNs is influenced by man
y challenging factors. These factors must be overcome before e
fficient communication can be achieved in WSNs.

Node deployment
Energy considerations
Data delivery model
Node/link heterogeneity
Fault tolerance
Scalability
Network dynamics
Transmission media
Connectivity
Coverage
Data aggregation/convergecast
Quality of service

Node Deployment
Node deployment in WSNs is application dependent a
nd affects the performance of the routing protocol.
The deployment can be either deterministic or random
ized.
In deterministic deployment, the sensors are manually
placed and data is routed through pre-determined path
s.
In random node deployment, the sensor nodes are scat
tered randomly creating an infrastructure in an ad hoc
manner.

Energy Considerations
Sensor nodes can use up their limited supply of energ
y performing computations and transmitting informati
on in a wireless environment. Energy conserving form
s of communication and computation are essential.
In a multi-hop WSN, each node plays a dual role as da
ta sender and data router. The malfunctioning of some
sensor nodes due to power failure can cause significan
t topological changes and might require rerouting of p
ackets and reorganization of the network.

Data Delivery Model

Time-driven (continuous)

Event-driven

Suitable for applications that require periodic data monitorin

g
React immediately to sudden and drastic changes

Query-driven

Respond to a query generated by the BS or another node in t

he network

Hybrid
The routing protocol is highly influenced by the data r
eporting method

10

Node/Link Heterogeneity
Depending on the application, a sensor node can have
a different role or capability.
The existence of a heterogeneous set of sensors raises
many technical issues related to data routing.
Even data reading and reporting can be generated fro
m these sensors at different rates, subject to diverse Q
oS constraints, and can follow multiple data reporting
models.

11

Fault Tolerance
Some sensor nodes may fail or be blocked due to lack
of power, physical damage, or environmental interfere
nces
It may require actively adjusting transmission powers
and signaling rates on the existing links to reduce ener
gy consumption, or rerouting packets through regions
of the network where more energy is available

12

Scalability
The number of sensor nodes deployed in the sensing a
rea may be on the order of hundreds or thousands, or
more.
Any routing scheme must be able to work with this hu
ge number of sensor nodes.
In addition, sensor network routing protocols should b
e scalable enough to respond to events in the environ
ment.

13

Network Dynamics
Routing messages from or to moving nodes is more ch
allenging since route and topology stability become i
mportant issues
Moreover, the phenomenon can be mobile (e.g., a targ
et detection/ tracking application).

14

Transmission Media

In general, the required bandwidth of sensor data will

be low, on the order of 1-100 kb/s. Related to the trans
mission media is the design of MAC.

15

TDMA (time-division multiple access)

CSMA (carrier sense multiple access)

Connectivity
High node density in sensor networks precludes them
from being completely isolated from each other.
However, may not prevent the network topology from
being variable and the network size from shrinking du
e to sensor node failures.
In addition, connectivity depends on the possibly rand
om distribution of nodes.

16

Coverage
In WSNs, each sensor node obtains a certain view of t
he environment.
A given sensors view of the environment is limited in
both range and accuracy.
It can only cover a limited physical area of the enviro
nment.

17

Data Aggregation/Convergecast
Since sensor nodes may generate significant redundan
t data, similar packets from multiple nodes can be agg
regated to reduce the number of transmissions.
Data aggregation is the combination of data from diffe
rent sources according to a certain aggregation functio
n.
Convergecasting is collecting information upwards
from the spanning tree after a broadcast.

18

Quality of Service
In many applications, conservation of energy, which i
s directly related to network lifetime.
As energy is depleted, the network may be required to
reduce the quality of results in order to reduce energy
dissipation in the nodes and hence lengthen the total n
etwork lifetime.

19

Routing Protocols in WSNs: A taxonomy

Routing protocols in WSNs
Network Structure
Network Structure

20

Protocol Operation
Protocol Operation

Flat routing

Negotiation based routing

SPIN
Directed Diffusion (DD)
Hierarchical routing
LEACH
PEGASIS
TTDD
Location based routing
GEAR
GPSR

SPIN
Multi-path network routing
DD
Query based routing
DD, Data centric routing
QoS based routing
TBP, SPEED
Coherent based routing
DD
Aggregation
Data Mules, CTCCAP

Reference

J. N. Al-Karaki and A. E. Kamal, Routing techniques in wirele

ss sensor networks: a survey, IEEE Wireless Communications,
vol. 11, no. 6, pp. 6-28, Dec. 2004.

21

Chapter 4.2
Flat Routing

22

Overview

In flat network, each node typically plays the same role and sen
sor nodes collaborate together to perform the sensing task.
Due to the large number of such nodes, it is not feasible to assi
gn a global identifier to each node. This consideration has led t
o data centric routing, where the BS sends queries to certain re
gions and waits for data from the sensors located in the selecte
d regions. Since data is being requested through queries, attribu
te-based naming is necessary to specify the properties of data.
Prior works on data centric routing, e.g., SPIN and Directed Di
ffusion, were shown to save energy through data negotiation an
d elimination of redundant.

23

4.2.1
SPIN
Sensor Protocols for Information via Negotiation

24

SPIN -Motivation

Sensor Protocols for Information via Negotiation, SPI

N

A Negotiation-Based Protocols for Disseminating Informatio

n in Wireless Sensor Networks.

Dissemination is the process of distributing individual

sensor observations to the whole network, treating all
sensors as sink nodes

25

Replicate complete view of the environment

Enhance fault tolerance
Broadcast critical piece of information

SPIN (cont.)- Motivation

Flooding is the classic approach for dissemination
Source node sends data to all neighbors
Receiving node stores and sends data to all its neighbo
rs
Disseminate data quickly
Deficiencies

26

Implosion
Overlap
Resource blindness

SPIN (cont.)-Implosion
A
Node
The direction
of data sending
The connect
between nodes

x
C

B
x

x
D

27

SPIN (cont.)- Overlap

r
q
Node
The direction
of data sending
The connect
between nodes
The searching
range of the
node

(q, r)

(s, r)

C
28

SPIN (cont.)- Resource blindness

In flooding, nodes do not modify their activities based

on the amount of energy available to them.

A network of embedded sensors can be resource-awar

e and adapt its communication and computation to the
state of its energy resource.

29

SPIN (cont.)

Negotiation

Before transmitting data, nodes negotiate with each other to

overcome implosion and overlap
Only useful information will be transferred
Observed data must be described by meta-data

Resource adaptation

30

Each sensor node has resource manager

Applications probe manager before transmitting or processin
g data
Sensors may reduce certain activities when energy is low

SPIN (cont.)- Meta-Data

Completely describe the data

Must be smaller than the actual data for SPIN to be beneficia

l
If you need to distinguish pieces of data, their meta-data sho
uld differ

Meta-Data is application specific

31

Sensors may use their geographic location or unique node ID

Camera sensor may use coordinate and orientation

SPIN (cont.)- SPIN family

Protocols of the SPIN family

SPIN-PP
It is designed for a point to point communication, i.e., hopby-hop routing
SPIN-EC
It works similar to SPIN-PP, but, with an energy heuristic
added to it
SPIN-BC
It is designed for broadcast channels
SPIN-RL
When a channel is lossy, a protocol called SPIN-RL is use
d where adjustments are added to the SPIN-PP protocol to
account for the lossy channel.
32

SPIN (cont.)- Three-stage handshake protocol

SPIN-PP: A three-stage handshake protocol for point-t

o-point media

33

ADV data advertisement

Node that has data to share can advertise this by transmitti
ng an ADV with meta-data attached
REQ request for data
Node sends a request when it wishes to receive some actua
l data
DATA data message
Contain actual sensor data with a meta-data header
Usually much bigger than ADV or REQ messages

SPIN (3-Step Protocol)

AD
V

A
DAT

REQ

34

B
AD
V

REQ

RE
Q V
DA
T
AAD

DAD
V
A TA

REQ

DAA
DTV
A

Q
TA
AE
R
D
Q
V
RT
VEA
AD
A
DAD

SPIN (3-Step Protocol)

DATA

DA
TA
DA
TA

A
TA
DA

B
A
DAT

Notice the color of the data packets sent by node B

35

SPIN (3-Step Protocol)

DATA

DA
TA
DA
TA

A
TA
DA

B
A
DAT

SPIN effective when DATA sizes are large :

REQ, ADV overhead gets amortized
36

SPIN (cont.)- SPIN-EC (Energy-Conserve)

Add simple energy-conservation heuristic to SPIN-PP

SPIN-EC: SPIN-PP with a low-energy threshold
Incorporate low-energy-threshold
Works as SPIN-PP when energy level is high
Reduce participation of nodes when approaching low-energy-t
hreshold
When node receives data, it only initiates protocol if it can p
articipate in all three stages with all neighbor nodes
When node receives advertisement, it does not request the da
ta
Node still exhausts energy below threshold by receiving ADV
or REQ messages

37

SPIN (cont.)- Conclusion

SPIN protocols hold the promise of achieving high performanc

e at a low cost in terms of complexity, energy, computation, and
communication
Pros
Each node only needs to know its one-hop neighbors
Significantly reduce energy consumption compared to floodi
ng
Cons
Data advertisement cannot guarantee the delivery of data
If the node interested in the data are far from the source, da
ta will not be delivered
Not good for applications requiring reliable data delivery,
e.g., intrusion detection

38

SPIN (cont.)- Reference

J. Kulik, W.R. Heinzelman, and H. Balakrishnan, Negotiationbased protocols for disseminating information in wireless senso
r networks, Wireless Networks, Vol. 8, pp. 169-185, 2002.

39

4.2.2
Directed Diffusion
A Scalable and Robust Communication Parad
igm for Sensor Networks

40

Overview

Data-centric communication

Data is named by attribute-value pairs

Different form IP-style communication
End-to-end delivery service
e.g.
How many pedestrians do you obser
ve in the geographical region X?

A sensor field
Sources

Directed
Diffusion

Sink Node

41

Event
Event

Overview (cont.)

Data-centric communication (cont.)

Human operators query (task) is diffused

Sensors begin collecting information about query
Information returns along the reverse path
Intermediate nodes aggregate the data
Combing reports from sensors

Directed Diffusion is an important milestone in the dat

a centric routing research of sensor networks

42

Directed Diffusion

Typical IP based networks

Requires unique host ID addressing

Application is end-to-end

Directed diffusion use publish/subscribe

43

Inquirer expresses an interest, I, using attribute values

Sensor sources that can service I, reply with data

Directed Diffusion (cont.)

Directed diffusion consists of

44

Interest - Query which specifies what a user wants

Data - Collected information
Gradient
Direction and data-rate
Events start flowing towards the originators of interests
Reinforcement
After the sink starts receiving events, it reinforces at least
one neighbor to draw down higher quality events

Data Naming

Expressing an Interest

Using attribute-value pairs

e.g.,

Type = Wheeled vehicle

// detect vehicle location
Interval = 20 ms
// send events every 20ms
Duration = 10 s
// Send for next 10 s
Rect = [-100,100, 200,400] // from sensors in this area

45

Interests and Gradients

Interest propagation

The sink broadcasts an interest

Exploratory interest with low data-rate
Neighbors update interest-cache and forwards it
Flooding
Geographic routing
Use cached data to direct interests

Gradient establishment

46

Gradient set up to upstream neighbor

Low data-rate gradient
Few packets per unit time needed

Gradient Set Up

Inquirer (sink) broadcasts exploratory interest, i1

Intended to discover routes between source and sink

Neighbors update interest-cache and forwards i1

Gradient for i1 set up to upstream neighbor

47

No source routes
Gradient a weighted reverse link
Low gradient Few packets per unit time needed

Exploratory Gradient
Exploratory Request
Gradient
Event
Event

Low Data-rate Interest

Low Data-rate Interest

48

Low Data-rate Interest

Data Propagation
A sensor node that detects a target
Search its interest cache
Compute the highest requested data-rate among all i
ts outgoing gradients
Data message is unicast individually
A node that receives a data message
Find a matching interest entry in its cache
Check the data cache for loop prevention
Re-send the data to neighbors

49

Reinforcement (1/4)

Positive reinforcement
Sink selects the neighboring node
Original interest message but with high data-rate
Neighboring node must also reinforce at least one neighbor
Low-delay path is selected
Exploratory gradients still exist: useful for faults

Event
Event

Reinforced gradient
Reinforced gradient

Source

A sensor field
Sink
50

Reinforcement (2/4)

Path failure and recovery

Link failure detected by reduced rate, data loss
Choose next best link (i.e., compare links based on infrequ
ent exploratory downloads)
Negatively reinforce lossy link
Either send interest with base (exploratory) data rate or all
ow neighbors cache to expire over time
Event
Event

D
M

Source

51

A
Sink

Link
Link A-M
A-M lossy
lossy
A
A reinforces
reinforces B
B
B
B reinforces
reinforces C
C
C
C reinforces
reinforces D
D
or
or
A
A negative
negative reinforces
reinforces M
M
M
M negative
negative reinforces
reinforces D
D

Reinforcement (3/4)

Multipath routing

Consider each gradients link quality

Using negative reinforcement

52

B
Source
Event
Event

Path Truncation
Loop removal
For resource saving
Ex:
B gets same data from both A and D, but
D always delivers late due to looping
B negative reinforces D, D negative rein
forces E, E negative reinforces B
Loop BE D B eliminated
Conservative negative reinforces useful for
fault resilience

Sink

Multiple paths

D
A

E
B

A removable loop

Design Considerations

Design Space for Diffusion

Diffusion element

Design Choices

Interest Propagation

Flooding
Constrained or directional flooding based on location
Directional propagation based on previously cached data

Data Propagation

Reinforcement to single path delivery

Multipath delivery with selective quality along different
paths
Multipath delivery with probabilistic forwarding

Data caching and

aggregation

For robust data delivery in the face of node failure

For coordinated sensing and data reduction
For directing interests

Reinforcement

Rules for deciding when to reinforce

Rules for how many neighbors to reinforce
Negative reinforcement mechanisms and rules

53

Directed Diffusion: Pros & Cons

Different from SPIN in terms of on-demand data querying mec

hanism
Sink floods interests only if necessary (lots of energy savings)
In SPIN, sensors advertise the availability of data
Pros
Data centric: All communications are neighbor to neighbor w
ith no need for a node addressing mechanism
Each node can do aggregation & caching
Cons
On-demand, query-driven: Inappropriate for applications req
uiring continuous data delivery, e.g., environmental monitori
ng
Attribute-based naming scheme is application dependent
For each application it should be defined a priori
Extra processing overhead at sensor nodes
54

Conclusions

55

Directed diffusion, a paradigm proposed for event monitoring s

ensor networks
Directed Diffusion has some novel features - data-centric disse
mination, reinforcement-based adaptation to the empirically be
st path, and in-network data aggregation and caching.
Notion of gradient (exploratory and reinforced)
Energy efficiency achievable
Diffusion mechanism resilient to fault tolerance
Conservative negative reinforcements proves useful

References

C. Intanagonwiwat, R. Govindan, and D. Estrin, Directed Diff

usion: A Scalable and Robust Communication Paradigm for Se
nsor Networks, in the Proceedings of the Sixth Annual Interna
tional Conference on Mobile Computing and Networks (MobiC
om00), August 2000.
C. Intanagonwiwat, R. Govindan, D. Estrin, J. Heidemann, and
F. Silva, Directed Diffusion for Wireless Sensor Networking,
IEEE/ACM Transactions on Networking, Vol. 11, No. 1, Feb. 2
003.

56

Chapter 4.3
Hierarchical Routing

57

Overview

In a hierarchical architecture, higher energy nodes can be used

to process and send the information while low energy nodes ca
n be used to perform the sensing of the target.
Hierarchical routing is mainly two-layer routing where one lay
er is used to select cluster heads and the other layer is used for
routing.
Hierarchical routing (or cluster-based routing), e.g., LEACH, P
EGASIS, TTDD, is an efficient way to lower energy consumpti
on within a cluster and by performing data aggregation and fusi
on in order to decrease the number of transmitted messages to t
he base stations.

58

4.3.1
LEACH
Low-Energy Adaptive Clustering Hierarchy

59

LEACH

LEACH (Low-Energy Adaptive Clustering Hierarchy), a cluste

ring-based protocol that minimizes energy dissipation in sensor
networks.
LEACH outperforms classical clustering algorithms by using a
daptive clusters and rotating cluster-heads, allowing the energy
requirements of the system to be distributed among all the sens
ors.
LEACH is able to perform local computation in each cluster to
reduce the amount of data that must be transmitted to the base s
tation.
LEACH uses a CDMA/TDMA MAC to reduce inter-cluster an
d intra-cluster collisions.
60

LEACH (cont.)
Sensors elect themselves to be local cluster-heads at a
ny given time with a certain probability.
Each sensor node joins a cluster-head that requires the
minimum communication energy.
Once all the nodes are organized into clusters, each cl
uster-head creates a transmission schedule for the nod
es in its cluster.
In order to balance the energy consumption, the cluste
r-head nodes are not fixed; rather, this position is selfelected at different time intervals.

61

LEACH

62

LEACH: Adaptive Clustering

Periodic independent self-election

Probabilistic
CSMA MAC used to advertise
Nodes select advertisement with strongest signal strength
Dynamic TDMA cycles

All nodes marked with a given symbol belong to the same cluster, and
the cluster head nodes are marked with a .
63

Algorithm
Periodic process
Two phases per round:

Setup phase

Advertisement: Execute election algorithm

Members join to cluster

Cluster-head broadcasts schedule

64

Steady-State phase
Data transmission to cluster-head using TDMA
Cluster-head transfers data to BS (Base Station)

Algorithm (cont.)
Fixed-length cycle

Setup phase

Steady-state phase
Time slot Time slot Time slot
1
2
3

Advertisement phase

65

Self-election of cluster
heads
Cluster heads compete
with CSMA

Cluster setup phase

Members
compete with
CSMA

Broadcast schedule

Cluster head Broadcast

CDMA code to members

Algorithm Summary

Set-up phase

Node n choosing a random number m between 0 and 1

If m < T(n) for node n, the node becomes a cluster-head where

T (n )

66

P
if n G
1 P [r * mod(1/ P )]
0
otherwise ,

where P = the desired percentage of cluster heads (e.g., P= 0.05), r=the c

urrent round, and G is the set of nodes that have not been cluster-heads in
the last 1/P rounds. Using this threshold, each node will be a cluster-head
at some point within 1/P rounds. During round 0 (r=0), each node has a p
robability P of becoming a cluster-head.

Algorithm Summary (cont.)

Set-up phase
Cluster heads assign a TDMA schedule for their members w
here each node is assigned a time slot when it can transmit.
Each cluster communications using different CDMA codes t
o reduce interference from nodes belonging to other clusters.
TDMA intra-cluster
CDMA inter-cluster
Spreading codes determined randomly
Broadcast during advertisement phase

67

Algorithm Summary (cont.)

Steady-state phase

68

All source nodes send their data to their cluster heads

Cluster heads perform data aggregation/fusion through local
transmission
Cluster heads send aggregated data back to the BS using a si
ngle direct transmission

An Example of a LEACH Network

While neither of these diagrams is the optimum scenario, the se

cond is better because the cluster-heads are spaced out and the
network is more properly sectioned

Good case scenario

69

Bad case scenario

Node
Cluster-Head Node
X Node that has been cluster-head in the last 1/P rounds
Cluster Border

Conclusions

Advantages
Increases the lifetime of the network
Even drain of energy
Distributed, no global knowledge required
Energy saving due to aggregation by CHs

Disadvantages
LEACH assumes all nodes can transmit with enough power t
o reach BS if necessary (e.g., elected as CHs)
Each node should support both TDMA & CDMA
Need to do time synchronization
Nodes use single-hop communication
70

Reference

W. Heinzelman, A. Chandrakasan, and H. Balakrishnan, Ener

gy-efficient communication protocol for wireless sensor netwo
rks, Proceedings of the 33rd Hawaii International Conference
on System Sciences, January 2000.

71

Chapter 4.4
Location Based Routing

72

Overview

Sensor nodes are addressed by means of their locations.

The distance between neighboring nodes can be estimated on the basis of

incoming signal strengths.
Relative coordinates of neighboring nodes can be obtained by exchangin
g such information between neighbors.

To save energy, some location based schemes demand that nod

es should go to sleep if there is no activity.
More energy savings can be obtained by having as many sleepi
ng nodes in the network as possible.
Hereby, two important location based routing protocols, GEAR
and GPSR, are introduced.

73

Geographical and Energy Aware Routing (GEAR)

Greedy Perimeter Stateless Routing (GPSR)

4.4.1
GEAR
Geographical and Energy Aware Routing

74

Geographical and Energy Aware Routing

(GEAR)

The protocol, called Geographic and Energy Aware Routing (G

EAR), uses energy aware and geographically-informed neighb
or selection heuristics to route a packet towards the destination
region.
The key idea is to restrict the number of interests in directed dif
fusion by only considering a certain region rather than sending
the interests to the whole network. By doing this, GEAR can c
onserve more energy than directed diffusion.
The basic concept comprises of two main parts
Route packets towards a target region through geographical a
nd energy aware neighbor selection
Disseminate the packet within the region

75

Energy Aware Neighbor Computation

Each node N maintains state h(N, R) which is called learned co

st to region R, where R is the target region
Each node infrequently updates neighbor of its cost
When a node wants to send a packet, it checks the learned cost
to that region of all its neighbors
If a node does not have the learned cost of a neighbor to a regio
n, the estimated cost is computed as follows:
c(Ni, R) = d(Ni, R) + (1-)e(Ni)
where
= tunable weight, from 0 to 1.

d(Ni, R) = normalized the largest distance among neighbors of N

e(Ni) = normalized the largest consumed energy among neighbors
of N
76

Energy Aware Neighbor Computation (con

t.)
When a node wants to forward a packet to a destinatio
n, it checks to see if it has any neighbor closer to desti
nation than itself
In case of multiple choices, it aims to minimize the lea
rned cost h(Nmin, R)

It then sets its own cost to:

h(N, R) = h(Nmin, R) + c(N, Nmin)
c(N, Nmin) = the transmission cost from N and Nmin

77

Forwarding Around Holes

BT= 5
F

CT=2

xH

h(C,T) = h(B,T)+c(C,B)
5
C

is set to 1. Initially, at time 0, at node S, among all neighbors of S, B, C, D

are closer to T than S. h(B,T)=c(B,T)= 5 , h(C,T)=c(C,T)=2, h(D,T)=c(D,T)= 5 .
After that , h(C, T) = h(B, T) + 1
78

Recursive Geographic Forwarding

Once the target region is reached, the packets are disseminated

within the region by recursive geographic forwarding
Forwarding stops when a node is the only one in a sub-region

Ni

79

Recursive Geographic Forwarding (cont.)

When network density is low recursive geographic for

warding is subject to two pathologies: inefficient trans
missions and non-termination

80

Recursive Geographic Forwarding (cont.)

Inefficient Transmission

Recursive geographic forwarding vs. Restricted flooding

A

Restricted flooding 1 time for

Recursive Geographic
sending and 4 times for receiving
Forwarding 3 times for sending
= consuming
and 3 times for receiving =
5 units of energy
consuming 6 units of energy

B
D

81

Recursive Geographic Forwarding (cont.) P

athologies

Non-Termination

In the recursive geographic forwarding protocol, packet forwa

rding terminates when the target subregion is empty.
K
E

B
A

C
L
F

82

Recursive Geographic Forwarding (cont.) P

roposed solution for pathologies

Solution:
Node degree is used as a criteria to differentiate low d
ensity networks from high density ones
Choice of restricted flooding over recursive geographi
c forwarding if the receivers node degree is below a t
hreshold value

83

Conclusion
GEAR strategy attempts to balance energy consumpti
on and thereby increase network lifetime
GEAR performs better in terms of connectivity after i
nitial partition

84

References

Y. Yu, D. Estrin, and R. Govindan, Geographical and Energy-Aware

Routing: A Recursive Data Dissemination Protocol for Wireless Sens
or Networks, UCLA Computer Science Department Technical Repor
t, UCLA-CSD TR-01-0023, May 2001.
Nirupama Bulusu, John Heidemann, and Deborah Estrin. Gps-less l
ow cost outdoor localization for very small devices. IEEE Personal
Communications Magazine, 7(5):28-34, October 2000.
L. Girod and D. Estrin. Robust range estimation using acoustic and
multimodal sensing. In IEEE/RSJ International Conference on Intel
ligent Robots and Systems (IROS 2001), Maui, Hawaii, October 2001
.
Nissanka B. Priyantha, Anit Chakraborty, and Hari Balakrishnan. T
he cricket location-support system. In Proc. ACM Mobicom, Boston
, MA, 2000.
Andreas Savvides, Chih-Chieh Han, and Mani B. Strivastava. Dyna
mic fine-grained localization in adhoc networks of sensors. In Proc.
ACM Mobicom, 2001.
85

4.4.2
GPSR
Greedy Perimeter Stateless Routing

86

Greedy Perimeter Stateless Routing (GPS

R)

Greedy Perimeter Stateless Routing (GPSR) proposes the aggr

essive use of geography to achieve scalability
GEAR was compared to a similar non-energy-aware routing pr
otocol GPSR, which is one of the earlier works in geographic r
outing that uses planar graphs to solve the problem of holes
In case of GPSR, the packets follow the perimeter of the planar
graph to find their routes
Although the GPSR approach reduces the number of states a n
ode should keep, it has been designed for general mobile ad ho
c networks and requires a location service to map locations and
node identifiers.

87

Algorithm & Example

The algorithm consists of two methods:

greedy forwarding + perimeter forwarding

Greedy forwarding, which is used wherever possible,

and perimeter forwarding, which is used in the regions
greedy forwarding cannot be done.

88

Greedy Forwarding (cont.)

Under GPSR, packets are marked by their originator
with their destinations locations
As a result, a forwarding node can make a locally opti
mal, greedy choice in choosing a packets next hop
Specifically, if a node knows its radio neighbors posit
ions, the locally optimal choice of next hop is the neig
hbor geographically closest to the packets destination
Forwarding in this scheme follows successively closer
geographic hops, until the destination is reached

89

Greedy Forwarding (cont.)

x
y

90

Greedy Forwarding (cont.)

A simple beaconing algorithm provides all nodes with
their neighbors positions: periodically, each node tran
smits a beacon to broadcast MAC address, containing
its own identifier (e.g., IP address) and position
Position is encoded as two four-byte floating point qu
antities, for x and y coordinate values
Upon not receiving a beacon from a neighbor for long
er than timeout interval T, a GPSR router assumes that
the neighbor has failed or gone out-of-range, and delet
es the neighbor from its neighbor table

91

Greedy Forwarding (cont.)

The Problem of Greedy Forwarding

D
v

|xD|<|wD|and|yD|
x will not choose to
forward to w or y
using greedy
forwarding

void
w

x
x

92

The Right-Hand Rule: Perimeters

In previous works, use the right-hand rule to map perimeters b

y sending packets on tours of them. The state accumulated in th
ese packets is cached by nodes, which recover from local maxi
ma in greedy forwarding by routing to a node on a cached peri
meter closer to the destination.
This approach requires a heuristic, the no-crossing heuristic, to
force the right-hand rule to find perimeters that enclose voids i
n regions where edges of the graph cross
x

93

Right-Hand Rule Does Not Work with

Cross Edges

z
u

x originates a packet to u
Right-hand rule results in the
tour x-u-z-w-u-x
94

Remove Crossing Edge

z
u

Make the graph planar

w
x

Remove (w,z) from the

graph
Right-hand rule results in
the tour x-u-z-v-x
95

Make a Graph Planar

A graph in which no two edges cross is known as plan
ar. A set of nodes with radios, where all radios have id
entical, circular radio range r, can be seen as a graph:
each node is a vertex, and edge (n, m) exists between
nodes n and m if the distance between n and m, d(n,
m)r.
Convert a connectivity graph to planar non-crossing g
raph by removing bad edges
Ensure the original graph will not be disconnected
Two types of planar graphs:

Relative Neighborhood Graph (RNG)

Gabriel Graph (GG)

96

Planarized Graphs (cont.)

Relative Neighborhood Graph (RNG)

97

Planarized Graphs (cont.)

Gabriel Graph (GG)

w
u

98

Planarized Graphs (cont.)

An algorithm for removing edges from the graph that
are not part of the RNG or GG would yield a network
with no crossing links
The RNG is a subset of the GG

It is because RNG removes more edges

Hereby, the RNG is used

If the original graph is connected, RNG is also connec
ted

99

Connectedness of RNG Graph

Key observation

Any edge on the minimum spanning tr

ee of the original graph is not removed
Proof by contradiction: Assume
(u,v) is such an edge but removed in R
NG
w

v
100

Planarized Graphs (cont.)

Original

Gabriel Graph (GG)

The full graph of a radio

The GG subset of
network, 200 nodes, uniformly
the full graph
randomly placed on a 2000 x
2000 meter region, with a radio
range of 250 m.
101

Relative
Neighborhood Graph
(RNG)

The RNG subset of the

full and GG graphs.

Combining Greedy and Planar Perimeters

All data packets are marked initially at their originators as gree

dy mode
GPSR packet headers include a flag field indicating whether th
e packet is in greedy mode or perimeter mode
Packet sources also include the geographic location of the desti
nation in packets
Only a packets source sets the location destination field, it is le
ft unchanged as the packet is forwarded through the network
Upon receiving a greedy-mode packet for forwarding, a node s
earches its neighbor table for the neighbor geographically close
st to the packets destination
When no neighbor is closer, the node marks the packet into per
imeter mode
102

GPSR

greedy fails

Greedy Forwarding

Perimeter Forwarding

have left local maxima

greedy works

greedy fails

103

Combining Greedy and Planar Perimeters

(cont.)
D

Lf
e0

104

Lp

If forwarding node to D < Lp to D,

returns a packet to greedy mode

Conclusion
GPSRs benefits all stem from geographic routings us
e of only immediate-neighbor information in forwardi
ng decisions.
GPSR keeps state proportional to the number of its nei
ghbors, while both traffic sources and intermediate DS
R routers cache state proportional to the product of the
number of routes learned and route length in hops.

105

References

B. Karp and H. T. Kung, Greedy Perimeter Stateless Routing f

or Wireless Networks, Proc. 6th Annual ACM/IEEE Int'l. Con
f. Mobile Comp. Net., Boston, MA, pp. 243-54, August 2000.
G. G. Finn, Routing and addressing problems in large metropo
litan-scale internetworks, Tech. Rep. ISI/RR-87-180, Informat
ion Sciences Institute, March 1987.
S. Floyd and V. Jacoboson, The synchronization of periodic ro
uting messages, IEEE/ACM Transactions on Networking, Vol.
2, pp. 122-136, April 1994.
B. Karp Greedy perimeter state routing, Invited Seminar at t
he USC/Information Sciences Institute, July 1998.
J. Saltzer, D. P. Reed, and D. Clark, End-to-end arguments in
system design, ACM Transactions on Computer Systems, Vol.
2, No. 4, Pages: 277-288, November 1984.
106

Chapter 4.5
QoS Based Routing

107

Overview
QoS is the performance level of service offered by a n
etwork to the user.
The goal of QoS is to achieve a more deterministic net
work behavior so that the information carried by the n
etwork can be better delivered and the resources can b
e better utilized.
In QoS-based routing protocols, the network has to ba
lance between energy consumption and data quality.
In particular, the network has to satisfy certain QoS m
etrics, e.g., delay, energy, bandwidth, etc. when delive
ring data to the BS.

108

Parameters of QoS Networks

Different services require different QoS parameters

Multimedia
Bandwidth, delay jitter & delay
Emergency services
Network availability
Group communications
Battery life

Generally the parameters that are important are:

109

bandwidth
delay jitter
battery charge
processing power
buffer space

Challenges in QoS Routing

Dynamically varying network topology
Imprecise state information
Lack of central coordination
Hidden node problem
Limited resource
Insecure medium

110

4.5.1
TBP (Ticket-Based Probing)
QoS of Bandwidth

111

Ticket-Based Probing

Distributed multi path QoS routing scheme

Bandwidth-constrained routing and delay-constrained routing

There are numerous paths from source to destination, w

e shall not randomly pick several paths to search
We shall not use any flooding path-discovery approach
es, which may send routing messages to the entire netw
ork
Multipath search is tolerant to imprecise information
We want to make an intelligent hop-by-hop path selecti
on to guide the search along the best candidate paths

112

Ticket-Based Probing (cont.)

S
D

113

Ticket-Based Probing (cont.)

A ticket is the permission to search one path. The sour
ce node issues a number of tickets based on the availab
le state information
Utilizes tickets to limit the number of paths searched d
uring route discovery

A ticket is the permission to search a single path

More tickets, more QoS constraints are required

Probes (routing messages) are sent from the source to

ward the destination to search for a low-cost path that
satisfies the QoS requirement
Each probe is required to carry at least one ticket

114

Ticket-Based Probing (cont.)

P1
(1 )

)
P4(2

j P
3(3
)

115

3)

P2(
2)

P3(

P1(1)

Ticket-Based Probing (cont.)

A

Demand = 3

S
3
B

2
6

x
2

116

5
E

Ticket-Based Probing (cont.)

A

Demand = 4

(1.1,3)
3

(1.1,3)
3

3
(1.2,1) 2
B
117

2
2

6
(1.2,1)

5
E

(1.2,1)

Ticket-Based Probing (cont.)

A

Demand = 4

(1.1,3)

(1.1.1,2)

3
(1.2,1)

B
118

(1.1.2,1)
(1.1.2,1) 5
(1.2,1)
2

2
6

(1.2,1)

Ticket-Based Probing (cont.)

T1

T1

D
T2

T2
T1

119

T2

Ticket-Based Probing (cont.)

A

Demand = 4

x
3

(1,4)

(2.1,3)

(2.2,1)

x2

120

(2.1,3)

3
6

(2.1,3)

(2.2,1)

(2.1,3)

(2.2,1)

Conclusion

The routing overhead is controlled by the number of tickets, w

hich allows the dynamic tradeoff between the overhead and the
routing performance. Issuing more tickets means searching mo
re paths, which results in a better chance of finding a feasible p
ath at the cost of higher overhead.
A distributed routing process is used to avoid any centralized p
ath computation that could be very expensive for QoS routing i
n large networks.
This approach not only increases the chance of success but also
improves the ability to tolerate the information imprecision bec
ause the intermediate nodes may gradually correct a wrong dec
ision made by the source.
121

Conclusion (cont.)

Ticket-based probing scheme achieves a balance between the si

ngle-path routing algorithms and the flooding algorithms. It do
es multipath routing without flooding.
The basic idea is to achieve a near-optimal performance with m
odest overhead by using a limited number of tickets and makin
g intelligent hop-by- hop path selection.

122

References

S. Chen and K. Nahrstedt, On finding multi-constrained paths, in Pro

c. IEEE ICC98, pp. 874-879.
R. Guerin and A. Orda, QoS-based routing in networks with inaccurate
information: Theory and algorithms, in Proc. IEEE INFOCOM97, Jap
an, pp. 75-83.
Q. Ma and P. Steenkiste, Quality-of-service routing with performance g
uarantees, in Proc. 4th Int. IFIP Workshop Quality of Service, May 199
7, pp. 115-126.
Z. Wang and J. Crowcroft, QoS routing for supporting resource reserva
tion, IEEE J. Select. Areas Commun., Sept. 1996.
S. Chen and K Nahrstedt, Distributed Quality-of-Service Routing in A
d Hoc Networks, IEEE J. Select. Areas Commun, vol.17, no. 8, pp. 14
88-1505, Aug. 1999.

123

References

T. Hea, J. A Stankovic, C. Lu, and T. Abdelzaher, SPEED: a st

ateless protocol for real-time communication in sensor network
s, in Proc. IEEE International Conference on Distributed Com
puting Systems, pp. 46-55, May 2003.
G. S. Ahn, A. T. Campbell, A. Veres, and L.H. Sun. SWAN: Se
rvice Differentiation in Stateless Wireless Ad Hoc Networks, I
n Proc. IEEE INFOCOM'2002, June 2002.

124