CLUSTERING IN WIRELESS SENSOR NETWORKS

flangeeasyMobile - Wireless

Nov 21, 2013 (3 years and 6 months ago)

61 views







CLUSTERING IN WIRELESS SENSOR

NETWORKS


B
Y

K
ALYAN

S
ASIDHAR

R
ESEARCH

P
ROBLEM


Understanding existing clustering algorithms and finding the problems
stated and addressed



Compare the pros and cons of each algorithm



Simulate algorithms and compare performance with and without clustering
mechanism




INTRODUCTION TO CLUSTERING




Grouping of similar objects or sensors in our context


distance or proximity


Logical organizing



Topology control approach


Load balancing, network scalability



Types of clustering


Static: local topology control



Dynamic: changing network parameters



Single hop and multi hop



Homogeneous and heterogeneous

HEED[1]

ADVANTAGES OF CLUSTERING


Transmit

aggregated

data

to

the

data

sink



reducing

number

of

nodes

taking

part

in

transmission



Useful

energy

consumption



Scalability

for

large

number

of

nodes



Reduces

communication

overhead

for

both

single

and

multi

hop


L
ITERATURE

S
URVEY

O
F

C
LUSTERING

A
LGORITHMS



HEED
: A hybrid energy efficient distributed clustering approach for ad
-
hoc sensor networks



MRECA
: Mobility resistant efficient clustering approach for ad
-
hoc
sensor networks



Energy efficient dynamic clustering algorithm for ad
-
hoc sensor
networks



LEACH
-
Energy efficient communication protocol for WSN



EEDC
-
Dynamic clustering and energy efficient routing technique for
WSN

Problem statement



Set of nodes, identify set of CHs that cover the entire network



Protocol distributed


Local information



One node
-
one cluster



Node
-
cluster head: single hop



CH
-
CH: multi hop using routing protocol

HEED

Assumptions


Sensor quasi
-
stationary



Links are symmetric



Energy consumption non
-
uniform for all nodes



Nodes
-
location unaware



Processing and communication capability
-
similar


Algorithm:


Cluster head selection



hybrid of residual energy (primary) and communication cost (secondary)
such as node proximity



Number of rounds of iterations



Tentative CHs formed



Final CH until CH
prob
=1



Same or different power levels used for intra cluster communication


Pros:


Balanced clusters



Low message overhead



Uniform & non
-
uniform node distribution



Inter cluster communication explained



Out performs generic clustering protocols on various factors


Cons:


Repeated iterations complex algorithm



Decrease of residual energy smaller probability


number of iterations increased



Nodes with high residual energy one region of a network


Future work:


Only two level hierarchy provided but can be extended to multilevel hierarchy



MRECA

Assumptions:


Sensor quasi
-
stationary





Nodes
-
location unaware





Every node as source and server


Algorithm:


Mobility resistant clustering approach



Deterministic time without iterations



Computed score value used to compute delay


Delay used CH announcement



Node mobility


Local maintenance performed instead of re
-
clustering


Pros


Clusters generated as node speed increased



Only one iteration against repeated iterations in HEED



Each node one message



saving on message transmission better energy efficiency



Robust against synchronization errors



Can be used for environmental monitoring and battlefield


applications

Cons



Inter cluster communication not explained



CH rotation mentioned but not explained ‘how




Future work


Extensive simulations on large scale
networks with elaborate power models,



Extensions to k
-
hop clusters and integration
of clustering with network applications



EEDC

Assumption:


Two tier hierarchy network


Routing limited to CHs route set up cost minimized



Sensors clustered


Algorithm:


Active node estimation and optimum probability of becoming cluster head


Received Signal power



Cluster formation


CH with a certain probability by wining a competition with neighbors



Data collection


Node
-
CH using MAC protocol
-
p
-
persistent CSMA



Data delivery


CH
-
BS
-
multi hop routing protocol



Pros


Number of clusters and CH
-
Dynamic


Energy dissipation
-
even distribution


Prolong network lifetime



most efficient for large
-
scale sensor network



Intra and inter cluster communication explained


Future work


Further investigating the applicability of the proposed clustering
technique and routing algorithm to more general wireless sensor
networks.


LEACH

Assumptions:


Fixed and remote base station



Nodes homogeneous and energy constrained



Radio channel is symmetric



E
A
-
E
B
=E
B
-
E
A



Sensing rate for all sensors fixed

Algorithm


CH position rotated among the nodes


energy load distributed .



Number of active nodes in the network and the optimal number of clusters
assumed
a priori



Nodes join a target number of CHs



Node
-
CH communication
-
TDMA


Pros


Incorporates data fusion into routing protocols


Amount of information to base station reduced



4
-
8 times effective over direct communication in prolonging network
lifetime



Grid like area

Cons


Only single hop clusters formed


Might lead to large number of clusters



No discussion on optimal CH selection



All CHs should directly transmit to the data sink




DYNAMIC CLUSTER



Energy efficiency distributed:


CH selection
-
both residual energy and P
T



Number of nodes
-
network size and P
T



CH
-
center of the cluster



Rotating CH to average power consumption



Breaking clusters and reforming


compensate for differences of power consumption in different areas



Unique route


Only CH with lowest ID and high residual energy


What is only one CH is present and that CH as low residual energy
?

Pros


Reduce flooding in route discovery



Avoid duplicate data transmission


Cons


Inter cluster communication not explained



Number of iterations needed for CH selection and cluster
formation not mentioned


CONCLUSIONS


Problem statement seems to be unique


Reduce energy consumption


Prolong network lifetime


Form set of clusters from a set of nodes


Cluster the whole network with the selected CH


Rotate CHs for energy distribution



Algorithms differ in CH selection and cluster formation



Some address intra and inter cluster communication



Some address real world applications



REFERENCES

[1]. A hybrid energy efficient distributed clustering
approach for ad
-
hoc sensor networks