clustering in sensor networks

Clustering in Sensor Networks

– The data collected by each sensor is communicated through the The data collected by each sensor is communicated through the

network to a single processing center that uses the datanetwork to a single processing center that uses the data

– ClusteringClustering sensors into groups such that sensors communicate sensors into groups such that sensors communicate

information only to clusterheads and then the clusterheads information only to clusterheads and then the clusterheads

communicate the aggregated information to the processing center, communicate the aggregated information to the processing center,

saving energy and bandwidthsaving energy and bandwidth

– The cost of transmitting a bit is higher than a computation; therefore, it The cost of transmitting a bit is higher than a computation; therefore, it

may be beneficial to organize the sensors into clustersmay be beneficial to organize the sensors into clusters

– Cluster-based control structures provides more efficient use of Cluster-based control structures provides more efficient use of

resources in wireless sensor networksresources in wireless sensor networks

Clustering can be used forClustering can be used for

– Transmission management Transmission management

– Backbone formationBackbone formation

– Routing Efficiency Routing Efficiency

Why Clustering?

– This paper proposes a distributed, randomized clustering algorithm to This paper proposes a distributed, randomized clustering algorithm to

organize the sensors in a wireless sensor network into clusters to minimize organize the sensors in a wireless sensor network into clusters to minimize

the energy used to communicate information from all nodes to the the energy used to communicate information from all nodes to the

processing centerprocessing center

– By the generation of hierarchy of clusterheads, the energy savings increase By the generation of hierarchy of clusterheads, the energy savings increase

with the number of levels in the hierarchywith the number of levels in the hierarchy

– Sensor detects events and then communicate the collected information to a Sensor detects events and then communicate the collected information to a

central location where parameters characterizing these events are central location where parameters characterizing these events are

estimatedestimated

– In the clustered environment, the data gathered by the sensors is In the clustered environment, the data gathered by the sensors is

communicated to the data processing center through a hierarchy of communicated to the data processing center through a hierarchy of

clusterheads clusterheads

– The processing center determines the final estimates of the parameters The processing center determines the final estimates of the parameters

using information communicated by the clusterheadsusing information communicated by the clusterheads

An Energy Efficient Hierarchical Clustering Algorithm for Wireless Sensor Networks[Bandyopadhyay+, 2003]

– The processing center can be a specialized device or one of the sensors The processing center can be a specialized device or one of the sensors

itselfitself

– In such clustered environment, sensor data is communicated over smaller In such clustered environment, sensor data is communicated over smaller

distances, the energy consumed in the network will be much lower than the distances, the energy consumed in the network will be much lower than the

energy consumption when every sensor communicates directly to the energy consumption when every sensor communicates directly to the

information processing centerinformation processing center

– The results in stochastic geometry are used to derive values of parameters The results in stochastic geometry are used to derive values of parameters

for the algorithm that minimize the energy spent in the sensor networkfor the algorithm that minimize the energy spent in the sensor network


A New, Energy-Efficient, Single-Level Clustering AlgorithmA New, Energy-Efficient, Single-Level Clustering Algorithm

– Each sensor becomes a clusterhead (CH) with probability Each sensor becomes a clusterhead (CH) with probability p p and advertises and advertises

itself as a clusterhead to the sensors within its radio range – these itself as a clusterhead to the sensors within its radio range – these

clusterheads are called clusterheads are called volunteer clusterheadsvolunteer clusterheads

– This advertisement is forwarded to all the sensors that are no more than This advertisement is forwarded to all the sensors that are no more than kk

hops away from the clusterheadhops away from the clusterhead

– Any sensor node that is not clusterhead itself receiving such advertisement Any sensor node that is not clusterhead itself receiving such advertisement

joins the cluster of the closest clusterheadjoins the cluster of the closest clusterhead

– Any sensor node that is neither a clusterhead nor has joined any cluster Any sensor node that is neither a clusterhead nor has joined any cluster

itself becomes a clusterhead – called itself becomes a clusterhead – called forced clusterheadsforced clusterheads

– Since the advertisement forwarding has been limited to Since the advertisement forwarding has been limited to kk hops, if a sensor hops, if a sensor

does not receive a CH advertisement within time duration does not receive a CH advertisement within time duration tt (where (where tt is the is the

time required for data to reach the CH from any sensor time required for data to reach the CH from any sensor kk hops away), it hops away), it

means that the sensor node is not within k hops of any volunteer CHsmeans that the sensor node is not within k hops of any volunteer CHs



– Therefore, the sensor node becomes a forced clusterheadTherefore, the sensor node becomes a forced clusterhead

– The CH can transmit the aggregated information to the processing center The CH can transmit the aggregated information to the processing center

after every t units of time since all the sensors within a cluster are at most k after every t units of time since all the sensors within a cluster are at most k

hops away from the CHhops away from the CH

– The limit on the number of hops allows the CH to reschedule their The limit on the number of hops allows the CH to reschedule their

transmissionstransmissions

– This is a distributed algorithm and does not demand clock synchronization This is a distributed algorithm and does not demand clock synchronization

between the sensorsbetween the sensors

– The energy consumed for the information gathered by the sensors to reach The energy consumed for the information gathered by the sensors to reach

the processing center will depend on the parameters the processing center will depend on the parameters pp and and kk

– Since the objective of this work is to organize sensors in clusters to Since the objective of this work is to organize sensors in clusters to

minimize the energy consumption, values of the parameters (minimize the energy consumption, values of the parameters (pp and and kk) must ) must

be found to ensure the goalbe found to ensure the goal



Assumptions made for the optimal parameters are as follows:Assumptions made for the optimal parameters are as follows:

– The sensors are distributed as per a homogeneous spatial Poisson process The sensors are distributed as per a homogeneous spatial Poisson process

of intensity of intensity λλ in 2-dimensional space in 2-dimensional space

– All sensors transmit at the same power level – have the same radio range All sensors transmit at the same power level – have the same radio range rr

– Data exchanged between two communicating sensors not within each others’ Data exchanged between two communicating sensors not within each others’

radio range is forwarded by other sensorsradio range is forwarded by other sensors

– A distance of A distance of dd between any sensor and its CH is equivalent to hops between any sensor and its CH is equivalent to hops

– Each sensor uses 1 unit of energy to transmit or receive 1 unit of dataEach sensor uses 1 unit of energy to transmit or receive 1 unit of data

– A routing infrastructure is in place; when a sensor communicates data to A routing infrastructure is in place; when a sensor communicates data to

another sensor, only the sensors on the routing path forward the dataanother sensor, only the sensors on the routing path forward the data

– The communication environment is contention- and error-free; sensors do not The communication environment is contention- and error-free; sensors do not

have to retransmit any datahave to retransmit any data


d / r

A New, Energy-Efficient, Hierarchical Clustering AlgorithmA New, Energy-Efficient, Hierarchical Clustering Algorithm

– This algorithm is extension of the previous one by allowing more than one This algorithm is extension of the previous one by allowing more than one

level of clustering in placelevel of clustering in place

– Assume that there are Assume that there are hh levels in the clustering hierarchy with level 1 being levels in the clustering hierarchy with level 1 being

the lowest level and level the lowest level and level hh being the highest being the highest

– The sensors communicate the gathered data to level-1 clusterheads (CHs)The sensors communicate the gathered data to level-1 clusterheads (CHs)

– The level-1 CHs aggregate this data and communicate the aggregated data The level-1 CHs aggregate this data and communicate the aggregated data

to level-2 CHs and so onto level-2 CHs and so on

– Finally, level-h CHs communicate the aggregated data or estimates based on Finally, level-h CHs communicate the aggregated data or estimates based on

this aggregated data to the processing centerthis aggregated data to the processing center



– The cost of communicating the information from the sensors to the The cost of communicating the information from the sensors to the

processing center is the energy consumed by the sensors to communicate processing center is the energy consumed by the sensors to communicate

the information to level-1 CHs, plus the energy consumed by the level-1 CHs the information to level-1 CHs, plus the energy consumed by the level-1 CHs

to communicate the aggregated data to level-2 CHs, …., plus the energy to communicate the aggregated data to level-2 CHs, …., plus the energy

consumed by the level-h CHs to communicate the aggregated data to the consumed by the level-h CHs to communicate the aggregated data to the

information processing centerinformation processing center

Algorithm DetailsAlgorithm Details

– The algorithm works in a bottom-up fashionThe algorithm works in a bottom-up fashion

– First, it elects the level-1 clusterheads, then level-2 clusterheads, and so onFirst, it elects the level-1 clusterheads, then level-2 clusterheads, and so on




– Level-1 clusterheads are chosen as follows:Level-1 clusterheads are chosen as follows:

o Each sensor decides to become a level-1 CH with certain probability Each sensor decides to become a level-1 CH with certain probability pp11

and advertises itself as a clusterhead to the sensors within its radio and advertises itself as a clusterhead to the sensors within its radio

rangerange

o This advertisement is forwarded to all the sensors within This advertisement is forwarded to all the sensors within kk11 hops of the hops of the

advertising CHadvertising CH

o Each sensor receiving an advertisement joins the cluster of the closest Each sensor receiving an advertisement joins the cluster of the closest

level-1 CH; the remaining sensors become forced level-1 CHslevel-1 CH; the remaining sensors become forced level-1 CHs

– Level-1 CHs then elect themselves as level-2 CHs with a certain probability Level-1 CHs then elect themselves as level-2 CHs with a certain probability

pp2 2 and broadcast their decision of becoming a level-2 CHand broadcast their decision of becoming a level-2 CH

– This decision is forwarded to all the sensors within This decision is forwarded to all the sensors within kk22 hops hops




– The level-1 CHs that receive the advertisement from level-2 CHs joins the The level-1 CHs that receive the advertisement from level-2 CHs joins the

cluster of the closest level-2 CH; the remaining level-1 CHs become forced cluster of the closest level-2 CH; the remaining level-1 CHs become forced

level-2 CHslevel-2 CHs

– Clusterheads at level Clusterheads at level 3, 4, 5,…,h3, 4, 5,…,h are chosen in similar fashion with are chosen in similar fashion with

probabilities probabilities pp33, p, p44, p, p55,...,p,...,phh respectively to generate a hierarchy of CHs, in respectively to generate a hierarchy of CHs, in

which any level-i CH is also CH of level (i-1), (i-2),…,1.which any level-i CH is also CH of level (i-1), (i-2),…,1.


Advantages:Advantages:

– It is considered one of the earliest clustering algorithms in sensor It is considered one of the earliest clustering algorithms in sensor

networks that incorporates energy efficiency into the design of the networks that incorporates energy efficiency into the design of the

algorithmalgorithm

– Since it is distributed algorithm, there is no need for clock Since it is distributed algorithm, there is no need for clock

synchronization between sensor nodessynchronization between sensor nodes

– It achieves not only better energy efficiency, but also better time It achieves not only better energy efficiency, but also better time

complexity compared to previous workcomplexity compared to previous work

– The sensor nodes considered are simple nodes with fixed power level of The sensor nodes considered are simple nodes with fixed power level of

transmissionstransmissions

– Since the algorithm is run periodically, the probability of becoming a Since the algorithm is run periodically, the probability of becoming a

clusterhead for each period is chosen to ensure that every node will get clusterhead for each period is chosen to ensure that every node will get

a chance to become clusterhead – providing the functionality for load a chance to become clusterhead – providing the functionality for load

balancing balancing



– Another approach to ensure load balancing is to trigger the algorithm Another approach to ensure load balancing is to trigger the algorithm

when the energy levels fall below a certain thresholdwhen the energy levels fall below a certain threshold

– Energy savings increases as the density of the sensor nodes increases Energy savings increases as the density of the sensor nodes increases

for single level clusteringfor single level clustering

– For the hierarchical clustering algorithm, the energy savings increase for For the hierarchical clustering algorithm, the energy savings increase for

(i) networks of sensors with lower communication radius, (ii) lower (i) networks of sensors with lower communication radius, (ii) lower

density of sensors in the network, and (iii) increase in the number of density of sensors in the network, and (iii) increase in the number of

hierarchy levelshierarchy levels


Disadvantages:Disadvantages:

– The energy consumption of clusterheads has not been addressed since these The energy consumption of clusterheads has not been addressed since these

nodes will involve with more computation and communication of data to nodes will involve with more computation and communication of data to

higher level clusterheads – consequence of non-uniform power consumption higher level clusterheads – consequence of non-uniform power consumption

on the performance of the overall sensor network in the long runon the performance of the overall sensor network in the long run

– An ideal network is assumed (contention- and error-free) which may not An ideal network is assumed (contention- and error-free) which may not

reflect the real life scenariosreflect the real life scenarios

– Possible load imbalance between different clustersPossible load imbalance between different clusters

– Overhead associated with the clusterheads selection is not consideredOverhead associated with the clusterheads selection is not considered

– How does the network cope with sensor node failures? How is detected and How does the network cope with sensor node failures? How is detected and

remedied?remedied?

– How does the network handle information sent by faulty sensors?How does the network handle information sent by faulty sensors?


Disadvantages:Disadvantages:

– How many forced-clusterheads can the sensor network handle? What is the How many forced-clusterheads can the sensor network handle? What is the

upper bound? What are the guarantees that forced-clusterhead will be able upper bound? What are the guarantees that forced-clusterhead will be able

to communicate with the neighboring clusterheads?to communicate with the neighboring clusterheads?

– Similarly, what is the upper bound on the number of sensor nodes within Similarly, what is the upper bound on the number of sensor nodes within

one cluster?one cluster?

– Energy is wasted by those sensor nodes closer to the processing center Energy is wasted by those sensor nodes closer to the processing center

than their CH, but still need to go through their CHthan their CH, but still need to go through their CH


Suggestions/Improvements/Future Work:Suggestions/Improvements/Future Work:

– What happens if a sensor node receives several join advertisements What happens if a sensor node receives several join advertisements

from multiple nearby clusterheads? How does the sensor node decides from multiple nearby clusterheads? How does the sensor node decides

which one to join? which one to join?

Possible solution:Possible solution: the decision can be made to join to the cluster with the the decision can be made to join to the cluster with the

minimum number of members such that sensor nodes are evenly minimum number of members such that sensor nodes are evenly

distributed among the clustersdistributed among the clusters

– Error and contention in communication is not considered Error and contention in communication is not considered

Possible solution: Possible solution: results may be verified with the real MAC protocol and results may be verified with the real MAC protocol and

traffic conditions under a simulator or a test-bedtraffic conditions under a simulator or a test-bed

– The capabilities of the processing center should be more than the The capabilities of the processing center should be more than the

regular sensor nodesregular sensor nodes



– Further energy efficiency can be achieved if the clusterheads can be in Further energy efficiency can be achieved if the clusterheads can be in

active or inactive mode (energy saving mode)active or inactive mode (energy saving mode)

– Depending on the distance from the clusterheads, the sensor nodes may Depending on the distance from the clusterheads, the sensor nodes may

choose to transmit data towards clusterhead in various power levels (for choose to transmit data towards clusterhead in various power levels (for

instance, low vs. high)instance, low vs. high)

– In multi-hop mode, the sensor nodes closest to the clusterhead have the In multi-hop mode, the sensor nodes closest to the clusterhead have the

most energy drainage due to data forwardingmost energy drainage due to data forwarding

Possible solution:Possible solution: a scheme allowing the sensor nodes to alternate a scheme allowing the sensor nodes to alternate

between single-hop and multiple-hop mode periodicallybetween single-hop and multiple-hop mode periodically


– LEACHLEACH (Low-Energy Adaptive Clustering Hierarchy) is a clustering-based (Low-Energy Adaptive Clustering Hierarchy) is a clustering-based

protocol that utilizes the randomized rotation of local cluster base stations protocol that utilizes the randomized rotation of local cluster base stations

to evenly distribute the energy load within the network of sensorsto evenly distribute the energy load within the network of sensors

– It is a distributed, does not require any control information from base station It is a distributed, does not require any control information from base station

(BS) and the nodes do not need to have knowledge of global network for (BS) and the nodes do not need to have knowledge of global network for

LEACH to functionLEACH to function

– The energy saving of LEACH is achieved by combining compression with The energy saving of LEACH is achieved by combining compression with

data routingdata routing

– Key features of LEACH include:Key features of LEACH include:

Localized coordination and control of cluster set-up and operationLocalized coordination and control of cluster set-up and operation

Randomized rotation of the cluster base stations or clusterheads and their Randomized rotation of the cluster base stations or clusterheads and their

clustersclusters

Local compression of information to reduce global communicationLocal compression of information to reduce global communication

Energy-Efficient Communication Protocol Architecture for Wireless Microsensor Networks (LEACH Protocol)[Heinzelman+ 2000, 2002]

– Considered microsensor network has the following characteristics:Considered microsensor network has the following characteristics:

The base station is fixed and located far from the sensorsThe base station is fixed and located far from the sensors

All the sensor nodes are homogeneous and energy constrainedAll the sensor nodes are homogeneous and energy constrained

– Communication between sensor nodes and the base station is expensive and no Communication between sensor nodes and the base station is expensive and no

high energy nodes exist to achieve communicationhigh energy nodes exist to achieve communication

– By using clusters to transmit data to the BS, only few nodes need to transmit for By using clusters to transmit data to the BS, only few nodes need to transmit for

larger distances to the BS while other nodes in each cluster use small transmit larger distances to the BS while other nodes in each cluster use small transmit

distancesdistances

– LEACH achieves superior performance compared to classical clustering algorithms LEACH achieves superior performance compared to classical clustering algorithms

by using adaptive clustering and rotating clusterheads; assisting the total energy of by using adaptive clustering and rotating clusterheads; assisting the total energy of

the system to be distributed among all the nodesthe system to be distributed among all the nodes

– By performing load computation in each cluster, amount of data to be transmitted to By performing load computation in each cluster, amount of data to be transmitted to

BS is reduced. Therefore, large reduction in the energy dissipation is achieved BS is reduced. Therefore, large reduction in the energy dissipation is achieved

since communication is more expensive than computationsince communication is more expensive than computation

LEACH [Heinzelman+ 2000, 2002]

Algorithm OverviewAlgorithm Overview

– The nodes are grouped into local clusters with one node acting as the local base The nodes are grouped into local clusters with one node acting as the local base

station (BS) or clusterhead (CH)station (BS) or clusterhead (CH)

– The CHs are rotated in random fashion among the various sensorsThe CHs are rotated in random fashion among the various sensors

– Local data fusion is achieved to compress the data being sent from clusters to the Local data fusion is achieved to compress the data being sent from clusters to the

BS; resulting the reduction in the energy dissipation and increase in the network BS; resulting the reduction in the energy dissipation and increase in the network

lifetimelifetime

– Sensor elect themselves to be local BSs at any any given time with a certain Sensor elect themselves to be local BSs at any any given time with a certain

probability and these CHs broadcast their status to other sensor nodesprobability and these CHs broadcast their status to other sensor nodes

– Each node decided which CH to join based on the minimum communication energyEach node decided which CH to join based on the minimum communication energy

– Upon clusters formation, each CH creates a schedule for the nodes in its cluster Upon clusters formation, each CH creates a schedule for the nodes in its cluster

such that radio components of each non-clusterhead node need to be turned OFF such that radio components of each non-clusterhead node need to be turned OFF

always except during the transmit timealways except during the transmit time

– The CH aggregates all the data received from the nodes in its cluster before The CH aggregates all the data received from the nodes in its cluster before

transmitting the compressed data to BStransmitting the compressed data to BS


Algorithm OverviewAlgorithm Overview

– The transmission between CH and BS requires high energy transmissionThe transmission between CH and BS requires high energy transmission

– In order to evenly distribute energy usage among the sensor nodes, clusterheads In order to evenly distribute energy usage among the sensor nodes, clusterheads

are self-elected at different time intervalsare self-elected at different time intervals

– The nodes decides to become a CH depending on the amount of energy it has leftThe nodes decides to become a CH depending on the amount of energy it has left

– The decisions to become CH are made independently of the other nodesThe decisions to become CH are made independently of the other nodes

– The system can determine the optimal number of CHs prior to election procedure The system can determine the optimal number of CHs prior to election procedure

based on parameters such as network topology and relative costs of computation based on parameters such as network topology and relative costs of computation

vs. communication (Optimal number of CHs considered is 5% of the nodes)vs. communication (Optimal number of CHs considered is 5% of the nodes)

– It has been observed that nodes die in a random fashionIt has been observed that nodes die in a random fashion

– No communication exists between CHsNo communication exists between CHs

– Each node has same probability to become a CHEach node has same probability to become a CH



– The operation of LEACH is achieved by The operation of LEACH is achieved by roundsrounds

– Each round begins with a set-up phase (clusters are selected) followed by steady-Each round begins with a set-up phase (clusters are selected) followed by steady-

state phase (data transmission to BS occurs)state phase (data transmission to BS occurs)

1.1. Advertisement Phase: Advertisement Phase:

– Initially, each node need to decide to become a CH for the current round based Initially, each node need to decide to become a CH for the current round based

on the suggested percentage of CHs for the network (set prior to this phase) on the suggested percentage of CHs for the network (set prior to this phase)

and the number times the node has acted as a CHand the number times the node has acted as a CH

– The node (n) decides by choosing a random number between 0 and 1The node (n) decides by choosing a random number between 0 and 1

– If this random number is less than T(n), the nodes become a CH for this roundIf this random number is less than T(n), the nodes become a CH for this round

– The threshold is set as follows:The threshold is set as follows:

1P

P

1 – P * (rmod )

0 Otherwise T(n) =

If n C G P = desired percentage of CHsr = current roundG = set of nodes that have not been CHs in the last 1/P rounds



1. Advertisement Phase: 1. Advertisement Phase:

– Assumptions are (i) each node starts with the same amount of energy and (ii) Assumptions are (i) each node starts with the same amount of energy and (ii)

each CHs consumes relatively same amount of energy for each nodeeach CHs consumes relatively same amount of energy for each node

– Each node elected as CH broadcasts an advertisement message to the restEach node elected as CH broadcasts an advertisement message to the rest

– During this “clusterhead-advertisement” During this “clusterhead-advertisement” phase, the non-clusterhead nodes phase, the non-clusterhead nodes

hear the ads of all CHs and decide which CH to joinhear the ads of all CHs and decide which CH to join

– A node joins to a CH in which it hears with its advertisement with the A node joins to a CH in which it hears with its advertisement with the highest highest

signal strengthsignal strength

2. Cluster Set-Up Phase:2. Cluster Set-Up Phase:

– Each node informs its clusterhead that it will be member of the clusterEach node informs its clusterhead that it will be member of the cluster

3. Schedule Creation:3. Schedule Creation:

– Upon receiving all the join messages from its members, CH creates a TDMA Upon receiving all the join messages from its members, CH creates a TDMA

schedule about their allowed transmission time based on the total number of schedule about their allowed transmission time based on the total number of

members in the clustermembers in the cluster



4. Data Transmission: 4. Data Transmission:

– Each node starts data transmission to their CH based on their TDMA scheduleEach node starts data transmission to their CH based on their TDMA schedule

– The radio of each cluster member nodes can be turned OFF until their The radio of each cluster member nodes can be turned OFF until their

allocated transmission time comes; minimizing the energy dissipationallocated transmission time comes; minimizing the energy dissipation

– The CH nodes must keep its receiver ON to receive all the dataThe CH nodes must keep its receiver ON to receive all the data

– Once all the data is received, the CH compresses the data to send it to BSOnce all the data is received, the CH compresses the data to send it to BS

Multiple ClustersMultiple Clusters

– In order to minimize the radio interference between nearby clusters, each CH In order to minimize the radio interference between nearby clusters, each CH

chooses randomly from a list of spreading CDMA codes and it informs its chooses randomly from a list of spreading CDMA codes and it informs its

cluster members to transmit using this codecluster members to transmit using this code

– The neighboring CHs radio signals will be filtered out to avoid corruption in the The neighboring CHs radio signals will be filtered out to avoid corruption in the

transmissiontransmission



– Localized coordination to enable scalability, and robustness for dynamic Localized coordination to enable scalability, and robustness for dynamic

networksnetworks

– Incorporates data fusion into the routing protocol in order to reduce the Incorporates data fusion into the routing protocol in order to reduce the

amount of information transmitted to BSamount of information transmitted to BS

– Distributes energy dissipation evenly throughout the sensors, thus increasing Distributes energy dissipation evenly throughout the sensors, thus increasing

the system lifetime of the networkthe system lifetime of the network


DisadvantagesDisadvantages::

– How to decide the percentage of cluster heads for a network? The topology, How to decide the percentage of cluster heads for a network? The topology,

density and number of nodes of a network could be different from other networksdensity and number of nodes of a network could be different from other networks

– No suggestions about when the re-election needs to be invokedNo suggestions about when the re-election needs to be invoked

– The clusterheads farther away from the base station will use higher power and The clusterheads farther away from the base station will use higher power and

die more quickly than the nearby onesdie more quickly than the nearby ones



– Extensions can be included to have hierarchical clustering where each CH Extensions can be included to have hierarchical clustering where each CH

will communicate with “super-clusterhead” until the top layer of hierarchy in will communicate with “super-clusterhead” until the top layer of hierarchy in

which the data needs to be sent to BSwhich the data needs to be sent to BS

– The degree and remaining energy of a node may be considered as The degree and remaining energy of a node may be considered as

parameters to decide a clusterhead in a round. If a clusterhead with a limited parameters to decide a clusterhead in a round. If a clusterhead with a limited

power used up its power in a round, the data to be transmitting may be lostpower used up its power in a round, the data to be transmitting may be lost

– Since TDMA schedule is used, a large delay may be introduced between Since TDMA schedule is used, a large delay may be introduced between

event detection and notification at base station. Therefore, the protocol is not event detection and notification at base station. Therefore, the protocol is not

suitable for a real-time applicationsuitable for a real-time application


Computes the degree of a node based on the distance Computes the degree of a node based on the distance (transmission range) between the node and the other nodes(transmission range) between the node and the other nodes

The node with the maximum number of neighbors (maximum The node with the maximum number of neighbors (maximum degree) is chosen to be a clusterhead and any tie is broken by the degree) is chosen to be a clusterhead and any tie is broken by the node idsnode ids

Drawbacks: A clusterhead cannot handle a large number of nodes due to A clusterhead cannot handle a large number of nodes due to

resource limitationsresource limitations Load handling capacity of the clusterhead puts an upper bound on Load handling capacity of the clusterhead puts an upper bound on

the node-degreethe node-degree The throughput of the system drops as the number of nodes in The throughput of the system drops as the number of nodes in

cluster increasescluster increases

Related Work

Highest-Degree Heuristic [Gerla+ 1995, Parekh 1994]

The node with the minimum node-id is chosen to be a clusterheadThe node with the minimum node-id is chosen to be a clusterhead A node is called a A node is called a gateway gateway if it lies within the transmission range of two if it lies within the transmission range of two

or more clustersor more clusters Distributed gateway Distributed gateway is a pair of nodes that reside within different is a pair of nodes that reside within different

clusters, but they are within the transmission range of each otherclusters, but they are within the transmission range of each other

Drawbacks: Since it is biased towards nodes with smaller node-ids, leading to Since it is biased towards nodes with smaller node-ids, leading to

battery drainage battery drainage It does not attempt balance the load for across all the nodesIt does not attempt balance the load for across all the nodes

Related Work

Lowest-ID Heuristic [Baker+ 1981a, 1981b, Ephremides+ 1987]

Node-weightsNode-weights are assigned to nodes based on the suitability of a are assigned to nodes based on the suitability of a node being a clusterheadnode being a clusterhead

The node is chosen to be a clusterhead if its node-weight is higher The node is chosen to be a clusterhead if its node-weight is higher than any of its neighbor’s node-weights and any tie is broken by the than any of its neighbor’s node-weights and any tie is broken by the minimum node idsminimum node ids

Drawbacks: No concrete criteria of assigning the node-weightsNo concrete criteria of assigning the node-weights Works well for “quasi-static” networks where the nodes do not move Works well for “quasi-static” networks where the nodes do not move

much or move very slowlymuch or move very slowly

Related Work

Node-Weight Heuristic [Basagni 1999a, 1999b]

AA clusterhead can clusterhead can ideallyideally support nodes support nodes– Ensures efficient MAC functioningEnsures efficient MAC functioning– Minimizes delay and maximizes throughputMinimizes delay and maximizes throughput

A clusterhead uses more battery power A clusterhead uses more battery power – Does extra work due to packet forwardingDoes extra work due to packet forwarding– Communicates with more number of nodesCommunicates with more number of nodes

A clusterhead should be less mobileA clusterhead should be less mobile– Helps to maintain same configuration Helps to maintain same configuration – Avoids frequent WCA invocationAvoids frequent WCA invocation

A better power usage with physically closer nodesA better power usage with physically closer nodes– More power for distant nodes due to signal attenuationMore power for distant nodes due to signal attenuation

Related Work

Weighted Clustering Algorithm (WCA) [Chatterjee+ 2002]

WCA Steps

1.1. Compute the Compute the degreedegree ddvv each node each node vv

Coordinate distance, predefined transmission range. Coordinate distance, predefined transmission range.

2.2. Compute the Compute the degree-differencedegree-difference for every nodefor every node

For efficient MAC (medium access control) functioning.For efficient MAC (medium access control) functioning.

Upper bound on # of nodes a cluster head can handleUpper bound on # of nodes a cluster head can handle..

vvv

txvvdistvNvV

ranged''

,

',|)(|

|| dvv

Related Work


3.3. Compute the Compute the sum of the distancessum of the distances DDvv with all neighbors with all neighbors

Energy consumption; more energy for greater dist.Energy consumption; more energy for greater dist.

communication.communication.

Power required to support a link increases faster thanPower required to support a link increases faster than

linearly with distance.linearly with distance. (For cellular networks)(For cellular networks)

)('

',vNv

vvdistvD 1

2 3

4

5

6

7

12 1314

1516

17

WCA Steps

Related Work


4.4. Compute the average speed of every node; gives a measure ofCompute the average speed of every node; gives a measure of

mobilitymobility MMvv

where where and and are the are the

coordinates of the node at time and coordinates of the node at time and

Component with less mobility is a better choice for clusterheadComponent with less mobility is a better choice for clusterhead..

T

tYYXXM ttttTv

111

1 22

YX tt, YX tt 11,

v t 1t

Yt

Yt-1

XtXt-1

time

WCA Steps

Related Work


5.5. Compute the total (cumulative) Compute the total (cumulative) timetime PPvv a node acts as clusterhead a node acts as clusterhead

Battery drainage = Power consumedBattery drainage = Power consumed

6.6. Calculate the Calculate the combined weightcombined weight WWvv for each node for each node

WWvv = w = w11ΔΔv v + w + w22DDv v + w + w33MMv v + w + w44PPvv for each nodefor each node

7.7. Find min Find min WWvv;; choose node choose node vv as the cluster head, remove all as the cluster head, remove all

neighbors of neighbors of vv for further WCA for further WCA

8.8. Repeat steps 2 to 7 for the remaining nodesRepeat steps 2 to 7 for the remaining nodes

WCA Steps

Related Work


It is desirable to balance the loads among the clustersIt is desirable to balance the loads among the clusters

Load balancing factor (LBF) has defined as (should be high)Load balancing factor (LBF) has defined as (should be high)

i i

cLBF

xn

2

where,where,

nc is the number of clusterheads is the number of clusterheads

xi is the cardinality of cluster is the cardinality of cluster ii and and

nc

ncN is the average number of neighbors of a clusterheadis the average number of neighbors of a clusterhead

((N N being the total number of nodes in the system)being the total number of nodes in the system)

Load Balancing Factor (LBF)

Related Work


For clusters to communicate with each other, it is assumed that For clusters to communicate with each other, it is assumed that

clusterheads are capable of operating in clusterheads are capable of operating in dual dual power mode power mode

A clusterhead uses A clusterhead uses lowlow power mode to communicate with its immediate power mode to communicate with its immediate

neighbors within its transmission range and neighbors within its transmission range and highhigh power mode is used for power mode is used for

communication with neighboring clusterscommunication with neighboring clusters

ConnectivityConnectivity is defined as (for multiple component graph) is defined as (for multiple component graph)

Probability that a node is reachable from any other nodeProbability that a node is reachable from any other node

N

componentlargestofsizetyconnectivi

Connectivity

Related Work


[Baker+ 1981a] D.J. Baker and A. Ephremides, A Distributed Algorithm for Organizing Mobile Radio Telecommunication Networks, Proceedings of the 2nd International Conference on Distributed Computer Systems, April 1981, pp. 476-483.

[Baker+ 1981b] D.J. Baker and A. Ephremides, The Architectural Organization of a Mobile Radio Network via a Distributed Algorithm, IEEE Transactions on Communications COM-29(11), 1981, pp. 1694-1701.

[Bandyopadhyay+ 2003] S. Bandyopadhyay and E.J. Coyle, An Energy Efficient Hierarchical Clustering Algorithm for Wireless Sensor Networks, IEEE INFOCOM 2003, San Francisco, CA, March 30 – April 3, 2003.

[Basagni 1999a] S. Basagni, Distributed Clustering for Ad hoc Networks, Proceedings of International Symposium on Parallel Architectures, Algorithms and Networks, June 1999, pp. 310-315.

[Basagni 1999b] S. Basagni, Distributive and Mobility-Adaptive Clustering for Multimedia Support in Multi-hop Wireless Networks, Proceedings of Vehicular Technology Conference, VTC, Vol. 2, 1999-Fall, pp. 889-893.

[Chatterjee+ 2002] M. Chatterjee, S. K. Das and D. Turgut, WCA: A Weighted Clustering Algorithm for Mobile Ad hoc Networks, Journal of Cluster Computing (Special Issue on Mobile Ad hoc Networks), Vol. 5, No. 2, April 2002, pp. 193-204.

[Ephremides+ 1987] A. Ephremides J.E. Wieselthier and D.J. Baker, A Design Concept for Reliable Mobile Radio Networks with Frequency Hopping Signaling, Proceedings of IEEE, Vol. 75(1), 1987, pp. 56-73.

References

[Gerla+ 1995] M. Gerla and J.T. Tsai, Multicluster, mobile, multimedia radio network, Wireless Networks, Vol. 1, No. 3, 1995, pp. 255-265.

[Heinzelman+ 2002] W. Heinzelman, A.P. Chandrakasan and H. Balakrishnan, An Application-Specific Protocol Architecture for Wireless Microsensor Networks, IEEE Transactions on Wireless Communications, Vol. 1, No. 4, October 2002, pp. 660-670.

[Heinzelman+ 2000] W. Heinzelman, A.P. Chandrakasan and H. Balakrishnan, Energy-Efficient Communication Protocol for Wireless Microsensor Networks, IEEE Proceedings of the Hawaii International Conference on System Sciences, January 4-7, 2000, Maui, Hawaii.

[Parekh 1994] A.K. Parekh, Selecting Routers in Ad-hoc Wireless Networks, Proceedings of the SBT/IEEE International Telecommunications Symposium, August 1994.

References

clustering in sensor networks

Documents

sensor data

sensor networksthe data

energy savings

energy consumption

data processing center

single processing center

sensors itselfin

dataclustering sensors