a modified energy efficient communication protocol for wireless sensor network
TRANSCRIPT
8/20/2019 A Modified Energy Efficient Communication Protocol for Wireless Sensor Network
http://slidepdf.com/reader/full/a-modified-energy-efficient-communication-protocol-for-wireless-sensor-network 1/7
International Journal of Scientific Research Engineering & Technology (IJSRET)Volume 2 Issue 11 pp 717-723 February 2014 www.ijsret.org ISSN 2278 – 0882
IJSRET @ 2014
A Modified Energy Efficient Communication Protocol for Wireless
Sensor Network
Aman Deep SinghM. Tech (Computer Science & Engineering)
B.S.A College of Engineering and Technology Mathura (U.P) INDIA
ABSTRACTWireless distributed micro sensor systems will enable
the reliable monitoring of a variety of environments for
both civil and military applications. . Clustering sensor
nodes is an effective technique for achieving these goals.
LEACH (Low Energy Adaptive Clustering Hierarchy) is
one of popular cluster-Based structures, which has been
widely proposed in wireless Sensor networks. LEACHuses localized coordination to enable the scalability and
robustness for dynamic networks. LEACH utilizes the
randomization of local cluster based stations to evenly
distribute the load among the sensors in the network.
LEACH uses a TDMA based MAC protocol the
proposed protocol is simulated and the results show a
significant reduction in network energy consumption.
Keywords: WSN (Wireless Sensor Networks),
hierarchical clustering, LEACH Algorithm, network
lifetime.
I. INTRODUCTIONA wireless sensor network contains hundreds or
thousands of these sensor devices that have ability to
communicate either directly to the Base Station (BS) or
among each other. LEACH is one of the protocols thatgive the guarantee about the allocation of energy in the
sensor node homogeneously. A homogeneous sensor
networks can be defined as a network consisting of
identical nodes in terms of energy, processing
capabilities, and sensing range. Micro sensor network consist of many spatially distributed sensors, which are
used to monitor various kinds of ambient conditions liketemperature, humidity, etc and then transform them into
electric signal. A sensor is equipped with a radio
transceiver, a small microcontroller, and an energysource usually a battery. . The nodes in WSNs are
usually battery operated sensing devices with limited
energy resources and replacing or replenishing the
batteries is usually not an option. Usually, sensor nodes
are scattered in the sensing field.
Routing in WSNs is very challenging due to the specific
characteristics that distinguish WSNs from other
wireless networks. Network protocols must be designed
to achieve fault tolerance while minimizing energyconsumption .In addition since the limited wireless
channel bandwidth must be shared among the sensors in
the network, routing protocols should be able to performlocal collaboration to reduce channel bandwidth
requirements. Based on the network structure adopted
routing protocols for WSNs can be classified into fla
network routing location-based network routing. In fla
network routing, all nodes have the same functionality
and they work together to perform sensing and routing
tasks.
The Sensor Protocols for Information via egotiation
(SPIN) [4] and Directed Diffusion [5] fall into this
category. Hierarchical network routing divides the
network into clusters to achieve energy-efficient
scalability and one of the famous hierarchical network
routing protocol is low-energy adaptive clustering
hierarchy (LEACH) [1]. In location-based network
routing, location information of nodes is used to
compute the routing path. This information can be
obtained from global positioning system (GPS) devices
attached to each sensor node. Examples of location-
based network routing protocols include geography
adaptive routing (GAF) [1] and Geographic and Energy-
Aware Routing (GEAR) [6].
Sensor networks contain too much data for an end-user
to process. Therefore, automated methods of combining
or aggregating the data into a small set of meaningful
information is required [7, 8]. In addition to helping
avoid information overload, data aggregation, also
known as data fusion, can combine several unreliable
data measurements to, being the area where we want to
monitor some ambient Conditions. The information
collected by sensor nodes is routed to the Base Station
either directly or through other sensor nodes.
During the creation of network topology, the process of
setting up routes in WSNs is usually influenced by
8/20/2019 A Modified Energy Efficient Communication Protocol for Wireless Sensor Network
http://slidepdf.com/reader/full/a-modified-energy-efficient-communication-protocol-for-wireless-sensor-network 2/7
International Journal of Scientific Research Engineering & Technology (IJSRET)Volume 2 Issue 11 pp 717-723 February 2014 www.ijsret.org ISSN 2278 – 0882
IJSRET @ 2014
energy considerations. The classification performed onthe aggregated data might be performed by a human
operator or automatically. Both the method of per-
forming data aggregation and the classification algorithm
are application-specific.Sensor nodes are randomly scattered so multi-hop
routing is unquestionably defacto. Many research
projects and papers have shown that the hierarchical
network routing and specially the clustering mechanisms
make significant improvement in WSNs in reducing
energy consumption and overhead [7, 8] also have to
note that most of clustering protocols proposed for
WSNs assume that nodes are stationary. The reason for
sensor nodes to be taken as stationary is the assumption
of simple network topology. Clustering protocols can
reduces signaling overhead since they do not have to
manage the mobility pattern or location information of sensor nodes. As a result, it allows nodes saving more
energy leading to a longer network life time. However,
with some applications such as animal tracking, search
and rescue activities this assumption is not very realistic;
hence there are raising demands for clustering protocols
to support mobile nodes.
Clustering network is efficient and scalable way to
organize WSNs. A cluster head responsible for
Conveying any information gathered by the nodes in its
cluster and may aggregate and compress the data before
transmitting it to the sink. However, this added
responsibility results in a higher rate of energy drain atthe cluster heads. One of the most popular clustering
mechanisms, LEACH, addresses this by robabilistically
rotating the role of cluster head among all nodes.
However, unless each node selects its probability of
becoming a cluster head wisely, the performance of the
network may be far from optimal. The main focus of this
paper is modifying LEACH clustering algorithm. This
algorithm fully utilizes the location information of
network nodes in routing to reduce the routing cost.
By analyzing the advantages and disadvantages of
conventional routing protocols using our model of
sensor networks, we have developed LEACH (Low-
Energy Adaptiv Clustering Hierarchy), a clustering-
based protocol that minimizes energy dissipation in
sensor networks. The key features of LEACH are:
• Localized coordination and control for cluster
set-up and operation
• Randomized rotation of the cluster “base
stations” or “cluster -heads” and thecorresponding clusters.
• Local compression to reduce global
communication
The use of clusters for transmitting data to the base
station leverages the advantages of small transmidistances for most nodes, requiring only a few nodes to
transmit far distances to the base station. However
LEACH out-performs classical clustering algorithms by
using adaptive clusters and rotating cluster-heads
allowing the energy requirements of the system to be
distributed among all the sensors. In addition, LEACH is
able to perform local computation in each cluster to
reduce the amount of data that
II. RELATED WORKLEACH is a kind of cluster-based routing protocols
which uses distributed cluster formation. LEACHrandomly selects a few sensor nodes as cluster heads
(CHs) and rotates this role to evenly distribute the
energy load among the sensors in the network. The ideais to form clusters of the sensor nodes based on the
received signal strength and use local cluster heads as
routers to the sink. In LEACH, the Cluster Heads
compress data arriving from member nodes and send an
aggregated packet to the BS in order to reduce theamount of information that must be transmitted to the
BS. In order to reduce inter & intra cluster interference.
Clustering is the method by which sensor nodes in a
network organize themselves into hierarchical structuresBy doing this, sensor nodes can use the scarce networkresources such as radio resource, battery power more
efficiently. Within a particular cluster, data aggregation
and fusion are performed at cluster- head to reduce the
amount of data transmitting to the base station. Cluster
formation is usually based on remaining energy of
sensor nodes and sensor’s proximity to cluster -head [1]
Non cluster-head nodes choose their cluster- head righ
after deployment and transmit data to the cluster- head
The role of cluster-head is to forward these data and its
own data to the base station after performing data
aggregation and fusion. LEACH is one of the firsthierarchical routing protocols for WSNs.
III. LEACH AND LEACH-CLEACH (Low-Energy Adaptive Clustering Hierarchy)an energy-conserving routing protocol for wireless
sensor network.
8/20/2019 A Modified Energy Efficient Communication Protocol for Wireless Sensor Network
http://slidepdf.com/reader/full/a-modified-energy-efficient-communication-protocol-for-wireless-sensor-network 3/7
International Journal ofVolume 2 Issue 11 pp 717-
Figure 1: Structure of clustured W
The idea is to form cluster of sensor no
signal strength and use the cluster-head a
forward data of other nodes in cluster to th
The data processing is performed at
LEACH is a dynamic clustering mechani
divided in rounds/intervals with equal le
beginning of the round, cluster-headsrandomly among the nodes which hav
energy higher than the average remaining
the nodes.
LEACH is done into two steps, the setup
steady state phase. In setup phase th
organized into clusters and CHs are sel
cluster heads change randomly over tim
balance the energy of the network. This
choosing a random number between 0 and 1
selected as a cluster head for the currentrandom number is less than the threshold
which is given by
In this formula, p is the percentage of clust
all nodes in the network, i.e., the probabilit
is selected as a cluster head; r the number
selection; and G is the set of nodes that ar
in round 1/p. As we can see here, the select
heads is totally randomly.
After becoming cluster heads, the nod
messages to all nodes to inform the status oCluster head nodes decide which cluster
based on the receiving signal strength of th
The cluster-heads create schedules and se
nodes in the clusters. For the rest of the rou
send data to their respective cluster head n
cluster- heads aggregate and send the dat
station.
After each round, clusters-heads are re-gen
cientific Research Engineering & Technolo23 February 2014 www.ijsret.org ISSN
IJSRET @ 2014
SN
es based on
s a router to
base station.
luster-heads.
sm. Time is
ngth. At the
is generated e remaining
energy of all
hase and the
nodes are
ected. These
in order to
is done by
. The node is
round if the value T (n),
r heads over
y that a node
of rounds of
not selected
ion of cluster
es broadcast
them. Non head to join
se messages.
nd to all the
nd, the nodes
des, then the
to the base
rated to form
new clusters. The cluster-hea
to spend energy equally betwe
it can lengthen the sensor net
LEACH-centralized (LEAC
LEACH in operation exceLEACH-C, the cluster head
BS. During the setup phase
nodes information about th
remaining energy levels. BS
level to determine the candi
node. The average node ene
node has remaining energy fa
be removed from the candidat
Using the candidate set, BS
simulated annealing algorithm
problem of finding k opti
algorithm attempts to minimicluster head nodes are deter
nodes the information inclu
clusters member node and
each cluster. Nodes use this i
TDMA slot for data trans
randomized rotation of the
position such that it rotates a
order to not drain the batte
addition, LEACH perform
“compress” the amount ofclusters to the base station,
dissipation and enhancing systSensors elect themselves to b
given time with a certain pro
nodes broadcast their status t
network. Each sensor node de
wants to belong by choosi
requires the minimum commu
the nodes are organized into
creates a schedule for the
allows the radio components
node to be turned off at al
transmit time, thus minimizinthe individual sensors. Once t
data from the nodes in its clu
aggregates the data and then
data to the base station. Since
in the scenario we are examitransmission. However, sinc
cluster-heads, this only affects
LEACH can achieve over a
energy dissipation compared
with the base station, when u
gy (IJSRET) 278 – 0882
d rotation allows network
en sensor nodes and hence
ork life time.
-C) [1] is similar to
pt cluster formation. In selection is carried out a
BS receives from other
ir current locations and
uses the remaining energy
date set for cluster head
rgy is computed and the
lling below this value wil
set.
finds clusters using the
[11] to solve the NP-hard
mal clusters [12]. This
e the total energy that non ined, BS broadcast to al
ding cluster head nodes
ransmission schedule for
formation to determine its
ission LEACH includes
high-energy cluster-head
ong the various sensors in
ry of a single sensor. In
local data fusion to
ata being sent from the further reducing energy
em lifetime. local cluster-heads at any
ability. These cluster-head
o the other sensors in the
ermines to which cluster i
ng the cluster-head tha
nication energy2. Once al
clusters, each cluster-head
odes in its cluster. This
of each non-cluster-head
l times except during its
g the energy dissipated in he cluster-head has all the
ster, the cluster-head node
transmits the compressed
the base station is far away
ing, this is a high energy e there are only a few
a small number of nodes
factor of 7 reduction in
to direct communication
ing the optimal number of
8/20/2019 A Modified Energy Efficient Communication Protocol for Wireless Sensor Network
http://slidepdf.com/reader/full/a-modified-energy-efficient-communication-protocol-for-wireless-sensor-network 4/7
International Journal ofVolume 2 Issue 11 pp 717-
cluster-heads. The main energy savings of
protocol is due to combining lossy compre
data routing. There is clearly a trade-off
quality of the output and the amount of
achieved. In this case, some data from tsignals is lost, but this results in a substan
of the overall energy dissipation of the syst
IV. DISADVANTAGES OF LEAIn using LEACH protocol for cluster orga
features are still not supported. LEAC
homogeneous distribution of sensor nodes
area. This scenario is not very realistic. Let
scenario in which most of the sensor node
together around one or two cluster-heacluster formation algorithm will end up
more cluster member nodes A. This couldhead nodes a quickly running out of energy.
In addition, cluster heads are randomly
impossible the scenario illustrated in Figur
which two or even more cluster heads are
each other
Figure 2: Multiple cluster head in smal
In Figure 3, H1 and H2 is two cluster
▲and are their cluster members, respectiH2 are very closely located. Accord
communication model, the energy that a
consumes is the sum of that consumed in r
and that in sending data.
E ch = L Ebit N mem + L E (N mem +
+Lmd to Bs (1)
where L is the length of data, m the power
of transferring l bit of data, E bit the power
of processing 1 bit of data, Nmem th
members in a cluster, d to Bs the distance
cluster head and node Sink, LEelec Nme
that Nmem cluster members consume w
them send out length of l data to the clus
LENmem the power that the cluster he
cientific Research Engineering & Technolo23 February 2014 www.ijsret.org ISSN
IJSRET @ 2014
the LEACH
sion with the
between the
compression
e individual tial reduction
m.
H nization, few
assumes a
in the given
us consider a
are grouped
s LEACH’s by assigning
make cluster
selected, it
3 occurs, in
very close to
l region
heads, nodes
ely. H1 and ing to data
cluster head
eceiving data
1) + L E bit
consumption
consumption
number of
between the
m the power
hen each of
ter head, and
ad consumes
when it receives data of l
members. It follows from (1)
that cluster heads H1 and
transfer is:
Where Nmem1 and Nmem2 t
clusters H1 and H2, dh1toBs
between the two cluster heads
the total energy consumed by
E h1 + E h2 = L E bit (N m
mem1 + N mem2 + 2) + 2L
h2toBS )
When H1 and H2 are very clo
d h1toBS = d h2toBS
Then (4) becomes
E h1 + E h2 = LE bit (N memem1 + N mem2 + 2) + 2LE
As we can see, in this case th
of two clusters is only LE bit
the case that there is only onbecause LE bit + Lmd h1t
therefore, the total energy cotwo cluster heads is approxi
there is only one cluster head.
It is clear now that whenrandomly selected within a sm
loss occurs. The amount of lo
proportional to the number o
Of course, there is a precondit
is, cluster heads are very closebetween them becomes neglig
V. PROTOCOL PER
1. All senor nodes are identisame amount of initial eneenergy at the same rate an
residual energy and control
distance. Every node has t
different MAC protocol a
communication channels arconsumption of transferring d
is the same as that of transf
gy (IJSRET) 278 – 0882
ength l from its cluster
that the amount of energy
H2 consume during data
he number of members in
and dh2to Bs the distance
and node Sink, Therefore
he two clusters is
m1 + N mem2 ) +LE (N
bit + L m (d h1toBS + d
e, we can have
m1 + N mem2 ) + LE (N it + 2Lm d h1toBS
total energy consumption
+ Lmd h1tobs greater than
cluster head. In addition bs is much greater than
sumption when there are ately twice of that when
ultiple cluster heads are all area, a big extra energy
st energy is approximately
cluster heads in the area
ion on this conclusion, tha
ly located and the distance ible.
ORMANCE
cal and charged with the rgy. All nodes consume
are able to know thei
transmission power and
he capability to suppor
d data processing. Al
identical. The energy ta from node A to node B
rring the same amount of
8/20/2019 A Modified Energy Efficient Communication Protocol for Wireless Sensor Network
http://slidepdf.com/reader/full/a-modified-energy-efficient-communication-protocol-for-wireless-sensor-network 5/7
International Journal ofVolume 2 Issue 11 pp 717-
date from node B to node A.
2. Every node can directly communicat
other node
4. Every node has data to transfer in ever
Thedata transferred by sobering nodes arcan be fused.
5. Sensor nods are static.
WSNs are autonomous networks. Sens
independent with each other. The coordin
nodesis done through wireless communi
costs much. This is one of the major rea
LEACH protocol selects cluster heads rand
discussed before, this approach may cause
energy because of unbalanced cluster head
To solve this problem, we propose a new
selecting cluster heads. We assume that:1. The network satisfies the pre-conditions
LEACH protocol.
2. After deployment, sensors are able t
positions through GPS, or before depl
positions are accurately decided.
3. All nodes are able to adjust data transm
If necessary they can communicate with the
to acquire the initial setting information of t
If we modify the procedure of the calcula
during the cluster head generation such that
are produced progressively, then a node c
it is suitable to be a new cluster headlocations of existing cluster heads and its
More specifically, if the node is very
existing cluster head, then this node will
attempt to be a cluster head.
As shown in Figure the network is divid
parts. Nodes in region G1 will compete
cluster head. When a node is selected as a c
will broadcast the information to nodes nea
region G2 will receive the message. Thus,
this region compete for being cluster head
information of the cluster head in regioninto consideration. If a node in G2 is close
head in G1, the node will be discarded. The
in all other regions will be generated in the
The cluster heads generated with this appr
be close to each other. However, becausequit the competition for cluster head, the to
cluster heads can be reduced, which is
saving the network energy. Our approach t
problem is when a node is excluded in the
selection, a message is broadcast to other
cientific Research Engineering & Technolo23 February 2014 www.ijsret.org ISSN
IJSRET @ 2014
with every
time frame.
related and
r nodes are
tion between
ation, which
sons that the
omly. As we
the waste of
distribution.
approach to
f applying
know their
yment, their
ission power.
base stations
he network.
tion of T (n)
cluster heads
uld decide if
ased on the wn location.
lose to any
give up the
ed into three
for being a
luster head, it
rby. Nodes in
hen nodes in
, the location
1 will taken to the cluster
cluster heads
ame way
ach will not
some nodes tal number of
ot good for
solving this
cluster head
nodes and T
(n) will be modified to increa
nodes being selected as cluste
is
K is the number of nodes t
cluster head selection due to t
initial value of 0. When k in
well, which will ensure suf
heads will be generated by the
Figure 3: Selectin
To facilitate the explanation
we introduce the following no
Bs The base station or node Si
Si The i-th sensor node
Hj The j-th cluster head
Mem (Cj) Members of the j-th
Mem (Cj)I The i-th membersLoc (Si) Location of the i-th s
Delay (Si) Time delay that the
start to compete for a cluster h
Num(Giveup) Number of disc
|| Operation of concatenation
To evaluate the performance
100 X 100 network configurat
each sensor node is assigned
the amount of transmission enamplifier energy (E amp ) is 1
performance evaluation are
energy consumption and data
of nodes alive. Each performa
varying the number of cluster-To determining quality of dat
amount of data (number of d
gy (IJSRET) 278 – 0882
e the probability of others
heads .The modified T(n)
at are excluded from the
e location reason, with an
reases, T (n) increases as
ficient number of cluster
progressive algorithm.
cluster heads
f our improved algorithm
ations:
nk
cluster
f the j-th cluster nsor node
i-th sensor node
ead
arded cluster heads
f LEACH, we consider a
ion with 101 nodes, where
an initial energy of 2.0 J
rgy is 50 nJ /bit , transmi 00 pJ /bit. The criteria for
the network lifetime, the
aggregated at BS and no
ce criteria is evaluated by
heads from 1 to 8. a we need to measure the
ata signals represented by
8/20/2019 A Modified Energy Efficient Communication Protocol for Wireless Sensor Network
http://slidepdf.com/reader/full/a-modified-energy-efficient-communication-protocol-for-wireless-sensor-network 6/7
International Journal of Scientific Research Engineering & Technology (IJSRET)Volume 2 Issue 11 pp 717-723 February 2014 www.ijsret.org ISSN 2278 – 0882
IJSRET @ 2014
an aggregate signal) received at the base station. We
tracked the rate at which the data are transferred to the
base station and the amount of energy required to get the
data to the base station.
VI. CLUSTER HEAD SELECTION:After the deployment of sensor nodes, we first acquire
all nodes’ location information (through GPS technologyor known prior to its deployment) and report it to the
base station. The base station decides Delay (Si) for
every node based on the geographic distribution of all
sensor nodes. Delay (Si) = 0 for those in the region to
start first. As illustrated in Figure 4, nodes in G1 start to
compete for cluster heads at time 0, then nodes in G2
start with a delay, and then nodes in G3 start with a
delay after nodes in G2 are finished, and so on. During
the process, nodes need to send their locationinformation to the base station:
SRiR BS: Loc (SRiR)
The base station needs to send the delay information to
each node:
Bs SRiR: Delay (SRiR)
Set Num (Giveup) to 0. Start with the nodes in G1. If a
cluster head is generated from G1, broadcast a Hello
package and Num (Giveup).
Hj broadcast: Hello, Num (Giveup)
When nodes in G1 are finished, consider nodes in G2.
Now the cluster heads generated in G1 are referencepoints. The distance between a node in G2 and any
cluster head in G1 is a factor in selecting the node as a
cluster head, as well as the random value of T (n). If all
conditions are satisfied, then broadcast the Hello
message and Num (Giveup).Hj -> broadcast: Giveup, Num (Giveup)
Otherwise, only broadcast Num (Giveup). When nodes
in other region receive this message, they will increment
Num (Giveup) by 1, and then modify T (n) to increasethe probability of being selected as cluster head. Repeat
the above process until all nodes in the network areconsidered.
VII. SIMULATIONSIn this section, we evaluate the performance of My
LEACH protocol implemented with MATLAB 100
sensor nodes are randomly distributed in an area of 100
m x 100 m. BS is put at the location with x = 175, y =
50.The bandwidth of data channel is set to 1 Mbps, the
length of data messages is 500 bytes and packet header
for each type of packet was 25 bytes. The number round
is set to 500s. When a node uses energy down to its
energy threshold, it can no longer send data and is
considered as a dead node.
REFERENCES1. Lu, Ye Ming and Vincent W. S. Wong, “An energyefficient multipath routing protocol for wireless sensor
networks” research articles. Int. J.Commun. Syst.
20(7):747--766, 2007.
2. W. Heinzelman , J. Kulik, H. Balakrishnan, “Adaptiveprotocols for information dissemination in wireless
8/20/2019 A Modified Energy Efficient Communication Protocol for Wireless Sensor Network
http://slidepdf.com/reader/full/a-modified-energy-efficient-communication-protocol-for-wireless-sensor-network 7/7
International Journal of Scientific Research Engineering & Technology (IJSRET)Volume 2 Issue 11 pp 717-723 February 2014 www.ijsret.org ISSN 2278 – 0882
IJSRET @ 2014
sensor networks” Proceedings of ACM/IEEE
MobiCom’99, Seattle, WA, U.S.A., August 1999.
3. W. B. Heinzelman, A. Chandrakasan and H.
Balakrishnan, “Energy-Efficient CommunicationProtocol for Wireless Micro sensor Networks,” Proc.
33rd Hawaii Int’l. Conf. Sys. Sci., Jan. 2000.
4. M. Bani Yassein, A. Al-zou'bi, Y. Khamayseh, W.
Mardini “Improvement on LEACH Protocol of WirelessSensor Network (VLEACH)”, International Journal of Digital Content Technology and its Applications
Volume 3, Number 2, June 2009.
5. Sabarish B A, Guru Moorthy S M, Dhivya M A,
Sakthi Naveen K, Vaishnavi S “A survey on clustering
protocols in Wireless Sensor Networks” InternationalJournal Of Advances In Computing and Information
Technology, published on April 2012
6. M. J. Handy, M. Haase, D. Timmermann, \Low
Energy Adaptive Clustering Hierarchy with
Deterministic Cluster-Head Selection", IEEE MWCN,
2002.
6. W. B. Heinzelman “An Application-Specific Protocol
for Wireless Micro sensor Networks”, IEEETRANSACTIONSON WIRELESS
COMMUNICATIONS, OCT 2002.
7..Zhang, Yu-quan; Wei, Lei, “Improving the LEACH protocol for wireless sensor networks”, Wireless Sensor Network, 2010. IET-WSN. IET International Conference
8.T. Murata and H. Ishibuchi, “Performance evaluation
of genetic algorithms for flowshop scheduling problems”Proc. 1st IEEE Conf. Evolutionary Computation, June
1994.
9. P. Agarwal and C. Procopiuc, “Exact and
approximation algorithms for clustering,” in Proc. 9th
Annual. ACM-SIAM Symp. Discrete Algorithms,
Baltimore, MD, Jan. 1999
10. M.Joa- Ng, I.T.Lu, “A Peer -to-peer Zone-based Two-
level link state routing for mobile Ad Hoc Networks”,IEEE Journal on Selected Areas in Communications,
Special Issue on Ad-hoc Networks, Aug1999.