Content uploaded by Ramanaiah OBV
Author content
All content in this area was uploaded by Ramanaiah OBV on Mar 01, 2023
Content may be subject to copyright.
978-1-4673-7758-4/15/$31.00 ©2015 IEEE
YA-LEACH: Yet Another LEACH for Wireless
Sensor Networks
W.T. Gwavava
CSE Department
JNTUH College of Engineering
Hyderabad, India
wellytg@gmail.com
O.B.V. Ramanaiah
CSE Department
JNTUH College of Engineering
Hyderabad, India
obvramanaiah@gmail.com
Abstract— Wireless Sensor Networks (WSN) is an ever growing
field owing to their extensive variety of applications. WSN have
limited battery energy, consequently, an energy efficient routing
protocol is of major concern. LEACH (Low Energy Adaptive
Clustering Hierarchy) is one of the fundamental clustering
protocols in WSN that provides high energy efficiency. Many
protocols have been derived from LEACH which mostly address
a single inherent challenge. Instead of focusing on a single
parameter, combining independent techniques can compound
gains in energy efficiency. This paper proposes Yet Another
LEACH (YA-LEACH) which uses centralised cluster formation to
ensure optimal clusters and allow Cluster Heads (CH) to extend
operation into multiple rounds to achieve energy savings. Also,
the proposed protocol will have an alternative (vice) CH that
takes over when CH residual energy reaches a critically low
level. The extended protocol was simulated on NS2 with results
showing an improvement in network lifetime and overall data
throughput.
Keywords— LEACH protocol; V-LEACH; NS2; Wireless
Sensor Networks; Energy Efficiency
I. INTRODUCTION
A WSN is a wireless ad hoc network of sensor nodes that
are mainly used for monitoring applications. These sensor
nodes have very limited battery energy: hence energy
conservation is an important parameter to increase the overall
lifetime of the network [1]. Energy dissipation in sensor nodes
comes from: the sensing subsystem, the processing subsystem,
and the communication subsystem with the major source of
energy dissipation coming from communication [2]. In WSN
applications, the sensed data is gathered at regular intervals
and sent to Base Station (BS). Hierarchical routing protocols
are best known to minimize energy consumption by forming
clusters of nodes with a Cluster Head (CH), where the sensed
data from the Cluster Member (CM) nodes can be aggregated
and compressed before forwarding to the BS [3]. This avoids
the transmission of redundant data to BS and thereby reducing
energy consumed by the nodes.
The LEACH (Low Energy Adaptive Clustering Hierarchy)
[4] protocol is one of the most common energy-efficient
clustering protocols that uses this approach. Many protocols
are derived from LEACH; these are summarised in surveys
[5]–[9]. These versions address challenges inherent in
LEACH that mainly related to cluster formation which may
affect different performance metrics (network lifetime, data
throughput, network latency). Study of these versions show
how improvements are made to LEACH by considering
parameters not limited to the following: cluster formation
methodology (centralised or distributed), CH selection criteria
(random, energy-based, position-based, connectivity), node
roles (CH, CM, vice-CH, relay node), the round period length
of setup and steady phases, and hop communication (single,
multi-hop). These parameters affect the performance of the
protocol and have to be optimised as much as possible in
relation to performance metric goals. Most of the various
LEACH versions address a single parameter point with results
showing improvement in network lifetime.
This paper proposes a new protocol which further improves
on network lifetime performance and data throughput. To
achieve these goals, this paper looks at the various parameters
affecting the performance metrics given above. Our study
proposes YA-LEACH by combing the benefits of some of the
versions of LEACH while addressing their inherent challenges.
This paper proposes to address high cluster and CH energy
consumption which arises from poor clustering as a result of
random distributed CH selection. By using centralised
clustering, better clusters can be achieved which minimises
cluster and CH energy dissipation. This technique is proven in
LEACH-Centralised (LEACH-C) [10]. However, it has high
energy setup costs since all nodes send their information to BS;
our protocol addresses this by using CHs to send this
information after the initial setup round and also by reducing
the number cluster setups.
Given that centralised cluster formation produces better
clusters that require less energy for data transmission [11],
maintaining these optimal clusters for extended rounds
provides additional energy savings and at the same time
reduces setup costs. This principle of prolonging CH
replacement has been proposed in T-LEACH [12] and proven
to provide better network lifetime performance. However, T-
LEACH still inherits poor clustering as stated above,
centralised clustering addresses poor cluster formation. To
further prolong CH replacement, the proposed mechanism of
CH replacement will have an alternate CH, in which CH
transfers over its CH role to when it is at a predetermined
minimal energy level. The alternate CH will only assume the
role of CH until that round ends, at which point new clustering
done.
The principle in which alternate CH is implemented is
proposed in Vice Cluster Head LEACH (V-LEACH) [13], [14]
where the focus is the challenge of naming the successor in
the event of CH death as a result of high energy dissipation in
CHs. High energy costs in CHs arise from data aggregation
and transmission to BS and can deplete CH energy before the
end of the round, in this event vice-CH ensures that cluster
data collected for the remainder of the round reaches BS. In
contrast from having the vice-CH takeover after CH death, it
is proposed to use a minimal energy level to changeover to
vice-CH. This ensures data already collected by the old CH
will be able to reach BS and that the old CH will still be useful
as a CM (except in the event CH failure). Simulation results
show that the proposed YA-LEACH has an improvement in
performance in terms of network lifetime and overall data
throughput as compared to LEACH.
The remainder of this paper is organised as follows: Section
II discusses the principle of LEACH protocol; in Section III,
LEACH-C centralised cluster formation principle is discussed;
in Section IV, V-LEACH is introduced; Section V, we discuss
T-LEACH. Section VI, introduces the proposed protocol YA-
LEACH with Section VII giving details of the YA-LEACH
algorithm, in Section VIII, simulation and analysis of the
implemented proposal is given; Section IX concludes the
paper.
II. LEACH
The main approach of LEACH [4] includes self-organizing
algorithms for distributing cluster formation, adaptive CH
position changing and data aggregation techniques to reduce
redundancy. It is broken up into rounds, with each round
beginning with a setup phase in which clusters are organized
and then followed by a steady-state phase, where data
transmission to the BS is done. The setup phase is shorter than
the steady state phase.
A. Setup Phase
1) Advertisement Phase
At the beginning of the round, all nodes use a distributed
algorithm to determine whether they can become CH or not.
This is done by choosing a number between 0 and 1 which if
above a threshold value that node will become cluster head
else it will listen for CH advertisements. The threshold is set
as:
1
0
where P = percentage of CHs, r = current round, and G is
the set of nodes that have not been cluster-heads in the last 1/p
rounds. Nodes that succeed in becoming CHs will broadcast
advertisement messages to the other nodes which will respond
with join request message by considering the closest CH.
2) Cluster Set-Up Phase
Nodes that have not become CH will respond to
advertisement messages by sending join request message to
the CH. Nodes will do this by considering the closest CH to
itself and only send to that CH. CHs will wait to receive join
requests keeping their receivers on.
B. Steady Phase
1) Schedule Creation
Once the CH nodes receive all the join request messages
from nodes that would like to be included in the cluster, a
TDMA (Time Division Multiple Access) schedule is created.
The schedule will be created using the number of nodes in that
cluster and is broadcast back to CM nodes.
2) Data Transmission
With the cluster fixed and TDMA schedule received by all
the nodes, data transmission will begin. The CH will keep its
radio receiver on to receive data from CM nodes which will
wake up during their allocated TDMA slot. When all CM
nodes have finished sending their data to CH, the CH will
perform some data aggregation and compression on the data
before forwarding the data to the BS. After some
predetermined time for the round, the clusters will reorganise
themselves again beginning with the setup phase.
C. Problems of LEACH
1) CH responsible for sending data to BS, its failure
reduces network robustness. Proposals to solve this
problem can be seen in work done in [13], [14].
2) CH selection is random which does not consider
nodes residual energy. Proposals to solve this are found in
[10], [15], [16].
3) Dynamic clustering does not guarantee uniform CH
distribution. Proposals to solve this are seen in [10].
4) CH Transmission to BS consumes more energy for
nodes located further away causing poor energy balancing.
Proposals to solve this are found in [16],[17].
5) CH rotation and cluster formation is done after each
round which poses significant energy costs. A proposal to
solve this are seen in work done in [12],[18].
III. LEACH-C
LEACH Centralised [10] (LEACH-C) is an improvement
of the LEACH protocol. LEACH-C uses a centralised cluster
formation algorithm; simulated annealing. This may produce
better clusters by dispersing the CH nodes throughout the
network. During the set-up phase of LEACH-C, all the nodes
send information about their current location and residual
energy level to the BS which computes the average node
energy, and nodes having energy below this average cannot be
CHs for the current round. Once the CHs and associated
clusters are found, the BS broadcasts a message which
contains the cluster information to all nodes. Nodes which
have their own ID matching the broadcast cluster head ID
become CHs; otherwise node matches its cluster, determines
its TDMA slot for data transmission and goes to sleep until it
is time to transmit data. The network will then enter into the
steady state phase which is similar to that of LEACH.
A. Pros and Cons
The main advantage LEACH-C brings over LEACH is that
the centralised cluster formation ensures more optimal clusters
which guarantee the distribution of CH. The optimal clusters
also ensure minimal energy is used for data transmission in
every round [11]. In LEACH, random clustering does not
guarantee of CH distribution and may give poor clusters
which may affect energy performance in that round.
One major drawback of LEACH-C is that in the setup
phase all nodes have to send their information to BS directly
which consumes a lot of energy, this reduces the gains of
optimal clusters and lowers network lifetime performance. In
contrast, the proposed protocol will have a reduced number of
setups phases hence reducing the costs. Also, it will use CHs
to send network information to BS at the time when new
clustering has to be performed (except for in network
initialisation). This will minimise the cost in the setup phase,
and energy savings can be achieved.
IV. V-LEACH
V-LEACH [13] is a version of LEACH protocol which has
the cluster with; a CH which does aggregation and sending
data that is received from the CM nodes to the BS; a vice-CH
which is the node that will become a CH of the cluster in case
of CH death; CM nodes that gather data from environment
and send it to the CH.
In the original LEACH, the CH is always on to receive data
from CMs. It aggregates the data and then sends it to the BS.
BS might be located far away from some CH. CH nodes will
die earlier than the other nodes because of its operation of
receiving, sending and overhearing. In the event CH node dies,
all gathered data will never reach the BS.
The V-LEACH protocol uses an alternative CH node (vice-
CH) that will take the role of the CH in the event that the CH
dies. This ensures that cluster nodes data will always reach the
BS; there will be no need to select a new CH each time the
CH dies. This will extend the overall network lifetime [13].
Fig. 1 shows the V-LEACH clustering structure of nodes.
Base
Station
Cluster Head
Vice Cluster Head
Cluster Member Node
Fig. 1 Clustering in V-LEACH
A further extension was to include the process of CH
selection criteria on the basis of three factors i.e. Minimum
distance, maximum residual energy, and minimum energy
[14].
A. Pros and Cons
The main advantage V-LEACH brings is cluster robustness
in the event of CH death. It allows clusters to continue
functioning and sending data in the event of CH death. There
will be no need to select a new CH minimising setup costs.
Although vice-CH takes over as CH the data collected by CH
will still be lost, vice-CH only guarantees the continuation of
that round. There is no solution given in the case of vice-CH
death [8].
In contrast, the proposed protocol does not wait for CH
death to transfer the role to vice-CH but transfer occurs when
CH residual energy value reaches a threshold level. This
allows for CH to transfer its data to vice-CH and use the
remainder of its energy as a CM node. These ensure data
throughput and better energy balancing.
V. T-LEACH
The method of threshold-based CH replacement proposed
in T-LEACH [12], uses the a threshold residual energy value
to determine CH rotation instead of at the end of the each
round. As the frequency of rounds and of cluster head
replacement increases, energy consumption increases due to
the communication of advertisement messages. In T-LEACH,
the decision to perform new clustering is done based on the
additional residual energy in each node to replace CH. When
current CHs maintain residual energy at a level above the
predetermined threshold, CHs are not replaced even when it is
time to replace them at the end of each round. This is delayed
until the residual energy falls below the threshold, thus
making it possible for nodes to continuously play the roles of
CHs. This will minimise the number of CH selection phases,
reducing the CH replacement costs and thereby extending the
network lifetime. Fig. 2 shows how the round time period
looks like where Tsp (setup phase time) and Tdp (data
transmission phase time) which is longer than Tsp. Tdp is
divided in Tintra the time where data is collected by CH and
Tinter the time where data is sent to BS. Prolonging CH
replacement reduces Tsp and it associated costs.
T
sp
T
intra
T
inter
T
intra
T
inter
T
intra
T
sp
T
dp
...
... ...
T
dp
T
dp
T
round
T
inter
Fig. 2 The Round and phases in LEACH protocol
A. Pros and Cons
This approach focuses on minimising setup costs in the
network. Energy consumed setting up clusters is non-essential
to WSN application and has to be minimised. Reducing the
number of setup costs in the whole network lifetime provides
additional energy savings. This protocol shows it is better at
balancing energy and the network lifetime.
T-LEACH still inherits probability of poor clustering as in
LEACH and prolonging such poor clusters may also lead to
higher energy dissipation which balances off the gains. In
contrast, the proposed protocol centralised cluster formation
guarantees optimal clusters which give additional energy
savings and thus prolonging these clusters compounds the
energy savings. Also in T-LEACH threshold is calculated as
the threshold level of CH replacement energy refer to [12],
however in proposed protocol the threshold value is calculated
as the residual energy required to transfer one round cluster
data to vice-CH which is much lower. This further prolongs
CH replacement giving additional savings as it reduces setup
phases even further, also the availability of vice-CH enables
possibility lower thresholds as it ensures cluster robustness.
VI. PROPOSED YA-LEACH
The operation of V-LEACH given in [14], [13] is similar to
that of LEACH, its major improvement point is the inclusion
of an alternative CH which will take over in the event of the
CH death. LEACH uses distributed random cluster formation
algorithm which may lead to poor clustering [11], this in turn
will cause faster depletion of energy in some CHs. As the
length of a round is predetermined and fixed for all rounds
this means when the network lifetime is near the end, nodes
elected as CH will not have enough energy to complete a
round. This also means that the CH will die and render that
cluster’s data useless, for the remaining period of the round all
data will be lost. V-LEACH ensures that cluster will not
become useless as an alternative CH node will pick up from
where the CH left, this ensures that some cluster data will
always reach the BS. The selection of vice-CH is done by the
CH soon after it receives join request messages from CMs,
where it will consider minimum distance, maximum residual
energy, and minimum energy as given in [14]. Once selected
the steady state phase will begin, during the early stages of the
network lifetime selected vice-CH heads remain unused and
are discarded as new cluster formation is carried out. This
vice-CH selection is an extra overhead to cluster setup [6].
In this paper, we propose YA-LEACH in which we have
taken into consideration that the randomly distributed nature
of LEACH cluster formation algorithm does not ensure
optimal CH distribution [19]. Firstly, we have proposed to use
the centralized approach introduced in LEACH-C which tries
to ensure optimal clusters, this will ensure fair CH distribution.
LEACH-C however, has higher energy overhead in the
communication of cluster formation information in each round.
In contrast, to avoid communication costs adjustment has been
made so that nodes will communicate their residual energy to
BS via CH. Centralized cluster formation ensures highly
optimal CH distribution, which translates to minimum energy
consumption for data transmission [10].
Secondly, instead of selecting new CHs at the end of each
round the CH is allowed to extend operation into another
round provided it has enough residual energy. In our proposal,
the residual energy value is calculated as the minimum energy
required for CH to transfer its role to vice-CH. This will be
the energy of aggregating data of one round and energy to
transmit it to the vice-CH. This allows maintenance of clusters
for longer periods as compared to T-LEACH. However,
further prolonging CH replacement leads to the depletion of
most the CH energy. To provide some energy balancing the
CH transfer of its role to vice-CH done at some minimal
residual energy helps. This allows CHs to save some of its
energy to become a CM until it dies. CMs require much less
energy as these nodes mostly sleep and only turn on the radio
in their TDMA slot.
Prolonged CH replacement further reduces the number of
cluster formations which have a high energy cost and the
transfer of role to vice-CH before CH death provides for
energy balancing as there is high energy dissipation in CH
from extended round use. Maintaining optimally formed
clusters for extended rounds helps achieve additional energy
savings. Optimal clusters already give additional energy
saving from minimum energy consumption for data
transmissions. When making the decision to prolong
replacement at the end of the round CH will only consider if it
has greater residual energy than that required to complete one
round data aggregation and transfer to vice-CH as stated
above. Since CH will transfer its role to vice-CH when its
residual energy it at the threshold level, vice-CH will ensure
the completion of the round and data transmission to BS, after
which at the end of that round new clusters are formed. This
will be the basis for which new clustering is done.
VII. YA-LEACH ALGORITHM
A. Set-up phase (clusters are organized)
1) Cluster Head Selection & Cluster Formation
In the initial setup phase it will use the centralized approach
presented in LEACH-C [10], optimal clusters are formed
using simulated annealing and then communicated to the
nodes. For the subsequent cluster setups, the network
information (nodes residual energy) will be communicated via
the CH with BS handling the clustering.
2) Vice Cluster head selection
The basis for selection of vice-CH will consider the
minimum distance from the CH which will ensure minimal
communication energy consumption and also maximum
residual energy which is necessary for communication to BS
when vice-CH takes over. This is done as part of the cluster
formation process.
B. Steady-State (data transmission)
1) Schedule creation
Each node will receive information about which cluster it
belongs to and the CH and vice-CH. Each node will determine
its TDMA slot for data transmission and go to sleep until it is
time to transmit data. CH and vice-CH keep TDMA schedule
of all the nodes. For extended rounds, the TDMA schedule is
sent by CH as no changes to the clusters will have occurred.
2) Data transmission
The CH will gather data from CM nodes using TDMA
schedule and will aggregate all the data before forwarding it to
the BS at the end of the round.
3) Round Extension
This is when the CH will decide whether or not to continue
being the CH for the next round. At the end of each round, the
CH will make a decision as to whether it should carry on
being the CH. The decision is made by checking whether the
CH residual energy is enough to complete one round data
aggregation and transfer to vice-CH. CH will continuously
extend as long as residual energy is above the threshold.
C. Role of vice cluster head
The vice-CH will be on stand-by for the CH and will
incorporate low duty cycling by sleeping (turning off radio
equipment). It will behave as a normal CM node collecting
data and sending it to the CH in its allotted TDMA slot. In the
case that the CH’s energy is at a threshold level (transfer
energy stated in VI) it will communicate with the vice-CH to
become the current CH and then transfer its data over to the
vice-CH. CM nodes will be aware of vice-CH and when a
takeover has occurred they will transmit to vice-CH in their
allocated TDMA slot. Vice-CH will send a signal to the nodes
that it is now new CH and begin the data transmission. The
vice-CH will only be the CH for the completion of that round
and send the data to the BS upon which new cluster formation
will be done for the next round. This will be the way new
clustering is done.
A second mechanism the vice-CH will have is to determine
whether the CH is still alive is by periodically receiving Hello
messages from CH. If it misses a series of consecutive Hello
message it will try sending its own. If it fails to establish CH
presence it will be an assumption of CH failure upon which it
will take over as CH. At the end of that round, new clustering
will be done to ensure that new optimal clusters are formed.
This will further ensure cluster robustness.
The following Fig. 3 shows a flowchart of all the phases of
YA-LEACH
Cluster formation
START
Cluster Head Selectio n
Is CH
energy
> Transfer
_threshold
energy
YES
Data Transmission
NO
Vice-CH selectio n
Schedule Creation
Centralized
Fig. 3 Flowchart of the phases of YA-LEACH
VIII. SIMULATION RESULTS AND ANALYSIS
The network simulation software used is NS2. LEACH
protocol module is not available in NS2 so it has to be patched
in first and we then implement YA-LEACH based on this
LEACH protocol. Details on how to patch can be found in
[20], [21]. Finally, we compare and verify the whether the
performance of YA-LEACH gives better results base on the
following performance metrics
1) Network lifetime – this is the time until the remaining
nodes in the network cannot form a cluster or all the
nodes residual energy value are at 0J.
2) Data throughput – the total data received at Base
Station during network lifetime.
3) Energy consumption – the rate of energy consumed
by the network.
4) Cluster Setups – the number of times clustering is
done by the network performs
A. Simulation Parameters
Comparison of the YA-LEACH will be done against the
LEACH protocol using the following parameters
TABLE 1
SIMULATION PARAMETE RS
Parameter Value
Number of sensor nodes 100
Number of cluster heads 5
Simulated annealing algorithm 1000
Initial energy of node 2J
Distribution area of nodes 100m×100m
Base Station Location (50,175)
Size of packet header 25Bytes
Data size of packet 500Bytes
Eelec 50 nJ/bit
ε
f
riss
_
am
p
10 pJ/bit/m2
εtwo
_
ra
y_
amp 0.0013 pJ/bit/m4
E
f
usion 5 nJ/bit
B. Simulation Results
TABLE 2
NUMBER OF CLUSTER SETUPS
Protocol Total No. of
Rounds
No. of Cluster
Setups
Round
Clustering %
LEACH 25 25 100%
YA-LEACH 37 16 43%
Fig. 4 Comparison of Network Lifetime
Fig. 5 Comparison of data received at Base Station
C. Simulation Analysis
Fig. 4 shows that in YA-LEACH first node dies earlier than
LEACH. Extending CH rounds means that energy dissipation
of CH is higher than CM nodes this results in the uneven
energy dissipation. CH nodes transfer their duty to vice-CH
and only act as CMs for the rest of their life, this tries to
provide energy balancing. Results show that these nodes still
die off earlier, however energy savings are achieved from
reducing the number of setup phases as shown in Table 2.
Centralised clustering ensures CH distribution across sensor
field is optimal, even as nodes die out. This means data can
still be transmitted at a lower energy cost which provides
additional energy savings. YA-LEACH, although having
some uneven energy dissipation shows a significant
improvement in the lifetime performance of 40% over
LEACH.
Fig. 6 Comparison of Energy Consumption of Network
Fig. 5 shows the data received by the Base Station from the
clusters. YA-LEACH a fairly higher data throughput as
compared to LEACH. It always tries guarantees cluster data
will reach the BS even when CH energy becomes too low to
communicate with BS. Results show a 30% overall increase in
data received at the BS as a result of optimised clusters and
increased network lifetime. Fig. 6 shows the rate of energy
consumption, YA-LEACH consumes less energy and balances
off usage more than LEACH. This shows that there are
additional energy savings from prolonging CH replacement
and that optimal clusters use less energy.
Table 2 shows the number of setup phases that are done
during network lifetime. Where round clustering is defined as
the percentage of total possible cluster setups that can be done.
It shows that in YA-LEACH there is more than 50% reduction
of cluster setups which translates to energy being saved by
avoiding the cluster setup costs. The results show that the
proposed YA-LEACH protocol provides better performance
in the given performance metrics.
IX. CONCLUSION
In this paper, we study a well-known WSN clustering
protocol LEACH and some of its descendant versions. Our
study proposes a new protocol YA-LEACH. This combines
benefits derived from the various versions. It uses centralized
cluster formation to achieve optimal clusters which use
minimal energy for data transmission. It maintains these
optimal clusters for as long as possible, to compound energy
savings. Prolonging CH replacement also reduces the number
of cluster formations which gives energy savings by avoiding
the cost of frequent cluster setups. New clustering occurs
when CH transfers its role to the vice-CH head which further
prolongs CH replacement and safeguards CH failure.
Transferring CH role to vice-CH before CH death provides
energy balancing, this allows CH to use its remaining energy
as CM. This also improves data throughput to BS, as collected
data will be ensured to reach BS. Simulation results show that
there is an increase in the network lifetime performance and,
as a result, higher data throughput. It is noted that with
clusters being prolonged data collected in the cluster tends to
become redundant. Future work will look at realizing energy
savings in steady phase by using data negotiation techniques
to minimize redundant data transmission.
REFERENCES
[1] C. Kumar and A. Swaroop, “A Comparative Study of Hierarchical
Routing Protocols in Wireless Sensor Networks,” Comput. Sustain.
Glob. Dev. (INDIACom), 2015 2nd Int. Conf., March, pp. 1018 – 1023,
2015.
[2] J. Al-Karaki and A. Kamal, “Routing Techniques in Wireless Sensor
Networks: a Survey,” Wirel. Commun. IEEE, December, pp. 6–28,
2004.
[3] S. Das and P. D. Ganjewar, “Energy Efficient Cluster based
Hierarchical Routing Protocols in Wireless Sensor Network – a
Survey,” Multidiscip. J. Res. Eng. Technol., vol. 1, no. 1, pp. 6–22,
2014.
[4] W. R. Heinzelman, A. Chandrakasan, and H. Balakrishnan, “Energy-
efficient communication protocol for wireless microsensor networks,”
Proc. 33rd Annu. Hawaii Int. Conf. Syst. Sci., pp. 3005–3014, 2000.
[5] M. Aslam, N. Javaid, A. Rahim, U. Nazir, A. Bibi, and Z. A. Khan,
“Survey of extended LEACH-based clustering routing protocols for
wireless sensor networks,” Proc. 14th IEEE Int. Conf. High Perform.
Comput. Commun. HPCC-2012 - 9th IEEE Int. Conf. Embed. Softw.
Syst. ICESS-2012, no. July 2015, pp. 1232–1238, 2012.
[6] R. M. B. Hani and A. A. Ijjeh, “A Survey on LEACH-Based Energy
Aware Protocols for Wireless Sensor Networks,” J. Commun., vol. 8,
no. 3, pp. 192–206, 2013.
[7] R. K. Kodali, P. N. Sarma, and A. D. Diffusion, “Energy Efficient
Routing Protocols for WSN ’ s,” pp. 4–7, 2013.
[8] D. Prabha and V. K. Arora, “A Survey on LEACH and its Descendant
Protocols in Wireless Sensor Network,” Int. Conf. Commun. Comput.
Syst., pp. 162–167, 2014.
[9] A. Rahman, S. Anwar, I. Pramanik, and F. Rahman, “A Survey on
Energy Efficient Routing Techniques in Wireless Sensor Net works”
Icact, pp. 200–205, 2013.
[10] W. B. Heinzelman, A. P. Chandrakasan, H. Balakrishnan, “An
application-specific protocol architecture for wireless microsensor
networks,” IEEE Trans. Wirel. Commun., vol. 1, no. 4, pp. 660, 2002.
[11] N. Pantazis, S. Nikolidakis, D. D. Vergados, and S. Member, “Energy-
Efficient Routing Protocols in Wireless Sensor Networks: A Survey,”
IEEE Commun. Surv. Tutorials, vol. 15, no. 2, pp. 551–591, 2013.
[12] J. Hong, J. Kook, S. Lee, D. Kwon, and S. Yi, “T-LEACH: The method
of threshold-based cluster head replacement for wireless sensor
networks,” Inf. Syst. Front., vol. 11, no. 5, pp. 513–521, 2009.
[13] Yassein, “Improvement on LEACH Protocol of Wireless Sensor
Network (VLEACH),” Int. J. Digit. Content Technol. its Appl., vol. 3,
no. 2, pp. 132–136, 2009.
[14] A. Ahlawat and V. Malik, “An extended vice-cluster selection approach
to improve v leach protocol in WSN,” Int. Conf. Adv. Comput. Commun.
Technol. ACCT, pp. 236–240, 2013.
[15] M. Kamath and N. Maninarayana, “Cluster Head Selection Algorithm to
Minimize Energy Consumption in Wireless Sensor,” Int. J. Innov. Res.
Comput. Commun. Eng., vol. 2, no. 2, pp. 29–36, 2014.
[16] F. Xiangning, S. Yulin, “Improvement on LEACH Protocol of Wireless
Sensor Network,” 2007 Int. Conf. Sens. Technol. Appl., pp. 260, 2007.
[17] V. Loscrì, G. Morabito, and S. Marano, “A two-levels hierarchy for
low-energy adaptive clustering hierarchy (TL-LEACH),” IEEE Veh.
Technol. Conf., vol. 3, pp. 1809–1813, 2006.
[18] Z. Peng and X. Li, “The improvement and simulation of LEACH
protocol for WSNs,” Proc. 2010 IEEE Int. Conf. Softw. Eng. Serv. Sci.
ICSESS 2010, pp. 500–503, 2010.
[19] J. F. Yan and Y. L. Liu, “Improved LEACH routing protocol for large
scale wireless sensor networks routing,” 2011 Int. Conf. Electron.
Commun. Control. ICECC 2011 - Proc., pp. 3754–3757, 2011.
[20] L. Tian, H. Du, and Y. Huang, “The simulation and analysis of LEACH
protocol for wireless sensor network based on NS2,” Syst. Sci. Eng., no.
1, pp. 530–533, 2012.
[21] W. B. Heinzelman, “Application-Specific Protocol Architectures for
Wireless Networks,” Massachusetts Institute of Technology, 2000.
