CITATION: Usman, M. J., Xing, Z., Chiroma, H., Gital, A. Y. U., Abubakar, A. I., Usman, A. M.

, & Herawan,
T. (2014). Modified Low Energy Adaptive Clustering Hierarchy Protocol for Efficient Energy Consumption in
Wireless Sensor Networks. International Review on Computers and Software (IRECOS), 9(11), 1904-1915.

Modified Low-Energy Adaptive Clustering Hierarchy Protocol for Efficient Energy

Consumption in Wireless Sensor Networks for Healthcare Applications

Mohammed Joda Usman1, Zhang Xing1, Haruna Chiroma2a*, Tutut Herawan3

1
School of Electronics and Information Engineering,
Liaoning University of Technology, Jinzhou 121001, China
2*
Department of Artificial Intelligence
3
Department of Information System
University of Malaya, Kuala Lumpur, Malaysia
a
Department of Computer Science
Federal College of Education (Technical) Gombe, Nigeria
[email protected], [email protected], [email protected], [email protected],
*
Corresponding Author: Haruna Chiroma, Department of Artificial Intelligence,
University of Malaya, Kuala Lumpur, Malaysia. +60143873685.

Abstract:
In healthcare system, the sensor nodes are usually deployed in an unattended field or
environment and replacement of batteries is very difficult if not impossible. In this paper, Low-
Energy Adaptive Clustering Hierarchy (LEACH) protocol was modified (MoLEACH) to
improve energy efficiency of the LEACH for healthcare applications. In cluster head selection,
the MoLEACH consider the residual energy of each node for calculation of the threshold value
for the next round, unlike the original LEACH that uses the residual energy of the network.
Comparative simulation analysis between the MoLEACH and LEACH in testing different
parameters such as first node dead, half node dead, the effect of the number of nodes to the
network lifetime, and energy distribution was performed. The simulation results show that the
number of nodes affects the network lifetime in which increments of number of nodes decrease
the network lifetime. In small area, minimum number of nodes is better for network lifetime in
both MoLEACH and LEACH protocols. The MoLEACH shows improvement of energy
efficiency over the LEACH in energy distribution. The MoLEACH was found to improve the
energy efficiency of the LEACH, hence, prolong the network lifetime, thus achieved high
residual battery capacity. The MoLEACH proposed in this study can be used effectively in
healthcare applications, thereby reduces the need for frequent recharging or replacement of
batteries. The MoLEACH is an alternative to the LEACH in healthcare application systems such
as in in-home monitoring, in-hospital monitoring, ambulatory monitoring, vital sign monitoring
in-hospitals, monitoring elderly people at home care, monitoring in mass-casualty disasters and
clinical monitoring for automatic patient monitoring without disturbing patient comfort by the
need for frequent recharging or replacement of batteries.
Keywords: Wireless Sensor Networks; LEACH Protocol; Healthcare; Cluster-Based Routing;
Energy Efficiency; Low-Energy Adaptive Clustering Hierarchy

1. Introduction
Recently, a new class of network called Wireless Sensor Networks (WSNs) [1-2] emerged.
The aim of this technology is to provide efficient and effective connection between the physical
and virtual world. Propose applications of sensor networks include: environmental monitoring,
natural disaster prediction and relief, security, healthcare, manufacturing, transportation, and
home appliances. The WSN applied to healthcare applications are referred to as Healthcare
Wireless Sensor Networks (HCWSNs) [3]. The difference between HCWSNs and WSNs
technology application is the criticality and the excellence, reliability of data transmission, which
has a number of characteristics that differentiate HCWSNs from the standard WSNs and WLAN.
Hence reliability is of paramount importance in the HCWSNs.
A WSNs is a network consisting of numerous sensor nodes with the capability of sensing
wireless communications and perform computing capabilities. These sensor nodes are scattered
in an unattended environment (i.e. Sensing area) to sense the physical world. The sensed data can
be collected by a few sink nodes which have access to infrastructure networks like the Internet.
Subsequently, end user can remotely fetch the sensed data by accessing the infrastructure
networks. These networks have the capabilities of collecting audio, seismic and other type of
data and collaborate to perform a high level task in the network [3-4].
It has been observed that about 70% of energy is consumed during the data transmission
phase. Thus, energy consumption is one of the major problems in WSNs applications. In
healthcare system, the sensor nodes are usually deployed in an unattended field or environment
and replacement of batteries is very difficult if not impossible. The sensor nodes are used to
monitor patients, track doctors, and nurses, transmit accurate sense information about some
hospital resources to the Base Station (BS) for analysis. These communication activities cause
the depletion of the sensor node energy quickly, reduce network lifetime, and hence expose the
patient's life into risk due to the need of real time information about the patient condition and the
current location of the doctors/nurses in case of any emergency [2,4-6].
As far as wireless environment is concerned, it consumes significant amount of energy.
Sensor node should consume as little energy as possible for receiving and transmitting data.
Currently, researchers put their efforts in sensor networks focusing on the issues involving the
development of an energy efficiency protocol. Recently, many algorithms and protocols about
energy efficiency have been proposed in the literature. It was found that the cluster-based model
is better than single-hop or multi-hop model in the optimization of energy efficiency in WSNs as
discussed in [7]. The Low-Energy Adaptive Clustering Hierarchy (LEACH) was proposed to
improved energy efficiency in WSNs. Yet, the LEACH has been susceptible to the limitations
[7]: LEACH is not efficient for large-scale networks. Fix percentage of cluster-heads for any size
of network (5%). The protocol may lead to concentration of cluster-heads in one area of the
network. The LEACH assumes that all nodes can communicate over one hop (directly) with the
BS. Uniform energy dissipation assumed in any given round. All nodes start with equal energy
residual levels.
In this paper, we propose to modify the LEACH (MoLEACH) protocol to improve
efficient energy consumption of the LEACH in order to increase the lifetime of the networks.

2. Related work
A survey of clustering algorithms for WSNs was presented by Anastasi et al. [2]. The authors
presented a taxonomy and classification of typical clustering schemes. The summary of the
different clustering algorithms for WSNs based on classification of variable convergence time,
protocols and constant convergence time were highlighted. Abbasi et al. further compared
performances of clustering approaches based on energy conservation, convergence rate, cluster
stability, cluster overlapping, location-awareness and support for node mobility. RMCP groups
wireless sensor nodes were integrated into clusters to detect signals with the goal of prolonging
the MASN lifetime, load balancing, and scalability. It was stated that the RMCP differs from
LEACH protocols because it takes into consideration the energy level determination of sensor
nodes, event-triggered and energy-aware cluster formation [8]. Malan et al. [3] presented a
comparison survey between different clustering protocols. The study briefly analyzed LEACH-
based protocols as well as proactive and reactive algorithms in WSNs. The main characteristics
of these protocols were compared and relevant application areas were pointed out. The
operations of clustering protocols were discussed in the survey presented by Hande et al. [9], the
strengths and limitations of each one of these algorithms were analyzed. Seven popular
clustering algorithms of the WSNs such as the LEACH, TL-LEACH, EECS, TEEN, APTEEN,
etc were considered. In addition, the survey compared these clustering protocols in terms of
energy consumption and network lifetime.

3. Methodology

3.1 The Original Low-Energy Adaptive Clustering Hierarchy

LEACH algorithm is a cluster members elect a cluster head to avoid excessive energy
consumption [10]. LEACH is a self-organizing, adaptive clustering protocol that uses
randomization to distribute the energy load evenly among sensors in the network as shown in
Fig.1. In LEACH, the nodes organize themselves into local clusters, with one node acting as the
local BS or cluster-head. If the cluster heads were chosen as priori and fixed throughout the
system lifetime similar to conventional clustering algorithms, it is easy to see that the unlucky
sensors chosen to be cluster-heads would die quickly, ending the useful lifetime of all nodes
belonging to those clusters.
 Fig. 1 The LEACH protocol for Wireless Sensor Network.

LEACH protocol includes the sensor nodes, which randomly organize into a local cluster
with one becoming the cluster head. The LEACH consists of 2 phases: set-up phase and steady-
phase. In the setup phase, sensors can be selected randomly among themselves as a local cluster
head with a certain probability. By doing so, the network may balance energy dissipation across
the whole network. The optimal number of cluster heads is 5% of the total. After the cluster
heads are selected, the heads advertise to all sensor nodes in the network that they are new
cluster heads. Once the nodes receive the advertisements, they decide which head they belong. In
the steady-phase, sensors sense and transmit data to the sink through their cluster heads. In the
same manner, other clustering routing protocols, transmits their information to the sink, and thus,
they make mono-level hierarchy in WSNs. After a certain period spent in the steady state, the
network goes into the setup phase again and enters another round of selecting cluster heads [10].
Heinzelman et al. [7] LEACH forms clusters using a distributed algorithm in which nodes
make autonomous decisions without any central control. In the beginning of round r + 1 (which
start at time (t)) with probability T(n), let each sensor n selects itself to be CH. T(n) is chosen
such that the expected number of CH nodes for this round is k. Thus, if there are N nodes in the
network then:
(1)

Ensuring that all nodes are CH, the same number of times requires each node to be a CH
one in rounds on average. If Cn(t) is the indicator function to determine whether node has
been a CH in the most recent (r mod rounds i.e Cn(t) = 0, if node has been a CH and one
otherwise), then each node should choose to become a CH at round r with probability

(2)
This means that only nodes that have not previously been a CH and considered to have more
energy available from the node that recently have intensive energy function can be a CH at round
r + 1. The expected number of nodes that have not been a CH in the first rounds r is N – K × r.
After rounds, all nodes expected to have been CH once, following which they are eligible to
perform this task in the next sequence of rounds. Since Cn(t) is one if node n is eligible to be a
CH at time t and zero otherwise, the term represent the total number of nodes that
are eligible to be a CH at time t and

(3)
This ensures that all nodes are having approximately equal energy after rounds. Using Eqns.
(2) and (3), the expected number of CH per round is

(4)

Each CH dissipates energy receiving signals from the nodes beam forming the signals and
transmitting the aggregate signal to the BS. Since the BS is located away from the nodes, it can
be assumed that the distance of BS is greater than the crossover distance and the energy
dissipation follows a two-ray ground model (e.g d4 power loss). The energy dissipated in the CH
node during a single frame can be expressed as:
(5)
Where l is the number of bits in each data and d4toBS is the distance from the CH to the BS.
Finally, when a node has selected itself as a CH based on Eqn. 2, it broadcasts an advertisement
message informing all nodes that it is a cluster head. This advertisement is done using Carrier
Sense Multiple Access (CSMA) protocol. Non-cluster heads use these messages from the cluster
heads to choose the cluster they want to belong for this round based on the received signal
strength of the advertisement message.
3.2 The Propose Modification of the Low-Energy Adaptive Clustering Hierarchy protocol
LEACH uses a distributed algorithm for the formation of the cluster head node, each node
make decisions by itself without a central control. In the case of MoLEACH, the cluster head is
selected randomly and nodes become the cluster head if cluster heads have the same initial
energy. To increase the lifetime of the networks, the MoLEACH uses a probability function
while considering the use of node residual energy for cluster configuration whereas LEACH only
utilizes a probability function. In the case of the nodes which do not have the same amount of
energy (for second round and afterward), the nodes with higher residual energy will become a
CH more often than nodes that have less energy to ensure that all nodes die approximately at the
same time. This can be achieved by setting the probability of becoming a cluster head as a
function of a node’s energy level relative to the aggregate energy remaining in the network,
rather than purely as a function of the number of times the node has been head clustered, thus,
 (6)

where p is the desired percentage of cluster heads, r is the current round number, G is the set of
nodes that have not been CHs in the last rounds, Eresidual is the current residual energy of node
and

(7)

where Etotal is the total energy of the entire network and by using these probabilities, the nodes
with higher energy will become cluster heads rather than nodes with less energy. The expected
number of cluster head nodes is given by Eqn. (8)

(8)

The optimal value of cluster, k can be determined as explained in the original LEACH. To
calculate the threshold value (Eqn. 6) for nodes to become CH for the next round, it is essential
for each node to transmit their current residual energy to BS so that the total energy for entire
network can be calculated using Eqn. 7. This process is done in pre set-up phase which is the
improvement on the original LEACH [7]. The the flow chart and pseudo code of the proposed
MoLEACH are depicted in Fig. 3 and Fig. 4. The three phases of the MoLEACH are described
as follows:
3.2.1 Set-up Phase
Each node generates a random probability (pn) at the beginning of a new round and computes
the threshold value T (n) by using Eqn. 6. If r = 1 (i.e. the first round) and Eresidual for each node
are the same. In case of pn < p, the node is selected as a CH. A selected CH broadcasts an
advertised message over neighbor nodes. The neighbor nodes collect advertised messages during
a given time interval and then send a “join REQ” message to the nearest CH. The cluster head
receives the “join- REQ” messages and builds a cluster member list and a Time Division
Multiple Access (TDMA) schedule. Subsequently, broadcast over neighbor nodes. The member
node receives and save the message for data transfer.
3.2.2 Steady-state phase

The timeline of the proposed MoLEACH algorithm, the steady-state operation is divided into
frames. The main activities in this phase are sensing and transmission of sensed data. Each
sensor nodes senses and transmits the sensed data to its CH according to TDMA table as shown
in Fig. 2. In addition, for next round to occur, it is required for the cluster node to transmit
together their current energy and their ID’s. When data and energy status has been received, the
CH performs data fusion and aggregation in order to reduce the amount of data. Finally, each CH
transmits data to BS along the CH-to-CH routing path which have been formed during the setup
phase.
 Fig. 2 MoLEACH algorithm TDMA frame
3.2.3 Pre set-up Phase

Before the last frame of a round completes, the CH sends BS the residual energy value of each
nodes that belong to its own cluster. BS collects all residual energy values from CHs, finds the
total residual energy value (Etotal) of the network, and Etotal back to CHs. The CH broadcasts
Etotal over cluster nodes. Each nodes save the value of Etotal for the next computation of T(n)
and the current round is terminated.
Fig. 3 Pre set-up phase for MoLEACH algorithm
BEGIN
1) Specify the probability, p and the number of nodes; (nrNodes) // p is set to 0.05
2) Eresidual (n) = E0 , n = 1,2,…,m; // initial energy for each nodes
SET-UP PHASE
3) Compute threshold value T (n);
4) While (cluster head counts < Needed cluster head, p) && (node < nrNodes)
5) Assign a random number to each nodes;
6) If (random number < threshold value);
7) Cluster head = TRUE; //only 5% nodes become cluster head
8) Else
9) Cluster head = FALSE;
10) End if
11) If ( CH (n) = TRUE) then
12) BC (ADV) ; //Broadcast and advertisement message
13) Join (IDi); // non-cluster head node i join into the closest cluster head
14) Cluster c; // form cluster c
15) End if
STEADY-STATE PHASE
15) If (CH (n) = TRUE) then
16) Receive (IDi, DataPCK); // Received data packet and residual energy from members
17) Aggregate (IDi, DataPCK); // aggregate receive data and residual energy
18) SendtoBS (IDi, DataPCK); // transmit receive data and residual energy to BS
19) Else
20) If (MyTimeSlot = TRUE) then
21) SendtoCH (IDi, DataPCK); // transmit sensed data
22) Else
23) SleepMode (n) = TRUE; // node n in sleep mode
24) End if
25) End if //round done for first round
PRE SET-UP PHASE
26) statusNodeReceive (IDi, DataPCK); // received data and residual energy from cluster head
27) for (n = 0; n < total node; n++)
28) calculate TotalEnergy = ; // Calculate maximum residual energy
29) BC (MaxEnergy); // Broadcast maximum residual energy to the network through CH
30) Go to 1; // Start for second and next round

Fig. 4 Pseudo-code of the MoLEACH Algorithm

3.4 Implementation of MoLEACH

3.4.1 Snapshot of OMNET++ Simulation for moLEACH Algorithm
The development of moLEACH under OMNeT++ requires a NED (module description) and a
C++ (functionality description) for implementation. For that purpose, we use the solar-LEACH
simulation developed by Voigt et al. [11]. The NED file for moLEACH algorithm was developed
using GNED editor in OMNeT++ version 4.1. GNED is a graphical environment, facilitating the
development and debugging of simulation topologies. Fig. 5 is the graphical and text
representations of model topologies that can be used interchangeably so that editing,
visualization and parameter setting can be done in the text or graphical views. GNED supports
two views of the model topology: graphical editing view, which is the default view, the simulator
starts up and display the NED source code editor (see Fig. 5).
 Fig. 5 Graphical and source view for LEACH implementation in GNED editor

3.4.2 Experiment Setup

The layout of the MoLEACH network is shown in Fig. 6; the nodes are randomly distributed
in the field region.
 Fig. 6 Node Detail for LEACH network

The network started the setup phase by sending status of the nodes to the BS. Each node has a
random number to calculate the threshold value and determine the CH as depicted in Fig. 7 .

Fig. 7 Each node sends its status to the BS

The red nodes are the CHs that have been chosen for comparing the random number to the
threshold value as shown in Fig. 8.
 Fig. 8. Cluster head formation in LEACH network simulation

After a selection of CHs, the algorithm Moloch proceeds with a steady - state phase where
TDMA schedule is created as shown in Fig. 9.

Fig. 9 TDMA schedule creation

Fig. 10 shows the data sensed by nodes which is transmitted to CH and thus, the data were
aggregated to BS and Fig. 11 shows the nodes after a couple of rounds.
Fig. 10 Data gathering by the nodes
 Fig. 11 Snapshot after a couple of rounds
3.4. 3 Implementation of LEACH Algorithm for evaluation
The LEACH C++ implementation consists of two different files: the header (.h) file, which
contains forward declarations of variables, structures and subroutines, and source (.cc) file,
which contains the implementation of the LEACH algorithm.
3.4.4 Initialization of The Module Network Parameters
The initialization of the network parameters was made through the omnetpp.ini file, a
simulation initialization file required to set the initial simulation values. The parameters and their
corresponding values are:
3.4.5 Generic parameters: (omnetpp.ini)
i. solar.numNodes: Number of sensor nodes in the network
ii. solar.xMax: Maximum x-position on grid of the node
iii. solar.yMax: Maximum y-position on grid of the node
iv. solar.rounds: Number of rounds for which the solution is to be run
v. solar.frames: Number of frames sent to head by each node in cluster

3.4.6 Simulation Parameters

For the purpose of simulation, the simple radio model was adopted from [7]. The described
model in the referred source assumes energy dissipation of Eelec = 50 nJ/bit to run the
transmitter or receiver circuitry and Eamp = 100 pJ/bit/m2 for the transmit amplifier to achieve
an acceptable (Signal to Noise Ratio). The first order radio model is as shown in Fig. 12.
Other parameters used in the simulation are described in Table 1 and Fig. 13 display screenshot
of the parameter settings in the simulator.

Fig. 12. First order radio model [7]

Table 1 Parameters used in the simulation

Type Parameter Value

From (0,0) to 1) (500, 500), to 2)
Network Network Grid (3 × 3)
(1000,1000) and to 3) (1500, 1500)
Base Station 100, 150, 200 meter
Initial Energy 0.5 Joule/battery/node
Node Distribution (N) 100 nodes randomly distributed
Application Rounds (frame) 5, 10, 20, 25 TDMA frames
Broadcast packet size 25 bytes
Packet header size 25 bytes
Transmitter energy consumption (ETX-
Radio Model 50 nJoule/bit
elec)
Receiver energy consumption (ERX-elec)
50 nJoule/bit
(ETX-elec = ERX-elec = Eelec)
Amplifier energy consumption (Eamp) 100 pJ/bit/m2

It also assumes that, an r2 energy loss due to the channel transmission. Thus, to transmit and
receive a k-bit message a distance (d) using this radio model, the radio expends as expressed in
Eqn. (9) and (10):
 (9)

(10)

where λ represent a path-loss exponent. The value of these parameters makes the message
transmission and reception not low cost operations. Therefore, the protocols should minimize the
number of transmit and receive operations by means of switching its state between active and
sleep (or idle) when required in order to minimize the energy consumption.
3.5 Testing Parameters for the Simulation
In order to obtain multiple results, a series of test has been prepared for the different routing
protocols earlier described. The network grid shows the different areas of the network used in the
simulations (500x500 m2, 1000x1000 m2 and 1500x1500 m2 ) as shown in Fig. 12. Each of
these areas has been simulated in which the BS was set to 100, 150 and 200 meters to the closest
node of the network to find out the behavior of the protocols when the BS is placed at different
distances far from the sensor nodes.
For each of the conditions, each protocol has been simulated with rounds comprised of 5, 10,
20 and 25 TDMA frames, which results in a short or long steady-state phase (see Fig. 3). All the
parameters implemented in omnetpp.ini file in OMNeT++ were altered each time the simulation
is executed.
 Fig. 13. Omnetpp.ini file for parameter setting

Therefore, based on the initial energy parameter (see Table 1), a full-charged battery energy level
of 0.5 Joules has been chosen, since it is sufficient to see the differences in the results among the
different protocols evaluated. In all the simulations, the network always consists of 100 nodes.
Hence, in large area networks such as 1500x1500 m2, the node density is lower than in small area
networks such as 500x500 m2.

4 Results and discussion

4.1 First Node Dead

In the case of a short steady phase and a small area network, the MoLEACH gets better
results, by achieving more than 2 times the lifetime of the LEACH as shown in Fig. 14.

500 X 500 & 5 Frames

400
Rounds Done

300
200
100 MoLEACH
0 LEACH
100 150 200
BS Distance (m)
 (a)

1000 X 1000 & 5 Frames

400
350
300
Rounds Done
250
200
MoLEACH
150
100 LEACH
50
0
100 150 200
BS Distance (m)

(b)

1500 X 1500 & 5 Frames

400
350
300
Rounds Done

250
200
MoLEACH
150
100 LEACH
50
0
100 150 200
BS Distance (m)

(c)
Fig. 14. Number of rounds when the first node is dead in LEACH and MoLEACH for 5 frames
in different BS distances.

When the node density decreases or the area, network increases, the results of both protocols get
closer, but MoLEACH still have better results as shown in Fig. 15.
 500 X 500 & 10 Frames
200

Rounds Done
100
MoLEACH
0 LEACH
100 150 200
BS Distance (m)

(a)

1000 X 1000 & 10 Frames

200
Rounds Done

150
100
MoLEACH
50
0 LEACH
100 150 200
BS Distance (m)

(b)

1500 X 1500 & 10 Frames

150
Rounds Done

100

50 MoLEACH
LEACH
0
100 150 200
BS Distance (m)

(c)
Fig. 15 Number of rounds when the first node is dead in LEACH and MoLEACH for 10 frames
in different BS distance.

If the duration of the steady-state phase increases, the lifetime of both protocols will decrease as
observed in Fig. 16. The results show that the MoLEACH outperforms the performance of
LEACH as shown in Fig. 17.
 (a)

1000 X 1000 & 20 Frames

80
Rounds Done

60
40
MoLEACH
20
LEACH
0
100 150 200
BS Distance (m)

(b)

1500 X 1500 & 20 Frames

80
Rounds Done

60

40
MoLEACH
20
LEACH
0
100 150 200
BS Distance (m)
(c)

(c)

Fig. 16 Number of rounds when the first node dead in LEACH and MoLEACH for 20 frames in
different BS distance
Fig. 17 indicates the results of both protocols with 20 frames. It shows that the network lifetime
becomes shorter as previously mentioned, but MoLEACH still get the highest number of rounds
before first node dead. It can also be seen that the more the distance from the BS to the node, the
more network lifetime is decreased.

500 x 500 & 25 Frames

50

40
Rounds Done

30

20 MoLEACH

10 LEACH

0
100 150 200
BS Distance (m)

(a)

1000 x 1000 & 25 Frames

50
40
Rounds Done

30
20 MoLEACH

10 LEACH

0
100 150 200
BS Distance (m)

(b)
 1500 x 1500 & 25 Frames
50

40

Rounds Done 30

20 MoLEACH
LEACH
10

0
100 150 200
BS Distance (m)

(c)
Fig. 17 Number of rounds for the first node dead in LEACH and MoLEACH for 25 frames from
different BS distance

The results of the two protocols (MoLEACH and LEACH) obtain in a steady-state phase of 25
frames, shows that both protocols have short lifetime when the steady-state phase is long. The
results further indicated that even though the network lifetime decreases, MoLEACH network
was found to be 2 times better than LEACH. It can be observed in Figs.14-17 that in the long
steady-state phase, the network lifetime for both the protocols decreases as explained earlier. The
proposed MoLEACH prolong the network lifetime. The consideration of node residual energy
during cluster head selection, processing can maintain the balanced energy consumption of the
sensor network.
4.2 Half Node Dead
The analysis of the half node dead is performed to observe the overall performance of
LEACH and MoLEACH. Comparison among all the protocols in the rounds was achieved until
half of the nodes are dead. It is clear that the number of half node dead in MoLEACH occur in
the rounds about 2 times as in the first node dead as shown in Fig. 18. The more the network size
increase, the more the network lifetime of MoLEACH and LEACH protocols decreases.
 (a)

(b)

(c)

Fig. 18 Number of rounds when the half node dead in LEACH and MoLEACH for 5 frames
from different BS distance.

Steady-state that is formed by 10 frames in a small and large area network is explained based on
Fig. 19. In the case of a steady phase formed by 20 frames, the results of both protocols show a
similar behavior in the previous case. Their respective outcomes get closer to each other in a
large area network, but LEACH shows more constant behavior than the MoLEACH with the BS
placed at different distances.

(a)

(b)

(c)
Fig. 19 Number of rounds when the Half Node Dead in LEACH and MoLEACH for 10
frames from different BS distance

(a)

(b)

(c)
 Fig. 20 Number of rounds when the Half Node Dead in LEACH and MoLEACH for 20 frames
from different BS distance.

When the steady phase has double the number of frames of the previous case, the behavior of
both protocols remains the same, but decreasing the number of rounds will achieve almost half of
the protocol's behavior as shown in Fig. 20. This situation can be explained as an example of a
low-cost set-up phase in energy terms, whereas the high cost steady-state phase is due to a non-
optimal election of the cluster heads and the direct communication between cluster heads and
base station. It can also be observed that the farther the BS from the nodes, the more the network
lifetime is decreased.

(a)

(b)

Fig. 21 Number of rounds when the Half Node Dead in LEACH and MoLEACH for 25 frames
from different BS distance.

In this scenario the results of both protocols with the longest steady phase simulated outcomes
are really similar to the previous ones as expected, but decreasing the overall amount of rounds
achieved. By analyzing the result from Figs.18-21, modified LEACH still achieves a higher
number of rounds, but it decreases as the TDMA frame increases. The network lifetime for both
protocols becomes worse as the network size increases. As the network size and TDMA frame
increases, the result for both protocols become closer to each other.
4.3 Effect of Number of Nodes to the Network Lifetime
The simulation was performed to study the effect of different number of nodes on the network
lifetime. In the simulation 50, 75 and 100 nodes were implemented in 500 x 500 m2, 1000 x 1000
m2 and 1500 x 1500 m2 network sizes.

(a)

(b)

Fig. 22. First Node Dead results for different nodes implemented in different network size

The first node dead comparison for different number of nodes and network size is shown in
Fig. 22, when the number of nodes increases, the network lifetime decreases for both protocols.
For larger number of nodes and bigger network size, the network lifetime becomes closer to each
other, but MoLEACH has shown improvement over LEACH. It can also be observed that the
LEACH shows a stable network lifetime for bigger network size, whereas MoLEACH shows
that the network lifetime decreases linearly and outperform the LEACH.

(a)

(b)

Fig. 23 Half Node Dead results for different nodes implemented in different network size

Half Node Dead for both protocols decrease with the increment of the number of nodes as
depicted in Fig. 23. The network size increases, the network lifetime shows a slight decrement.
But as mentioned earlier in the above analysis, it is concluded that performance of modified
LEACH is still better than LEACH.
4.4 Improvement made by MoLEACH
The network lifetime improvement of MoLEACH over LEACH is described in this section.
From the simulation and analysis in the preceding sections, the network size and the TDMA
frames affect the performance of the network lifetime in WSNs. The percentage of improvement
made by MoLEACH over the original LEACH is computed and reported in Table 1.
 The improvement of MoLEACH is greater with short steady-state phase, small network size
and closeness of BS distance from the nodes. This is expected because the energy used to
aggregate data, when BS is closer is less than energy if the BS distance is far away. When the
steady-state becomes longer, the improvement becomes less, yet, MoLEACH shows a better
network lifetime than the LEACH. For First Node Dead improvement, the shorter steady-state
phase shows only a slight improvement over LEACH. The improvement becomes worse, when
the BS distance is far away both protocols almost have the same performance. It shows that the
steady-state phase improvement of MoLEACH is better than the original LEACH in short
distance.

Table 2. Percentage of improvement made by MoLEACH for First Node Dead

MoLEACH Improvement for Healthcare Applications

Area Network(m²) 500 x 500 1500 x 1500
BS distance (m) 100 150 200 100 150 200
5 frames (%) 67 65 71 43 60 50
25 frames (%) 9 6 4 12 17 10

4.5 Distribution of Energy in LEACH and Modified LEACH

The distribution of energy in LEACH algorithm is uneven between the node if the CH
becomes unreasonable. The proposed MoLEACH has encountered this problem by taking into
account the node’s residual energy in selecting the CH for the purpose of simulation, each node
is supplied with 0.5 Joule battery. Fig. 24 shows the energy distribution between each node for
LEACH and MoLEACH after the simulation.
 Fig. 24 Energy distributions in LEACH and MoLEACH

In LEACH, there is a different amount of energy for cluster formation at the beginning of
each round, as the total energy depends on the number of nodes that select themselves to be CH
and their locations from the network. A large number of CH nodes imply that more energy is
dissipated in the network. Thus, the MoLEACH shows more energy-efficient than the original
LEACH in which better cluster can be determined than the distributed algorithm used in the
LEACH. The probable reasons for the energy efficiency of the MoLEACH can best be
explained by the following two possibilities: (1) The MoLEACH used a probability function
while considering the use of node residual energy for cluster configuration which could have
probably improved the energy efficiency of the MoLEACH. (2) The probability function might
have improves the CH selection by taking into account the residual energy of each node for
threshold calculation before another round started. The MoLEACH can effectively replace the
LEACH in healthcare application systems such as in-home monitoring, in-hospital monitoring,
ambulatory monitoring, vital sign monitoring in-hospitals, monitoring elderly people at home
care, monitoring in mass-casualty disasters and clinical monitoring for automatic patient
monitoring without disturbing patient comfort by the need for frequent recharging or
replacement of batteries.

5. Conclusions

In this paper, we proposed to modify LEACH to improve its energy efficiency in WSNs for
healthcare applications. The original LEACH was modified in the research and code name as
MoLEACH. The MoLEACH and the LEACH has been implemented in the OMNeT++ and
evaluated in different simulation scenarios of large scope clustered type networks. The
evaluation of the simulation scenarios was performed using several testing parameters such as
TDMA frames in different network sizes, BS distances, number of nodes, network size, First
Node Dead and Half Node Dead. The simulated comparative analysis of the results clearly
indicates that the performance of MoLEACH in terms of network lifetime outperforms the
LEACH. Thus, the MoLEACH advance the energy efficiency of the original LEACH. Therefore,
it can be determined that the MoLEACH proposed in this study can be implemented in the
design of HCWSNs architecture for effective, robust, energy efficient and promising applications
in healthcare systems to significantly reduce the need for frequent recharging or replacement of
batteries.
Conflict of Interests
“The authors declare that there is no conflict of interests regarding the publication of this paper”

Acknowledgment. This work is supported by High Impact Research Grant University of Malaya
no vote UM.C/628/HIR/MOHE/SC/13/2.

References
1 Akyildiz IF, Su W, Sankarasubramaniam Y, Cayirci E: Wireless sensor networks: A survey.
Computer Networks, 2002, 38(4): 393-422.
2 Anastasi G, Conti MD, Francesco M, Passarella A: Energy conservation in wireless sensor
networks: a survey. Ad Hoc Networks, 2009, 7(3): 537–568.
3 Malan D, Fulford-Jones T, Welsh M, Moulton–CodeBlue S: An ad-hoc sensor network
infrastructure for emergency medical care. International Workshop on Wearable and
Implantable Body Sensor Networks, 2004, 4(5): 12-14.
4 Wang L, Xiao Y: A survey of energy-efficient scheduling mechanisms in sensor networks.
Mobile Network Application, 2006, 11(5): 723-740.
5 Tilak S, Abu-Ghazaleh NB, Heinzelman W: A taxonomy of wireless micro-sensor network
models. ACM SIGMOBILE, 2006, 6(2): 28-36.
6 Yick J, Mukherjee B, Ghosal D: Wireless sensor network survey. Computer Networks.
2008, 52(12): 2292-2330.
7 Heinzelman WR, Chandrakasan A, Balakrishnan H: Energy-efficient communication
protocol for wireless microsensor networks. Proc. of the 33rd Hawaii International
Conference on System Sciences, IEEE Explore. New Delhi, 2010, 1-10.
8 Young Jang K, Kim K, Yong Youn H: An energy efficient routing scheme for wireless
sensor networks. International Conference on Computational Science and its Application
(ICCSA). Kuala Lumpur, Malaysia 2007, 195-206.
9 Hande A, CemErsoy E. Wireless sensor networks for healthcare: a survey. Computer
Networks, 2010; 54(15): 2688–2710.
10 Tao L, Qing-Xin Z, Luqiao Z: An improvement for LEACH algorithm in wireless sensor
network. Proc. of the 5th International Conference on Industrial Electronics and
Application. IEEE Explore, 2010, 1811-1814.
11 Varga A: OMNeT++ Manual. http://www.omnetpp.org, 2002-03-16/2003-11-02.