Energy Optimisation using Distance and Hop-based Transmission
(DHBT) in Wireless Sensor Networks
Scheme and Simulation Analysis
T. S. PradeepKumar
1
, P. V. Krishna
2
, M. S. Obaidat
3
, V. Saritha
4
and K. F. Hsiao
5
1
VIT University, Vellore, India
2
Sri Padmavati Mahila University, Tirupati, India
3
Fellow of IEEE, Fordham Univeristy, U.S.A.
4
Sri VIdyanikethan Engineering College, Tirupathi, India
5
University of Jordan, Jordan, and Ming-Chuan University, Taiwan
Keywords: Sensor Node, Simulation, Distance, Multi-Hop, Power, Lifetime.
Abstract: Wireless Sensor networks operate in a low energy mode and consume less power. The ultimate challenge of
a sensor node is to make the lifetime of the node increase by which the energy consumption is so minimal.
This paper addresses a mechanism by which the distance between any source and destination nodes is used
by the source nodes to decide whether transmission of the message to destination node must be multi-hop or
direct transmission by simply boosting the node power. The network size and the topology are considered
for determining the threshold value for the distance based on which the decision is made. The simulation
results have shown that the proposed method, DHBT, can increase the lifetime of the sensor network by at
least 130% when compared to the legacy systems.
1 INTRODUCTION
Wireless sensor networks play a major role in
applications like surveillance, animal habitat,
agricultural monitoring and observation, nuclear
radiation observation, location tracking, smart
homes and industrial automation (Obaidat and
Misra, 2014). Usually small sized nodes are
deployed in an environment. These nodes are used to
monitor the environment, collect the data and report
to the gateway node or base station nodes. All this
work is done autonomously (Lee et al., 2009).
One of the greatest challenges is to increase the
lifetime of sensor network due to adversarial
conditions of deployment environments that limits
battery power of sensor nodes, as they are not
rechargeable easily.
As of now, many routing algorithms, data
aggregation, location tracking, and clustering have
been proposed and most of these minimize energy
consumption due to delay in routing; optimize the
nodes to go to sleep mode and thereby increasing the
network lifetime. The nodes in a network will
consume more power if a single node dies that
determines the lifetime of a network (Obaidat and
Misra, 2014, Jurdak et al., 2010).
Metrics like energy consumption, and latency,
routing issues are being directly affected by the
number of hops and distance of hops. If the hop
count is too large, the energy consumption can be
minimized by adopting multi-hop transmission, but
at the cost of increase in end-to-end delay and
overhead. If the hop count is small (in case of direct
transmission), the latency and end-to-end delay will
be very small and hence the energy consumption is
high at the source node as the hoping is avoided.
Having said this, both methods should exist in a
network to increase the lifetime of the sensor
network (Anastasi et al., 2009). An optimal schedule
should be formulated to optimize energy, thereby
increasing the lifetime of the network of dynamic
and static topology.
This paper addresses this issue to handle both the
multi-hop and direct transmissions. Based on the
hop count number and hop distance, the source node
either bypasses intermediate nodes or uses direct
transmission. Since either of the techniques is
selected at runtime, the lifetime of the network will
be increased at least by 130%. A strategy is used to
PradeepKumar, T., Krishna, P., Obaidat, M., Saritha, V. and Hsiao, K.
Energy Optimisation using Distance and Hop-based Transmission (DHBT) in Wireless Sensor Networks - Scheme and Simulation Analysis.
DOI: 10.5220/0006483400170023
In Proceedings of the 7th International Conference on Simulation and Modeling Methodologies, Technologies and Applications (SIMULTECH 2017), pages 17-23
ISBN: 978-989-758-265-3
Copyright © 2017 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
17
find the Hop count threshold and distance of the
hops in the case of multi-hop transmission.
The objective of this paper is to increase the
lifetime of the sensor network via distance based
transmission. The contributions of this paper are
listed below:
1. The source node determines all the possible
paths along with their corresponding hop count
and distance before starting the transmission.
2. If the destination is within a threshold distance
or minimum hop count threshold, then the
source node bypasses the hops and sends the
information to destination node by slightly
increasing its power. With this, the lifetime of
the network is increased as source node uses
minimal power rather than awakening the
intermediate nodes. This method will not be
effective large-scale deployments of sensor
nodes with huge number of intermediate nodes
exist between source and destination nodes.
The rest of the paper is organized as follows.
Section 2 presents the previous related research and
introduces the problem statement. In Section 3,
relevant network, traffic and energy models are
presented. Performance evaluation using simulation
analysis and comparison are given in Section 5 and
Section 6 concludes the paper.
2 RELATED WORK
The researchers have proposed many methods to
reduce the energy utilization of node in sensors.
They have taken solar power and combined many
routing protocols to increase the life span of battery
and this technique is named as Enhancement of
energy conservation technology (EECT). This EECT
method is used in large organization to monitor the
air conditioners, which use heat energy and sunlight.
So, heat energy is utilized to make the environment
pollutant free and the battery usage is reduced
(Thayananthan and Alzranhi, 2014). The
performance is also affected by latency, energy and
reliability. Data aggregation approach is used to
reduce the redundancy and cost. ACO (Ant Colony
Optimization) is used to reduce the energy
utilization and also help to improve the life of sensor
node. This is only used when the channel is secured
(Ye and Mohamadian, 2014). The MEB (Minimum
Energy Broadcast) is used to reduce the
consumption of energy. POS based hybrid algorithm
and local search is used in this method. To decrease
the transmission power, POS is applied. The r-shrink
method is modified to re-establish the network again
(Hsiao et al., 2013).
The other methods using POS is either non-linear
programming or linear programming in order to
minimize the battery usage (Akyildiz et al., 2002;
Zeng and Pei, 2009). Development of routing
protocol is done using multi-objective fitness
function and encoding. These all use clustering
algorithm to minimize the energy consumption.
Physical to network layer optimization is done to
increase lifespan of network and reduce battery
consumption. The important applications and its
requirements are gathered to overcome this problem
(Pazzi and Boukerche, 2008).
Clustering is used to improve scalability, lifetime of
network and to reduce traffic load (Alla et al., 2012).
Due to extra load of receiving and transmitting data
to base station, cluster heads utilize more energy.
When overhead is more, cluster heads are destroyed
and stop working. Thus, this will reduce the overall
performance of a network. To overcome this
problem, Differential Evaluation based algorithm is
used which will reduce the load and enhance the
lifetime of network. This will also increase the
efficiency of data transfer from source sensor node
to its destination (Kuila and Jana, 2014).
These all are the techniques, which were used to
optimize the energy utilization during transferring
data from a particular source sensor node to the
destination sensor node (Akkaya and Younis, 2005;
Beyme and Leung, 2014; Saleem et al., 2012). These
techniques when applied to any protocol in the
sensor network will increase the lifespan of our
network and the battery lifetime. But it is not
optimized more than 20%. Many other methods are
there on which researchers work to reduce more and
more power consumption during the data transfer in
sensor networks.
Jeong-Hun Lee designs a network through an energy
efficient protocol based on the resource constraints
available between the BS and the source node (Lee
and Moon, 2014).
3 RADIO MODEL AND
PROBLEM STATEMENT
The energy consumption model of a typical sensor
network is shown in Figure 1. Energy is being
consumed by the radio of the sending and receiving
node through its processor, transceiver and the
amplifier. Amplifier consumes energy only when
SIMULTECH 2017 - 7th International Conference on Simulation and Modeling Methodologies, Technologies and Applications
18
Figure 1: Energy Model of Typical Sensor network.
transmitting a packet while the radio consumes
power during the reception of a packet.
The aim of this paper is to design an algorithm
that uses distance and hop as a metric that optimizes
the energy based on the dynamic selection of any of
the metric during the data transmission and also to
reduce the battery consumption, which increases the
lifetime of the sensor nodes.
In the proposed algorithm, DHBT, the distance
and hop are two metrics that are selected
dynamically during each data transmission. If a
source node decides to send a packet or data to a
destination node, the source node computes the path
to the destination node. The number of hops in each
path, the distance along each path and the residual
energy of all the nodes within the path are
determined. Firstly, the source node selects the path
that requires minimum energy. Based on the hop
count and the path length the packet transmission
between the source and destination nodes happens
either through multi-hop transmission or direct
transmission. The Hop count threshold will be
computed based on an algorithm, which also
depends on the network size and the topology.
3.1 Assumptions
1. The nodes are static or stationary
2. All the nodes are homogenous (same energy
level while transmitting or receiving a packet)
3. The distance between the nodes is not
constant.
4. Nodes are placed on a plane (X and Y axis)
5. Power usage during sleep mode is negligible
6. A boundary is considered while evaluating
the sensor energy model. (This model may
not work for a larger boundary with huge
distance constraints).
7. Minimal number of intermediate nodes (δ)
Table 1: Variables Used.
Variables and their purpose
Pi
Paths, i ε 1….m, n<m
E
T
Energy Consumption while transmitting a
message, Joules
E
R
Energy consumed while receiving a
message, joules
E
F
Energy consumed while forwarding a
message, joules
N
Number of Hops
N
t
Hop count threshold
D
Minimum distance threshold by which the
source node send info to destination node
directly, m
E
e
Energy consumed by the radio, joules
L
Message length
d
0
Distance threshold, m
D[n]
Distance matrix
C[n]
Energy Matrix
E[n][m]
Stores all the paths and its optimal energy
value
3.2 DHBT Algorithm
Initialize number of nodes → Sn.
Identify two corner nodes of the
topology and compute all the paths
between the two nodes
fori=1 to n
Find all path→[A]
D[i] → distance(Pm)
Discard N = 1
Compute (fix) Nt, L
Return(Pm)
If N >Nt: Goto HOP
Elseif N <Nt: Transmit with N=1
Elseif N<Nt and distance > d
0;
Goto
ENERGY
end for
HOP: fori=1 to m
C[i]→hop(Pm) //find number of hop
in all paths and store in array C
Return Pk
end for
ENERGY: fori=1 to m
for j=I to m
E[i][1] →min(C[ i])
E[i][2] →min(D[i])
Return(E[i][j])
end for
endfor
The algorithm explains the computation of energy of
the path, hop count and paths between the corner
nodes. Then, the average path length L or the
characteristic path length of a graph G, denoted L, is
Energy Optimisation using Distance and Hop-based Transmission (DHBT) in Wireless Sensor Networks - Scheme and Simulation Analysis
19
the average distance between vertices, where the
average is taken over all pairs of distinct vertices. In
any graph of order n, there are |E(Kn)| distinct pairs
of vertices. So, for a graph G of order n,
)(2,1
)2,1(
)1(
2
GVvv
vvd
nn
L
(1)
Complexity analysis:
a. The problem belongs to NP-Hard class to
compute the output of this problem as it takes
infinite time complexity. Mean time
complexity is O (n^n).
b. This NP-hard problem is reduced to NP
complete problem, which can be solvable in real
time.
c. Brute force attack is used to compute the path as
all the paths between a source and destination
has to be computed. Also, the intermediate
nodes should be minimal, otherwise it will take
infinite time to compute optimal path and the
algorithm becomes complex.
4 MATHEMATICAL MODEL
The energy consumed during the transmission and
reception of packets is:
0
4
0
2
....
...
ddif,dεk+Ek
d<ifd,dεk+Ek
me
fe
(2)
E
R
= k . E
e
(3)
E
F
= E
R
+E
T
(4)
For 1 bit message the total energy consumed by the
N hop is given as:
N
=i
α
i
me
dε+)E(=E(N)
1
.12N
(5)
If N <N
t
, there will be a direct transmission, in that
case the equation boils down to:
(6)
Since the E (N) here depends on the distanced,
which is to the power of 4, it proves that when the
distance is huge, the energy occupation will also be
more. Here in this case, the intermediate nodes are
not transmitting anything and they may be in sleep
mode as the source node decides to use direct
transmission.
However, DHBT algorithm also uses distance as one
of the metrics in reducing the energy level.
If N <N
t
and if the distance is greater than the
threshold distance, then normal multi hop
transmission will be occurring, thereby minimizes
the energy that will be wasted in direct transmission.
4.1 Analysis of Graph Model
Let the network shown in Figure 2 is as a graph G.
The figure shows only one edge, but the graph we
consider is a multi-graph (A loopless graph that has
multiple edges between two nodes). Here, we model
this as multi-graph as our algorithm proposes three
metrics: distance, hop and energy. In each edge, hop,
energy and distance will be considered as weight.
The multi-graph will be solved based on the
algorithm and is shown in Table. 2
Let G = {V, E, r} where:
V is the vertices
E is the number of edges
r:E; Assigning to each edge an unordered pair
of end nodes
Figure 2: Path taken by DHBT.
Let Aij will be the adjacency matrix and Iij will be
the incident matrix to find whether the given nodes
are incident on the edges or not.
Iij = 1 if Ej is incident with Vi; else 0.
Let Eij represents the weighted matrix for Energy,
Dij represents the weighted matrix for distance and
Hij be the weighted matrix for Hop count.
ENMENEN
MEEE
MEEE
Eij
...21
.............
2...2221
1...1211
SIMULTECH 2017 - 7th International Conference on Simulation and Modeling Methodologies, Technologies and Applications
20
DNMDNDN
MDDD
MDDD
Dij
...21
.............
2...2221
1...1211
2. Find path using brute force from matrix Dij
All path -> [Dij]
check all the combinations v0-v1, v0-v2,…… ,
v0-v1,……..,v0-v1-v2-….v10.
3. Find path having shortest path using maximum
[Eij]
Min path-> [Eij]
4. Take the path having minmum distance and min
energy.
5 IMPLEMENTATION AND
RESULTS
The network is tested using Network Simulator 2
(NS2) and mannasim where the simulations were
carried out. Here are the parameters that were used
during the simulation as shown in Table 2. For
comparison of DHBT protocol, the nearer
resemblence protocol is DSR, so DSR is taken for
comparison. DSR also computes all the paths
between the source and destination. The path taken
by the DHBT and DSR protocol is shown in Figures
2 and 3, respectively.
Table 2: Simulation Parameters.
Parameter
Description
MAC layer
802.11
Protocol
DSR, DHBT, LEACH
Propogation method
TwoRayGround
Transport Agents
TCP with 1500 bytes
Number of nodes
11,50,100
(LEACH protocol handles
the Cluster head
automatically, whereas DSR
and DHBT handles clutser
head separately)
Topology
Random with X=776 and
Y=612)
Tranmission Power
18mW
Forwarding Power
17mW
Receving Power
19.7mW
Idle Power
5 µA/0.005mW
Sleep Power
2µA/0.002mW (Negligible)
Sensed parameters
Temperature and pressure
Data dissemination interval
0.1 seconds
Sensing type
On Demand/Programmed
Figure 3: Path taken by DSR Protocol.
Experimental Setup:
The nodes were deployed as shown in the Fig 3 with
x and y locations so that for all the protocols under
comparison we will have the same topology. Since a
simulator is used for validating the protocol, the
sensing happens through Gaussian distribution.
Pressure and temperature sensors were modelled and
sensed.
Energy Consumption:
In DHBT, the path taken by the protocol is 0->1->9-
>10, whereas the DSR handles it 0->5->7->9->10.
So DHBT is optimising the power consumed at least
by one hop and hence the energy occupied by DHBT
is less compared to DSR. The number of nodes is
tested from small to huge nodes (11, 50,100). In
most of the cases the energy compared by the DHBT
is lower by at least 25% compared to DSR protocol.
Since DHBT directly computes all the paths related
to hop, distance and energy, AODV is not preferred
for comparison as AODV selects the path based on
the network topology on an on-demand basis.
Throughput:
When we take the case of transmission of data from
source to destination, then the average rate of
successful transmission is taken as throughput.
Hence, we have analysed throughput of DSR and
DHBT protocols. Throughput of DHBT protocol is
comparatively high as compared to DSR.
Route Selection:
DSR protocol considers hop count to find the path
for forwarding packets and also takes another path to
send the acknowledgment of the packet received.
Energy Optimisation using Distance and Hop-based Transmission (DHBT) in Wireless Sensor Networks - Scheme and Simulation Analysis
21
Figure 4: Throughput vs. time.
But DHBT protocol will first consider the paths
having less number of hops then will consider the
distance and energy. And the acknowledgment of
the data received is also sent by the same path.
Therefore, DHBT is more effective if we consider
all the conditions as compared to DSR protocol.
Figure 4 shows this difference between them.
Figure 5: Average packet delivery Ratio vs. number of
cluster heads.
The packet delivery ratio, and the lifetime of the
network are shown in Figures 5 and Figure 6,
respectively. Both are normalized and the PDR is
relatively constant for various bandwidth increases.
This predicts the stability of the network during the
increase in the bandwidth. Moreover, the lifetime of
the network is good for LEACH and DHBT, but
DHBT has slightly 10% more lifetime over LEACH
and network with DSR lifetime is too low compared
to other protocols.
Figures 7 and 8 show the energy consumed and
dissipated by DSR and DHBT protocols for 50
nodes by varying the number of cluster heads.
Figure 9 shows the dissipation of energy for a long
time by DHBT whereas the energy depleted fast by
the other protocols.
Figure 6: Network Lifetime vs. nuber of cluster heads.
Figure 7: Energy consumption vs. number of cluster
heads.
Figure 8: Energy dissipation vs. time.
6 CONCLUSIONS
In this paper, we derived a new algorithm called
DHBT by which the distance between any source
and destination nodes is used by the source nodes to
decide whether transmission of the message to
destination node must be multi-hop or direct
SIMULTECH 2017 - 7th International Conference on Simulation and Modeling Methodologies, Technologies and Applications
22
transmission by simply boosting the node power.
DHBT scheme has reduced the overall energy
utilization for each transfer of data in a sensor
network. Energy and hop count is working well with
DHBT whereas the distance calculation depends on
the transmitter and the receiver, so this work does
not handles distance calculation. However, distance
can be accurately calculated in the future work.
Also, distance can be computed using a localization
algorithm for sensor networks and thus the nearest
location of the sensor node could be found out and
can be solved for energy calculation. Simulation
analysis was used to predict the performance of our
proposed schema and to compare its performance
with competing schemes. We found out that DHBT
has excellent performance. As future work, we plan
to conduct more simulation experiments on DHBT
under different scenarios in order to check further
the performance under different conditions and
environments.
REFERENCES
Anastasi, G., Conti, M., Di Francesco, M., and Passarella,
A., 2009. Energy conservation in wireless sensor
networks: A survey. Ad hoc networks, 7(3), pp. 537-
568.
Akkaya, K. and Younis, M., 2005. A survey on routing
protocols for wireless sensor networks. Ad hoc
networks, 3(3), pp.325-349.
Akyildiz, I.F., Su, W., Sankarasubramaniam, Y. and
Cayirci, E., 2002. A survey on sensor networks. IEEE
Communications magazine, 40(8), pp.102-114.
Alla, S.B., Ezzati, A. and Mohsen, A., 2012. Hierarchical
adaptive balanced routing protocol for energy
efficiency in heterogeneous wireless sensor networks.
INTECH Open Access Publisher.
Beyme, S. and Leung, C., 2014. A stochastic process
model of the hop count distribution in wireless sensor
networks. Ad Hoc Networks, 17, pp.60-70.
Hsiao, P.C., Chiang, T.C. and Fu, L.C., 2013. Static and
dynamic minimum energy broadcast problem in
wireless ad-hoc networks: A PSO-based approach and
analysis. Applied Soft Computing, 13(12), pp.4786-
4801.
Jurdak, R., Ruzzelli, A.G. and O'Hare, G.M., 2010. Radio
sleep mode optimization in wireless sensor
networks. IEEE Transactions on Mobile
Computing, 9(7), pp.955-968.
Kuila, P. and Jana, P.K., 2014. A novel differential
evolution based clustering algorithm for wireless
sensor networks. Applied soft computing, 25, pp.414-
425.
Lee, J.H. and Moon, I., 2014. Modeling and optimization
of energy efficient routing in wireless sensor
networks. Applied Mathematical Modelling, 38(7),
pp.2280-2289.
Lee, S., Choi, J., Na, J. and Kim, C.K., 2009. Analysis of
dynamic low power listening schemes in wireless
sensor networks. IEEE Communications
Letters, 13(1), pp.43-45.
Obaidat, M. S. and Misra, S., 2014. Principles of Wireless
Sensor Networks. Cambridge University Press, 2014.
Pazzi, R.W. and Boukerche, A., 2008. Mobile data
collector strategy for delay-sensitive applications over
wireless sensor networks. Computer
Communications, 31(5), pp.1028-1039.
Saleem, M., Ullah, I. and Farooq, M., 2012. BeeSensor:
An energy-efficient and scalable routing protocol for
wireless sensor networks. Information Sciences, 200,
pp.38-56.
Thayananthan, V. and Alzranhi, A., 2014. Enhancement of
energy conservation technologies in wireless sensor
network. Procedia Computer Science, 34, pp.79-86.
Ye, Z. and Mohamadian, H., 2014. Adaptive clustering
based dynamic routing of wireless sensor networks via
generalized ant colony optimization. IERI
Procedia, 10, pp.2-10.
Zeng Xian-quan and Pei Hong-wen, 2009. Modern and
Design of Adaptive Mobile Middleware Supporting
Scalability, International journal of computer
applications, 29(9), pp. 24-31.
Energy Optimisation using Distance and Hop-based Transmission (DHBT) in Wireless Sensor Networks - Scheme and Simulation Analysis
23