A Proposed Load Balance Approach to Improve Data Transmissions

in Wireless Sensor Networks

A PHD proposal

Presented by

YAQEEN JASIM ABD

Outline:

I. Introduction
II. Challenges In WSN
III. Problem Statement
IV. Aim
V. Network Model.
VI . localization methods
VII. anchor node
VIII. clustering model
IX. load balance
X. metaheuristic technique
XI. Related Work
XII. Research Methodology
XIII. References
I. Introduction

Wireless Sensor Networks (WSNs) have many applications in various life domains.
WSNs find these applications due to the advances in wireless communication
technologies, lowcost, low-power consumption and light weight. Each sensor
represents a multi-functional device. sensors are self-governing entities that
autonomously collaborate to sense, collect the data, process the information, and
transmit the collected data to the base station (BS). BS can process the data received
from the sensors and produces an action for the environment. WSNs are usually
composed of many sensors that work on small built in batteries. So, WSNs must be
reasonably deployed to meet their crucial applications requirements while saving
energy.

Wireless Sensor Networks (WSN) are characterized by their resource-constrained

nature. To mitigate the prime concern of developing an energy-efficient network,
clustering of the nodes has emerged as a very effective tool. Clustering is a technique
to balance load of network. Queuing theory can also have utilized. Clustering can not
only help in obtaining even load distribution among the network nodes but also help
in having the enhanced network lifetime and scalability. The power utilization has
verified as a major problem in wireless sensor networks. There are many conflicting
factors that affect the efficiency of clustering, i.e. distance between CH and the base
station, distance form node
to cluster head (CH), maximum residual energy of CHs, etc. The coordination among
these factors has a capability to insure the optimal power utilization by reducing
power consumption and load balancing among the nodes and CHs. A developed
mathematical decision-making method can be proposed to choose the best set of CHs
from the available alternatives that can fulfill the condition of coordination efficiently.

This text discusses the localization process in wireless sensor networks (WSN) and
the two main types of localization methodologies: centralized and distributed. It also
covers various techniques such as ToA, TDoA, AoA, and RSSI that are used to
estimate the distance between nodes. The localization issues are categorized into
range-based and range-free localization, and the role of GPS-enabled nodes in the
process is explored, along with their advantages and disadvantages. Additionally, the
text examines how factors such as path loss, noise, propagation, device
measurements, connectivity, power control, and tracking can affect measurements in
localization.
II. Challenges in WSNs
- Power Consumption: the major problem in sensor networks is the Power
management. It is essential to design a power aware algorithms and protocols for
WSN .
- Production cost: the sensor can be considered as disposable devices, due to the
several deployment models. The sensor networks could possibly contend with
traditional information gathering methods only if the specific sensor node could be
produced economically. The target price intended for a sensor node should preferably
be very low in price. In order to make a WSN feasible, the cost of a sensor should be
reduced. Due to this the cost of a sensor will be a very challenging issue.
- Computational power and memory size: Every sensor stores the information
separately and sometimes more than one node saved same information and transferred
it to the sink that wastes the power and storing capacity of nodes . So effective
schemes are necessary to decrease the redundancy in the WSN.
- Operating environment: Sensor nodes may work in a very challenging
environment. These nodes may work at the bottom of an ocean, in a home or large
building, vehicle traffic management etc.

III. Problem Statement

IV. Aim

V. Network Model : assumes the wireless sensor network with the following
characteristics:
(1) All the sensor nodes are deployed randomly across the sensing field and are static.
More illustratively, nodes once deployed cannot change their location
(2) The sensor nodes are homogeneous and equipped with a definite amount of initial
energy
(3) The sensor nodes are facilitated with the power control features to introduce
variations in the transmission power as and when needed
(4) The base station is also static and can be placed at any point in the network
accordingly
(5) The continuous data flow model is used here to define the working mode of the
sensor nodes

VI . localization methods

Localization is a crucial aspect of transmitting data across a network of sensor nodes.

It involves positioning the nodes, which typically consist of anchor nodes, displaced
nodes, and a central server, using radio frequency signals for communication. Anchor
nodes are equipped with GPS and can be powered by batteries or external sources,
with recharge circuits and motion detection sensors to detect movement unless they
are in sleep mode. Localization has various applications in fields such as location-
based services, robotics, cellular networks, and healthcare. Meta-heuristic localization
algorithms aim to reduce discrepancies in the localization process.
Absolute localization involves a group of nodes, referred to as anchors, that have
knowledge of their exact locations. The remaining nodes are then located based on the
anchors' positions. GPS-based techniques are commonly used for absolute
localization, with a GPS receiver-equipped sensor being utilized for this purpose. A
small subset of nodes with GPS receivers serve as reference beacons and establish an
absolute coordinate system. By applying a linear transformation to the relative
coordinate system, coordinates can be obtained using reference nodes in the absolute
coordinate system. The Genetic Algorithm is employed to classify the Direction-of-
Arrival into phases such as initialization, fitness evaluation, selection, cross over,
mutation, and optimization. This approach reduces computational load and improves
location accuracy.

Figer 1 Classification of localization techniques

Centralized Techniques. The central base station monitors and calculates the
distance between all nodes. After the calculation, the distance is forwarded back to the
nodes. The transmission of data in this process is responsible for latency, bandwidth
and energy usage. Centralized algorithms are more accurate than distributed networks,
since centralized algorithms have a global view of the entire network.

Distributed Techniques. In the distributed localization methods the sensor nodes

collect the measurements with different methods and determine the distance between
neighbors and anchor nodes. By interacting with each other the sensor nodes get their
own position in the network.

In range-free techniques, localization depends on the relationship between nodes

and topological information of sensor nodes. range-free localization methods include
1- centroid algorithm,
2- Monte Carlo localization,
3- approximate point in triangle test (APTT),
4- closest point-based method,
5- assumption- based coordinate method,
6- DV-HOP.
Centroid algorithm : This is one of the most straightforward range-free algorithms
because it only requires a minimal amount of computations and communication
expenses compared to other algorithms. In this algorithm, all unknown nodes
calculate their locations by finding the centroid of all received packets from the
beacon nodes within their communication range. The algorithm consists of three
stages:
 1) Beacon nodes send their locations to sensor nodes within their communication
range, while unknown nodes collect the received packets.
2) Unknown nodes determine their locations by calculating the centroid using the
beacon nodes' locations.
3) The centroid calculation is a simple operation that only uses the beacon nodes'
locations. The estimated location for the beacons.
in this figure is the centroid value.

Figer 2 show the centroid algorithm

The Approximate Point-in-Triangulation Algorithm (APIT) is an iterative

algorithm that uses beacon transmissions from anchor nodes to perform location
estimation. It employs an area-based approach by dividing the environment into
triangular regions between beaconing nodes. This allows a node's presence inside or
outside these regions to narrow down the area in which it can potentially be located.
By using combinations of anchor positions, the estimated area in which a node resides
can be reduced to provide accurate location estimates. The method used to narrow
down the probable location of a node is called the point-in-triangulation (PIT)
examination. In this method, a node selects three anchors from nearby anchors and
checks if it is inside the triangle formed by these three anchors. APIT repeats this PIT
examination with different anchor triplets until all combinations are exhausted or the
desired accuracy is achieved.
 Figure 3. APIT scenario for triangle formation

Figure 4. Performing point in triangulation test

Range-based localization is a technique used to determine the distance between

sensor nodes using various measurement metrics such as RSSI, ToA, and TDoA.

RSSI estimates the distance between two sensor nodes based on the strength of the
received signal. However, this method is limited by a decrease in signal strength and
poor data throughput at greater distances.
𝑑0
𝑅𝑆𝑆(𝑑0) = 𝑃𝑡 + 𝐾𝑒 − 10ŋ log ( ) + 𝛼 + 𝛽
𝑑1

ToA, also known as Flight Time, determines the transit time from a single transmitter
to a single receiver to estimate distance. Synchronizing two nodes is essential for
greater precision, but inaccurate clock synchronization can result in imprecise
positions.
TDoA detects the geographical location of radio frequency emitters and relies
heavily on reception quality and precise synchronization. This approach requires three
nodes to determine transmitter location and may be affected by multiple paths and
synchronization errors. Increasing the distance between nodes can improve accuracy
by improving the difference between arrival times.
Angle of Arrival which is a term used in signal processing and wireless
communication. It refers to the direction from which a radio wave or signal arrives at
a receiving antenna. By measuring the angle of arrival of a signal, it is possible to
determine the location of the transmitter. This technology is used in various
applications, such as radar systems, wireless communication networks, and navigation
systems.

VII. anchor node

An anchor node is a fixed or stationary node in a wireless sensor network (WSN) that
serves as a reference point for other nodes in the network. Anchor nodes are used to
provide location information and enable localization algorithms in WSNs. They are
typically equipped with higher transmission power and more accurate localization
hardware than regular nodes, allowing them to communicate with a larger number of
nodes and provide more precise location information. Anchor nodes play an important
role in many applications of WSNs, such as environmental monitoring, surveillance,
.and tracking of mobile objects

VIII. clustering model

clustering of the nodes has emerged as a very effective tool. If executed intelligently,
clustering can not only help in obtaining even load distribution among the network nodes but
also help in having the enhanced network lifetime and scalability. Clustering Technique in
wireless sensor networks formulates the energy-balanced clusters to improve the network
lifetime.

A large number of sensors make up WSN, but their battery power is limited.
Although WSN hubs can operate in harsh and dangerous environments, the battery
cannot be recharged or replaced in these situations. Therefore, it is crucial to conserve
energy in the network. To achieve this, Clustering Model used ,a special hub called
the cluster head (CH) and regular nodes (ON) are used. The CH has high energy and
is responsible for collecting data and transmitting it from other hubs to the base
station (BS) , As shown in the image below.
 Figure 5 show WSN clustering model

Clustering—Design Challenges

In a large portion of the WSN open-air applications in some critical situations, such as
those that are unattended, maintaining the battery is complicated. With this limitation,
it is very difficult to

extend the lifetime of the network. Alongside the previously mentioned drawbacks,
some different difficulties, which should be

satisfactorily addressed while designing clustering algorithms exist, are listed below :

1- The algorithm complexity increases linearly as the network develops.

2- CH selection is very important because it can directly affect the performance of the
system.

3- In most of the strategies, CH choice depends on a few parameters, for example,

energy level and hub location.

4- CH will receive the sensed data from the nodes on which the data aggregation
process is performed. That process is the reason that these steps form the key
structural challenge.

5- A clustering algorithm should handle all types of applications because WSNs are
completely application dependent.

6- The clustering algorithm design must support defense applications in which data
are highly confidential, for example, military applications and health monitoring.

7- While applying the clustering model in a larger network it becomes complicated

and energy consumption will be gradually increased.

8- Clustering algorithms are combined with many hybrid models to improve the
overall performance of the network. During that process interference of the network is
also gradually increased

The utilization of clustering in Wireless Sensor Networks (WSN) offers several

advantages :

Firstly, it allows for optimized management techniques to be implemented throughout

the network.

Secondly, clusters manage the topology and communication overhead of the network
from the sensor level, with only the Cluster Heads (CHs) acting as hubs

Finally, clustering ensures that communication bandwidth is maintained and prevents

the exchange of redundant messages.
IX. load balance
Load balancing in wireless sensor networks (WSN) is the process of distributing the
workload evenly among the sensor nodes in the network. This is done to ensure that
no single node is overloaded with too much data to process, which can lead to energy
depletion and node failure. Load balancing helps to improve the performance and
efficiency of the network, by reducing congestion and preventing nodes from running
out of energy too quickly. It also helps to extend the lifetime of the network, by
ensuring that each node is used optimally.

Clustering load balancing in wireless sensor networks (WSN) is a technique used to

distribute the workload among the sensor nodes in the network. In this technique, the
sensor nodes are grouped into clusters, and each cluster is assigned a cluster head. The
cluster head is responsible for collecting data from the member nodes and forwarding
it to the base station. This helps to reduce the energy consumption of individual nodes,
extend the network lifetime, and improve the overall efficiency of the network, As
shown in the image below .

Figure 6 show Clustering load balancing in wireless sensor networks (WSN)

X. metaheuristic technique
The metaheuristic approach commences by randomly selecting a set of solution
vectors that progressively improve with each iteration. Once the specific application
parameters, such as scaling factor and crossover rate, are established, the fitness of the
current solution set is determined using a well-designed fitness function. The iteration
counter is then initialized, and a selection is made from the population chosen, after
which the selected vectors undergo a variation phase involving mutation/crossover.
The updated vectors are then evaluated for their current fitness, and a survivor
function, utilizing a greedy selection strategy, is employed to finalize the population
for the next generation. This process of updating the set of solutions is repeated for a
predetermined number of iterations, and ultimately, the most recent population is
selected as the final solution. The use of an intelligently and carefully designed fitness
function is crucial in achieving further improved offspring in metaheuristic
techniques. A wide range of metaheuristic techniques, including genetic algorithm
(GA), genetic programming (GP), evolutionary programming (EP), evolution
strategies (ES), differential evolution (DE), particle swarm optimization (PSO), ant
colony optimization (ACO), and teaching-learning-based optimization (TLBO), can
be found in the literature.
The following illustration depicts the functioning of a metaheuristic approach .

Figure 7: General scheme of metaheuristic techniques.

Differential Evolution is a popular stochastic metaheuristic optimization technique

that is used for continuous search space. It involves defining initial parameters such as
scaling factor, crossover rate, and a set of initial solutions (population) along with the
number of iterations. Each solution vector, also known as a target vector, undergoes
mutation and recombination phases. The selection of the next-generation solutions is
performed after the generation of all trail vectors, unlike particle swarm optimization
and teaching-learning-based optimization. Once all target vectors are converted into
trial vectors, the greedy selection towards the next-generation solution is performed
between the pair of target and trial vectors. There are various mutation strategies
available such as random, best, and target-to-best, and two types of crossover
techniques- binomial and exponential crossovers. The binomial and exponential
crossover can be defined in specific ways :
1. Binomial Crossover

𝑉𝑗 𝑖𝑓 𝑟 ≤ 𝐶𝑝 𝑂𝑅 𝑗 = 𝛿,
𝑈𝑗 = {
𝑋𝑗 𝑖𝑓 𝑟 > 𝐶𝑝 𝐴𝑁𝐷 𝐽 ≠ 𝛿,

2. Exponential Crossover

𝑈𝑖 if fitness(Ui, G) ≥ fitness(Xi, G),

𝑋𝑖, 𝐺 + 1 = {
𝑋𝑖 otherwise.
The Genetic Algorithm is a metaheuristic based on the principles of natural selection
in biology. It is used for optimization purposes and involves a population of Np
candidates represented by chromosomes encoded as binary numbers with N bits. The
algorithm follows a specific procedure, including:

Initialization
Fitness𝑎
Selection 𝑃𝑎 = 𝑁𝑝
∑𝑏=1 Fitness𝑏

Crossover 𝐶𝑎 = ∑𝑁𝑝
𝑏=1 Fitness𝑏.

and mutation.

A chromosome structure is illustrated in Figure 7

Figure 7: Structure of a sample chromosome.

XI. Related Work

no Title Author and Problem Used Proposed
year statement equipment solution
1 Energy-Efficient Junfeng Chen , The article proposing a new The proposed
Clustering and Samson Hansen proposes a new approach to Energy-Efficient
Localization Sackey , Joseph approach to energy-efficient Clustering and
Technique Henry energy-efficient clustering and Localization
Using Anajemba , clustering and localization in Technique
Genetic Xuewu Zhang , localization in Wireless Sensor (ECGAL) uses
Algorithm in and Yurun He Wireless Sensor Networks residual energy,
Wireless Sensor 2021 Networks (WSN) using distance
Networks (WSN) using Genetic estimation, and
Genetic )Algorithm GA) coverage
Algorithm connection to
)GA) form the fitness
.function
2 Metaheuristic Sandip K. The problem The technique This research
Load- Chaurasiya,1 statement in this uses differential article proposes a
Balancing- Arindam Biswas paper is to evolution to Metaheuristic
Based ,2 Prasit Kumar formulate formulate Load-Balancing-
Clustering Bandyopadhyay, balanced energy-balanced Based Clustering
Technique in 3 clusters in clusters based Technique
Wireless Sensor Amit Banerjee wireless sensor on node (MLBCT) for
Networks ,4 and Rajib networks that proximity, wireless sensor
Banerjee5 achieve even distribution, and networks, which
2022 load distribution energy balances the
among the nodes distribution energy of the
across the nodes and ensures
sensing field even load
distribution
among the cluster
members
3 MLBC: Sukhchandan, The proposed Multi-Objective The proposed
Multi-objective Sushma Jain approach aims Optimization approach aims to
Load Balancing 2018 to minimize the (MOO) minimize the
Clustering average energy approach for average energy
technique in consumption clustering and consumption
Wireless while load balancing while maximizing
Sensor maximizing data in Wireless data delivery
Networks delivery Sensor reliability by
reliability by Networks minimizing the
minimizing the (WSNs) to inter-cluster
inter-cluster achieve energy .transmission cost
transmission efficiency and
cost reliable data
.delivery
4 the constraints The article discusses various
Localization in V. Sneha1, M. of WSN nodes mainly techniques and
Wireless Sensor Nagarajan2 and the impact discusses the methodologies
Networks: A 2020 of deployment different used in
Review location on data techniques and localization in
.transmission methodologies wireless sensor
used in networks
localization in
wireless sensor
networks
5 focuses on the examined the it provides an
An overview on Osamah localization of literature that overview of
wireless sensor Ibrahim nodes in order to already exists different
networks and Khalaf1, Bayan optimize their about wireless localization
finding optimal Mahdi Sabbar2 performance sensor networks techniques used in
location of 2019 and methods for wireless sensor
nodes determining networks
location )WSNs)
6 Optimization of Asha Jerlin Several multihop optimization of
Routing-Based Manuel 1 , technologies are routing in a routing-based
Clustering Ganesh Gopal employed to clustering clustering
Approaches in Deverajan 2 , improve environment approaches in
Wireless Sensor Rizwan Patan 3 previous and classifies wireless sensor
Network: and Amir H. research efforts, the approaches )networks (WSNs
Review and Gandomi which into parameter-
Open Research 2020 emphasized based,
Issues efficiency in optimization-
terms of cost based, and
and time, as well methodology-
as innovative based categories
.methods
7 Performance Ibrahim Nemer1 Localization, This paper In this text, we
evaluation of & Tarek which is crucial focuses solely examine various
range-free Sheltami1 & in all on localization range-free
localization Elhadi applications of techniques that methods of
algorithms Shakshuki2 & WSNs, refers to do not require a localization and
for wireless Abdulrahman the process of .specific range explore some
sensor networks Abu Elkhail1 & accurately applications
Mumin Adam1 identifying the where knowing
2020 positions of SNs the precise
.in the network location of SNs is
crucial and
requires careful
.consideration
8 A novel Elham Moridi The utilization In order to propose a new
hierarchical Majid of sensors in maintain their hierarchical fault
fault Haghparast hard-to-reach optimal management
management 2022 locations and performance in framework to
framework for their hardware the face of overcome the
wireless sensor constraints lead faults, networks limitations. The
networks: to the require a proposed method
HFMF occurrence of method that is complies with
faults and the tolerant to such clustering
inability to events. This can algorithms
replenish be achieved by
.energy enhancing
energy
efficiency and
implementing
effective fault
management
.techniques
9 A Metaheuristic Celestine Iwendi1, Recently In this work, to Our focus has
Optimization Praveen Kumar Internet of optimize the been on reducing
Reddy M2,*,
Approach for Thippa Reddy
Things (IoT) is energy the energy usage
Energy G3,*, Kuruva being used in consumption, of sensors within
Efficiency in the several fields most the IoT network,
IoT like smart appropriate which will result
 Networks Lakshmanna4 , Ali city,these Cluster Head in a longer
Kashif Bashir5,*, network (CH) is chosen lifespan for the
MD. Jalil Piran6
2021
maximazing the in the IoT .network
energy network The
consumption of suggested
sensors in the approach
IoT network that involves
will lead to an utilizing a
decrease in the combination of
network lifetime two meta-
heuristic
algorithms,
namely Whale
Optimization
Algorithm
(WOA) and
Simulated
Annealing (SA),
to identify the
most suitable
CH for IoT
network clusters
10 An artificial PONNAN, S., Artificial I improve the proposes a
intelligence SARAVANAN, ntelligence- coverage, quorum-based
based quorum A.K., IWENDI, based Q uorum s connectivity, grid system where
system C., IBEKE, E. ystems a re used latency, and the number
for the and to solve the lifespan of sensors in the
improvement of SRIVASTAVA, energy crisis in of the networks quorum is
the lifespan of 2021 real-time increased without
sensor wireless sensor actually
networks .networks increasing
quorums
themselves,
leading to
improvements in
throughput and
latency
11 A Differential Mohammed Sensor nodes a novel routing The objective of
Evolution Kaddi typically have protocol based this study is to
Based Zeyneb Khalili very limited on the present a new
Meryem Bouchra
Clustering and 2020
resources due to differential routing protocol
Routing the constraints evolution for WSN that
Protocol for of algorithm (DE) utilizes the
WSN miniaturization in WSN differential
and evolution
manufacturing algorithm (DE) to
cost. These reduce energy
limitations consumption and
include low prolong the
energy,
computing network's
capacity, data .lifespan
storage space,
and transmission
.rate
Research methodology

start

Application
environment

Parameter of WSN :

* area

*number and type of sensors node

*number and type of anchors node

Deploy sensors randomly

Deployment and utilize anchor nodes

Indicated BS by using some method

Use the methods of localization

IF anchor nod no
found information
about sensor node

Yes

Range-based Range-Free

AOA RSS TOA TDOA

approximate centroid
Monte Carl
point in algorithm
localization
triangle test
(APTT)

A
 A

Collect network location data

Clustering Approach

Select load balancing CH

implementation metaheuristic technique

GA DE

Generate initial Population

Population initialization and parameter initialization

Fitness evaluations Calculate objective functions

Crossover and mutation Search for the best solution

Selection Crossover and mutation

Calculate objective factions for offspring

and compare with parent
The maximum number
No
of iterations

Update best solution

yes

Output results
NO
stopping

Yes
End

stop