Optimal Wireless Sensor Networks Deployment in 3-Dimensional Terrains using Hybrid Population Based Algorithm
No Thumbnail Available
Date
2014-02
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Addis Ababa University
Abstract
Wireless sensor networks (WSNs) are composed of cooperating sensor nodes that can perceive the
environment to monitor physical phenomena and events of interest. Sensor deployment is a
fundamental issue in a WSNs to maximize coverage and quality of service with limited number of
sensor nodes. In order to maximize area coverage, sensors need to be placed in a position such that
the sensing capability of the network reach at high quality. Coverage is one of the main problems
in WSNs deployment. Previous research works on sensor deployment mainly focused on Two
Dimensional (2D) plane or in Three Dimensional (3D) volume coverage. But now, these studies
on sensor deployment extended to 3D surfaces or terrain, to achieve the highest overall sensing
quality. In our thesis, we worked to develop an optimal WSNs deployment on 3D surfaces to
maximize area coverage under constrained number of nodes.
Researchers have used different methods and algorithms to make sensor deployment. Populationbased
optimization algorithms find near-optimal solutions to the difficult optimization inspired by
natural probabilistic evolution. In our research work, Genetic Algorithm (GA) and Particle Swarm
Optimization (PSO) are selected to form Hybrid Algorithm (HA) to find optimal locations of
sensors based on a fitness function. We have selected the two algorithms to exploit the best features
of the algorithms in combination.
We have used two typical surfaces, rough and smooth, to compare the results of the GA, PSO and
HA in the optimal deployment of sensors. The fitness function used in the algorithms is calculated
based on coverage of all sensors in the region of interest (ROI). A simulating program for both
surface types and all the three algorithms has been developed using MATLAB.
In all the three PSO, GA and HA evaluations, we found that the HA has exceeded PSO with a
percentage of 26.12% and GA with 1.58% on rough surface. Similarly when we also compared
the results of these algorithms on smooth surface, HA has exceeded PSO with a percentage of
22.24% and GA with 3.42%. Next to the HA, GA has a very good performance than PSO.
Key words: Particle Swarm Optimization, Genetic Algorithm, WSNs, 3-D terrain,
Sensor Deployment.
Description
Keywords
Particle Swarm Optimization, Genetic Algorithm, WSNs, 3-D terrain, Sensor Deployment