Fault-Tolerant and Minimum Communication Cost Wireless Sensor Networks Using Modified Genetic Algorithm

The performance of a network, including wireless sensor networks, is determined by various aspects, including the position and function of its nodes. The node placement methods, especially relay, are an interesting topic as they were the key to enhancing WSN communication ability. Relay placement is usually associated with energy efficiency, coverage area, connectivity, and communication costs. Dozens of recent surveys have even explored other aspects, including reliability, scalability, and security. As the optimization aspects become more complex, researchers consider that the multi-objective optimization (MOO) approach is the most promising solution. This research aims to implement a Multi-Objective Optimization (MOO) method of relay placement on a large scale to construct a WSN with minimum cost and fault tolerance ability. From the point of view of WSN, there are at least two novelties: the area modeling and the restriction of the hop number. This area modeling allows each sensor to connect to at least two different relays and construct multi-paths to the sink. We constructed this placement method over a modified genetic algorithm and combined it with the explorative character of swarm intelligence. In this research, some chromosomes of GA were decomposed as particles of PSO to extend the solution. The proposed method shows positive results that combine the good aspects of both algorithms in planning and designing a WSN.