A Wireless Sensor Network Topology Design Method Based On An Evolutionary Algorithm Featured With Population Crossover Strategy

Recent technological improvements have made the deployment of small,expensive, low-power, distributed devices, which are capable of local processingand wireless communication, a reality. Such nodes are called as sensor nodes. Eachsensor node is capable of only a limited amount of processing. But whencoordinated with the information from a large number of other nodes, they have theability to measure a given physical environment in great detail. Thus, a sensornetwork can be described as a collection of sensor nodes which co-ordinate toperform some specific action. Unlike traditional networks, sensor networks dependon dense deployment and co-ordination to carry out their tasks.In this paper, author use an method based on evolutionary algorithm (EA) tosolve a wireless sensor network optimization problems. By clustering a sensornetwork into a number of independent clusters using CEA (Crossover-basedEvolutionary Algorithm)and NEA(Negotiable Evolutionary Algorithm), author cangreatly minimize the total communication distance, thus prolonging the networklifetime and the coverage. As the master nodes’ communication distance is limitedand unbalance-load, author should find the optimal path to the master nodes whichcan not be directly connected to the central node.To solve the above problem, author first establish a mathematical model ofWSN and presented a method based on an evolutionary algotithm featured withpopulation crossover strategy, and introduced negotiable evolutionary algorithm inWSN for the first time.Simulation results show that our algorithm can quickly find a good solution.This approach is also applicable to multiple network topologies (uniform ornon-uniform) or shortest distance optimization problems.