Research And Simulation Of Optimized Deployment In Wsn

With the development of Wireless Sensor Network(WSN) technologies, more and more WSN technologies are applied to the Smart Home, Smart Transportation and so on. WSN is a kind of ad hoc network, which is composed of sensors that have the ability of communication and computation in the manner of self-organizing or multi-hop. Nowadays, the research work about the WSN is focus on network technologies and communication protocols. However, there is little research about the optimized deployment strategies of WSN.To deploy some sensors randomly in the farmland or in the forest, the regular way is to deploy by planes. However, it may form some blindness and a large number of redundant sensors. Therefore, how to achieve maximum network coverage scale with the minimum number of sensors is becoming an urgent problem in the WSN deployment. A large amount of sensors are required in the broad farmland or in the large areas. However, most of the sensors are charged by battery which is disposable and energy limited. Therefore, how to balance the energy of WSN has become another problem in the WSN deployment.Consequently, In order to resolve these problems, the main work of this thesis is to optimize the deployment strategy of WSN. The research achievements are given as follows:Firstly, we study the representative deployment algorithms of WSNin the literature. The primary study is on the Particle Swarm Optimization algorithm and Ring-shaped algorithm based deployment strategy of WSN. The limitations of each algorithm are analyzed.Secondly, to speed the computing time of the KMPSO algorithm and improve the network coverage scale, we propose an optimized KMPSO algorithm which is named as TKMPSO. The problem of optimization of deployment can be regard as a problem of Combinatorial Optimization. Combined with K-Means, the TKMPSO algorithm can split the particle swarm in a reasonable manner. We can compute the precise position of each sensor with the TKMPSO algorithm, and it can be deployed to the target.Thirdly, in the circumstances of first deployment of WSN, to reduce the number of deployed sensors, we propose a cost-effective strategy based on computational geometry deployment strategy. The optimized strategy is derived from the thought of greedy algorigthm and Voronoi polygon. With the greedy thought, the next sensor can be deployed to the proper position quickly. With the Voronoi, sensors can be evenly deployed to the area, and what is more important, the Voronoi diagram can be reused by the ring-shaped algorithm. In the circumstances of redeployment of WSN, to resolve the position collision of the high-energy sensors, we propose an improved energy-effective deployment strategy. The proposed strategy is able to decide whether the ring area has high energy sensors, so as to avoid to deploy new sensor again at the area with high energy sensors. Thus, the number of sensors is reduced and the lifecycle of the network is extended.Finally, The Matlab and NS2 simulation tools are adopted to demonstrate the effectiveness of the optimized algorithms.