Research On Particle Swarm Optimization-based Node Positioning Algorithm For Wireless Sensor Networks

With the increasingly mature day by day and rapid development of semiconductor technology, sensor technology, wireless communication technology, embedded computing technology and micro-electromechanical systems, the multifunctional micro-sensor with low cost and low power can be a lot of production. Wireless sensor networks come into being.Wireless sensor networks are mainly used in some specific areas for data collection and processing. But these regions are often with poor environment, where the human can not arrive in, Moreover, wireless sensor nodes are often with small size, light and energy-limited in particular. So how to reduce energy consumption and prolong the network life cycle of wireless sensor networks has been the focus of our study.For some applications, the data of no location information is meaningless. As the special applications of wireless sensor networks, specific location information of some sensor nodes can not be gained, because they are usually scattered randomly by the air on the specific region. So for these unknown nodes, the current approach to solve these problems are useing some specific algorithm to locate the unknown nodes by the limited known locations of the nodes.A wireless sensor network localization algorithm based on Particle Swarm Optimization is proposed in this thesis to solve the problem of inaccurate positioning and large energy consumption for wireless sensor network node positioning.The algorithm combines the particle swarm optimization algorithm (PSO) and node localization algorithm to improve the positioning accuracy. Some energy is consumed in the iterative process of PSO algorithm, but in this thesis, the fitness function is relatively simple. However, the energy consumption of the new method is smaller than that using the least squares method to TDOA ranging manner positioning.The algorithm in this thesis, we use auto-adapted PSO to adjust the inertia factor w's value linearly, take the distance between unknown node and beacon node as the fitness function, and which can reduce the computation load. In this thesis, we use the TDOA range manner to compute the distance between the unknown node and beacon node, and then obtain a nonlinear simultaneous equation, then optimize the algorithm using an improvement Particle Swarm Optimization to iterative process step by step, finally obtain the optimal solution which can meet a certain precision or arriving the iterative number of times. Namely, it's the coordinate of the position of the unknown node. The simulation results on NS2 indicated:When the error of range increases, it is more obvious to enhance the tendency of the algorithm's positioning accuracy;With the same error of positioning, it's need less number of beacon nodes; In the case of the same number of nodes, positioning accuracy is improved.