Particle swarm optimization using elite opposition-based learning and application in wireless sensor network

Standard particle swarm optimization algorithm is easy to fall into local optimum and has slow convergence velocity and low convergence precision in the late evolutionary, so the paper proposes a new approach, called particle swarm optimization using adaptive elite opposition-based learning. Elite particle is chosen randomly to adopt opposition-based learning at each iteration and learning dimension of space decreases linearly with the evolution, which increases the ability of exploration and exploitation ability. This approach is helpful to jump out the local optimum and improve the probability of the optimal solution. The experiments' results show that our algorithm is much better at not only the convergence velocity but also the accuracy of solutions and the ability to escape from local optimum. And it is applied into coverage optimization of wireless sensor network verify its practicability, and the simulation results showed that the proposed approach obviously increases in WSN coverage. Copyright © 2014 American Scientific Publishers.