An Efficient Variable Dimension PSO Algorithm for Mobile Node Tour Planning in WSN

Recent advances in unmanned aerial vehicle (UAV) and wireless charging technologies have made it practical to introduce a mobile node into Wireless Sensor Networks (WSN). This mobile node with sink ability or charging equipment on it can travel around the entire network to gather data from isolate sensor nodes or charge the nodes wirelessly. This can help with the large scale WSN to be connected or keep working for a long time without replacing batteries. For the mobile node, routing design is an important problem. We can regard this problem as a special case of traveling salesman problem with neighborhoods (TSPN), which is NP-hard. In this paper, we propose a novel routing design algorithm base on Variable Dimensions Particle Swarm Optimization (VD-PSO). In this algorithm, every particle is a feasible solution of TSPN. Every dimension of the particle is the coordinates of a rendezvous point (the point where mobile node stay to gather data or charge the nodes). And the dimensionality is right the number of the rendezvous point. With the evolutionary method of the particles, we can derive the best route of the mobile node. Simulation results show that our scheme has fast convergence speed, and the result is quite approximate to the optimal solution.