| Research on energy balance of linear wireless sensor networks used in coal mine, there are three types of nodes in network, sensor nodes, relay nodes and sink node. Sensor nodes are randomly deployed across the monitored area, relay nodes are deployed on one side of the network and sink node is deployed at the end of the network. Sensor nodes send data directly to relay node, then relay node fuses data and forwards data to another relay node or sink node. The distance between different sensor nodes and relay nodes is different,it leads to the energy consumption imbalance of sensor nodes. It requires proper deployment of relay nodes’ position that balancing the energy consumption of sensor nodes and prolong the survival period of network. Particle swarm optimization is suitable for searching the global best value in solution space. Therefore, we use particle swarm optimization to find a reasonable relay node deployment scheme. However, the basic particle swarm optimization convergence accuracy is poor, it is necessary to improve the algorithm. To solve this problem, we propose an iterative multi-stage PSO and chaos ring PSO. Iterative multi-stage PSO process is divided into two steps: first, improving the inertia weight in basic PSO and an adaptive inertia weight is proposed; second, dividing the network into a plurality of regions and then using iterative multi-stage PSO to optimize each incremental region. Chaotic map can make particles’ search process random and traversal, improving the accuracy of algorithm; local ring topology can accelerate the convergence speed. By integrating chaotic map and local ring topology, chaos ring PSO algorithm has been proposed. Through simulation and analysis, we confirmed that iterative multi-stage PSO and chaos ring PSO are effective in sensor nodes energy consumption balancing optimization. And confirmed chaos ring PSO is better than iterative multi-stage PSO in the algorithm running time and relay nodes deployment optimization. |
PDF Full Text Request