Energy Prediction And Clustering Algorithm In Energy Harvesting Wireless Sensor Network

Wireless Sensor network(WSN) nodes exist serious energy constraint problems as its power supply by the limited capacity of the battery. Energy harvesting wireless sensor networks(EH-WSN) using energy harvesting techniques to collect available energy of environment for the power supply network and relieve the problem of limited sensor node energy. Due to the acquisition energy is vulnerable affected by random environmental change, has obvious daily and seasonal change, as well as the discontinuity and uncertainty. How to make the network work steady and consistently on the condition of collecting energy in variation, at the same time, maximizing performance is the main problem of EH-WSN research. Under the background of above mention, research on energy prediction and energy balance of heterogeneous clustering algorithm, and design a network performance optimization method. Main work includes:1) Energy prediction algorithm based on adaptive filter. The algorithm by recording the historical data of environmental energy harvesting of sensor nodes and using s-order adaptive filter to predict the future energy. Normalized least square root square of the error between the predicted values and the observed values as a parameter, update the filter coefficient adaptively, follow environment change in time, so as to realize high precision in energy prediction. The simulation results show that in comparison with the traditional algorithm, this algorithm is in a lower integrated resource usage, getting a higher prediction precision. 1)2) Distance and energy based heterogeneous cluster hierarchy algorithm(DEHCH) of energy balance based on EH-WSN. Aiming at the shortcomings of the original LEACH protocol, on three factors of the residual energy, the energy acquisition opportunity of node and the distance of the node to the base station cluster heads, cluster selection algorithm was improved. Contribute the closer distance to the base station, the greater the energy acquisition opportunity and residual energy of sensor node to become cluster head with a higher probability in relative. Simulation results show that DEHCH protocol can improve the load balanced ability of network and the performance of EH-WSN effectively.3) Considering the fixed network performance cannot accommodate the insufficient of energy harvesting of environment change, proposed an optimization algorithm of the adaptive environment energy following the network performance. Enable network to follow environmental energy available and network node average remaining energy state, real-time adjust the network performance optimization factor adaptively. Simulation results show that the network in different energy harvest environment and in the guarantee of a certain active rate of network nodes, the algorithm can automatically adjust the network data communication speed rate, and to achieve higher data throughput.