Application of Data Fusion Based on Genetic Algorithm and BP Neural Network in WSN

摘要

Energy-saving is one of the inevitable problems of the routing design in WSN, while Data Fusion technology is widely utilized in energy constraint WSN to reduce the amount of messages exchanged between sensor nodes. This paper proposes a new algorithm based on Integrated Genetic and BP Neural Network(IGBP), IGBP uses the global search capability of GA to remedy the deficiency of BP artificial neural network. First, IGBP generates the best individuals in different networks by GA algorithm. Then it chooses the most optimize individual measure by Mean Squared Error to construct the BP network which was supplied to train of the WSN. Using the optimize individual nodes as initialization value training the BP network, it will enhance the learning rates of convergence and avoid falling into the local minimums.The simulation results show that the IGBP algorithm has made great progress in balancing the consumption of energy so as to prolong the network lifetime.