无线传感网中一种智能数据融合算法的实现及仿真分析

Implementation and Simulation Analysis of an Intelligent Data Fusion Algorithm in Wireless Sensor Network

在无线传感网中,传感器节点一般都由自身装配的电池供电,难以进行电量补充,因此节约电量对于无线传感网来说至关重要。为了提高无线传感网能量使用效率,延长网络生存时间,提出了一种结合遗传算法和粒子群算法优化BP神经网络的智能数据融合算法GAPSOBP(BP Neural Network Data Fusion algorithm optimized by Genetic algorithm and Particle swarm)。GAPSOBP算法将无线传感网的节点类比为BP神经网络中的神经元,通过神经网络提取无线传感网采集的感知数据并结合分簇路由对收集的传感数据进行融合处理,从而大幅减少发往汇聚节点的网络数据量。仿真结果表明,与经典LEACH算法和PSOBP算法相比,GAPSOBP算法能有效减少网络通信量,节约节点能量,显著延长网络生存时间。

Sensor nodes in wireless sensor network（WSN）are usually powered by their own assembled batteries which are difficult to recharge,therefore,saving energy is crucial for WSN.In order to improve the energy utilization of wireless sensor network and prolong the survival time of the network,a data fusion algorithm based on BP neural network named GAPSOBP is proposed which is optimized by genetic algorithm and particle swarm optimization algorithm.In GAPSOBP algorithm,sensor nodes are analogy to BP neural network neurons.The sensing data collected by sensor nodes is extracted by BP neural network,and then combined with clustering routing to fuse data for the purpose of reducing the amount of data sent to the Sink node.Simulation results show that compared with the classical LEACH algorithm and PSOBP algorithm,GAPSOBP algorithm can effectively reduce network communication traffic,save node energy consumption and prolong the network lifetime.