Research On Routing Selection Algorithms For Sensor Networks Based On Neural Network And Simulated Annealing

Wireless sensor network is composed of a large number of sensor nodes deployed in the monitoring area and forms a self-organizing network through wireless communication.The sensor node reports the data collected by the sensor to the user management node through the base station through the routing protocol.Because the energy of sensor nodes is limited and cannot be easily supplemented,energy saving becomes a hot topic in routing protocol research.Based on the existing clustering routing protocols,this paper applies neural network and simulated annealing algorithm to the routing algorithm,proposes a routing algorithm of uniform clustering and balanced network load,and verifies its simulation.Firstly,in the clustering phase of the network,the hierarchy is divided according to the distance from the base station and the optimal number of cluster heads is calculated.In the self-organizing clustering stage,a self-organizing feature neural network based uniform clustering algorithm is proposed,which takes the position information of sensor nodes and the azimuth of the sensor node and the base station as the parameters of the input vector to uniformly cluster the network.The member nodes in the cluster send the collected data to the cluster-head nodes,and the cluster-head nodes conduct data fusion.This algorithm can effectively avoid isolated nodes,reduce the energy consumption of the network and improve the number of remaining nodes in the network.Secondly,a cluster head selection algorithm based on competitive neural network is proposed.This algorithm takes the relative residual energy of nodes,the number of neighbor nodes of nodes and the number of selected cluster heads as the parameters of the input vector,and selects the optimal cluster head for each cluster,which avoids the randomness of cluster head selection and takes full account of the residual energy of cluster head nodes,facilitating the routing and transmission between clusters and reducing the data packet loss rate.Finally,a link selection algorithm based on simulated annealing is proposed in the phase of network data transmission.The algorithm divides the network cluster-head nodes into three regions,and links each region with simulated annealing algorithm.In the process of data transmission,the cluster-head nodes send the collected data to their respective cluster heads,and the inter-cluster transmission is transmitted along the link from the farthest cluster-head node to the base station.Partition can effectively reduce the simulated annealing algorithm into the local optimal solution,and reduce the network transmission delay,simulated annealing algorithm can find the optimal path in routing selection,greatly reduce the energy consumption of data transmission,balance the network load,and extend the life of the network.