Research On The Energy Optimization Of WSN Based On Improved EEUC Protocol

In the research of wireless sensor networks,clustering routing protocol is an important branch.LEACH protocol proposed in 2000 has become the basis of many researches.As an important means to solve the energy hole problem,heterogeneous clustering routing protocol can greatly extend the life cycle of the network and has a broad application prospect.Based on the traditional uneven clustering EEUC protocol,this dissertation proposes corresponding optimization methods in the three stages of cluster head election,data fusion and routing transmission,which are used to reduce network energy consumption,prolong network life cycle and improve the accuracy of data transmission.The main work of this dissertation is as follows:(1)In the cluster head election stage,the EEUC protocol does not consider the residual energy of cluster head nodes and the uneven distribution of cluster heads.On the basis of dividing the cluster radius according to the distance between cluster heads and sink nodes,the energy weight and node density of nodes are added to make the energy consumption of all nodes more uniform,reduce the number of nodes dying in advance,and improve the overall life of the network.(2)In the stage of data fusion,it is found that better parameters can be obtained by using neural network to establish input-output model and training.In this dissertation,the structure and characteristics of BP neural network are analyzed,and the number of hidden layer nodes is determined by empirical formula and trial method.In gradient descent,the combination of accelerated gradient descent algorithm(rmsprop)and gradient descent method with momentum(Adam ladder)is used Degree reduction optimization algorithm improves the efficiency of training,and the required error accuracy can be achieved after less training.(3)In the route transmission stage,the combination of genetic algorithm and simulated annealing algorithm is applied to the route selection of multi hop route.In the experiment,it is found that genetic algorithm can jump out of local extremum to search the global optimum better,while simulated annealing algorithm has higher efficiency in the optimization of local extremum.Therefore,this dissertation combines genetic algorithm and simulated annealing algorithm to improve the route selection Selected optimization efficiency.In this dissertation,on the MATLAB simulation platform,compared with other existing algorithms,the results show that this algorithm can significantly improve the life cycle of the network,reduce the redundancy of information transmission,and improve the accuracy of data fusion.