Research On WSN Clustering Routing Algorithm Based On Deep Learning

Wireless Sensor Network(WSN)plays an irreplaceable role in the current information age.Micro sensor nodes are deployed in the monitoring area.Due to their limited energy,the nodes will lose their function due to the exhaustion of energy,which will have a great impact on the network operation.Nodes consume a lot of energy in wireless communication modules,so improving routing protocols is an effective way to reduce node energy consumption.Therefore,while ensuring the normal completion of the network work,the network life cycle can only be extended by reducing the energy consumption of nodes.The classic LEACH protocol uses the idea of clustering,and nodes take turns serving as cluster heads,which saves energy to a certain extent.Therefore,this paper improves on the basis of the LEACH protocol.The main research contents of this paper are as follows:(1)Aiming at the problem of uneven clustering in the LEACH protocol network clustering,this paper uses a clustering algorithm to evenly cluster the network,avoiding the occurrence of large clusters and extremely small clusters in the network.(2)LEACH selects the cluster head by generating random numbers without considering the energy of the node and the distance from the base station.Once a low-energy node or a node far from the base station is selected as the cluster head again,the death rate of the node is accelerated,affect the normal operation of the network.Aiming at the above problems,a competitive neural network is introduced in the selection of cluster heads,considering factors such as node energy and distance from the base station,and following the principle of winner being king,the competitive neural network model is used to compete to select the best cluster head.(3)After the cluster head is selected in the LEACH protocol,the cluster head allocates time slots for data transmission to the nodes in the cluster.If the nodes do not collect data,the network time slot will be wasted.A polling mechanism is introduced here to efficiently utilize the network bandwidth.The cluster head is responsible for the fusion task of data within the cluster,but no specific data fusion method is mentioned.This paper introduces a deep learning model.Using the layer-by-layer greedy method to train the deep autoencoder to extract the features of the data,fuse the data,reduce the data traffic in the communication process,and save the network energy.(4)In the LEACH protocol,the cluster head transmits data to the base station through singlehop communication.The cluster head farther away from the base station consumes more energy.This paper will transmit data to the base station by a combination of single-hop and multi-hop communication.Finally,a simulation experiment is done on the MATLAB platform,and the performance of the algorithm in this paper and the three comparison algorithms are compared according to the three indicators of the network.The simulation results show that the algorithm in this paper effectively balances the energy of the network and extends the network survival time.