Research On Intelligent Routing Algorithm Based On Energy Prediction In Wireless Sensor Networks

In recent years,Energy harvesting wireless sensor networks(EH-WSNs)are widely used in various fields.Unlike battery powered wireless sensor networks,EH-WSNs can harvest energy from the environment to power sensor nodes.Therefore,WSNs with energy harvesting have a long-term sustainable energy base.The abundant energy supply changes the energy model of wireless sensor nodes,making wireless sensor networks can spend more energy to obtain stronger communication ability.At the same time,the goal of node state control is also changed from prolonging the lifetime of node to avoiding premature death of node.In order to increase the working time of awake node and reduce the sleeping time of node,the method of node state control is proposed.These changes need a new design of routing protocol to further improve the communication ability and energy management ability of wireless sensor networks.Major studies of this paper are the energy dissipation model and the joint optimization of information transmission and energy utilization in EH-WSNs.Firstly,aiming at the prediction of energy harvesting in EH-WSNs,two methods with different complexity and prediction accuracy are proposed to predict solar energy harvesting.The prediction model based on long-short term memory(LSTM)and the traditional prediction model based on solar periodic and trend change law are verified with the actual solar radiation data,which proves the effectiveness and difference of the prediction.Secondly,a new energy aware opportunistic routing protocol(EAOR)is proposed according to the energy harvesting prediction method.Compared with traditional methods,EAOR has proved its excellent performance in EH-WSNs.Through experiments with different precision prediction methods,it is proved that the high precision prediction method can promote the performance of the routing protocol.In the considered EH-WSNs model,the imbalance of energy consumption and collection limits the network's sensing and communication capabilities.In order to solve this problem,this paper applies opportunistic routing protocol(OR)to EH-WSNs to make full use of the abundant energy provided by energy harvesting,convert the energy into spatial diversity gain,and improve the communication ability of the network.At the same time,through the research of the existing works,there are still some deficiencies in the prediction of energy harvesting and node energy modeling in EH-WSNs.What's more,there is no research to use the prediction data of energy harvesting to improve the performance of opportunistic routing protocol.Therefore,opportunistic routing in EH-WSNs is reduced to an optimization problem aiming at maximizing the energy efficiency of a single hop.To solve this problem,this paperproposes an energy aware opportunistic routing protocol based on long-short term memory network.The protocol innovatively considers the current residual energy of nodes and the short-term solar energy harvesting predicted by LSTM neural network as the key factors in the selection process of opportunistic routing and forwarding candidate nodes,so as to improve energy utilization and balance energy storage.Therefore,in order to jointly optimize the energy consumption and information transmission of nodes,a new metric considering the energy factor and relay forwarding ability is proposed to assist the selection of opportunistic forwarding candidate nodes.In this protocol,the residual energy and sleeping history of the node are considered in relay priority.Simulation results show that,compared with the traditional opportunistic routing based on geography,the proposed protocol improves the network throughput by 20%and reduces the retransmission rate by 11%.At the same time,it has a strong ability of energy balance between nodes.In order to verify the impact of the prediction method on the simulation results,this paper also designs a prediction method with lower prediction accuracy instead of LSTM neural network for simulation,which proves that more accurate energy harvesting prediction can improve the performance of the proposed routing protocol,and verifies the necessity of using high-precision prediction method.This paper describes the prediction model,system model and design details of energy aware opportunistic routing protocol.Simulation shows that the proposed prediction method has higher prediction accuracy.Compared with the traditional opportunistic routing protocol,the proposed routing protocol has obvious advantages in improving the communication ability and energy management ability of the network.