Research On Mobile Charging Strategy In Wireless Rechargeable Sensor Networks

Wireless sensor networks(WSN)is an important part of the Internet of Things project.It senses and transmits information by deploying a large number of micro-sensor nodes in the monitoring area.However,the limited battery capacity of sensor nodes is a key problem that restricts the long-term operation of WSN.In recent years,the introduction of mobile charger(MC)in WSN provides a new idea for solving the energy problem of WSN.This paper analyzes the advantages of online charging planning and models the charging planning problem for various scenarios composed of different number of MC deployments and whether MC assists sensors to collect data,and obtains different charging planning schemes.The main contributions of this thesis are as follows :(1)Aiming at the problem of charging planning for single MC with limited energy,a charging planning model is established on the basis of previous studies.Aiming at maximizing the sensor survival rate and MC energy utilization rate,a charging node selection algorithm based on attention mechanism(OCAM)and a charging algorithm based on fuzzy logic and Q-learning(FLQL)are proposed.OCAM adopts the attention mechanism to adaptively adjust the weight value of the parameters to select the charging node.FLQL quantifies the charging request by constructing a fuzzy logic system,and uses the Q-learning algorithm to plan the charging path of the MC.Simulation results show that OCAM algorithm and FLQL algorithm have good network performance.(2)In view of the limitations of traditional charging planning in improving sensor survival rate,the feasibility of joint data collection protocol is analyzed.In order to maximize sensor survival rate,a joint data collection algorithm based on FLQL(ADC FLQL)is proposed.The algorithm introduces the data collection protocol into FLQL.MC collects part of the sensing data of the sensor during the charging process to reduce the energy consumption of the sensor,and determines the data cache of the sensor through the fuzzy logic system.In order to verify the effectiveness of the designed algorithm,it is compared with the existing typical joint data collection charging planning scheme.Experiments show that ADC FLQL algorithm is superior to WcAMDG and VN-MOAC algorithms in improving sensor survival rate and MC energy utilization.(3)Aiming at the rapidly increasing charging demand of large-scale WRSN,a multi-MC cooperative charging model is established to maximize the sensor survival rate and the energy utilization rate of MC.Based on OCAM and FLQL,a distributed cooperative charging algorithm based on OCAM(OCAMD)and a cooperative charging algorithm based on FLQL(FLQLC)are proposed.OCAMD integrates a collaborative strategy to ensure that MCs and charging request nodes correspond one-to-one,while balancing the charging load of multiple MCs.FLQLC achieves coordinated charging of multiple MCs by constructing a multi-agent-based Q-learning algorithm.Finally,the feasibility of OCAMD algorithm and FLQLC algorithm in improving sensor survival rate and MC energy utilization is verified by simulation experiments.