Research On The Key Technologies Of Energy Harvesting Wireless Sensor Networks

As one of the important information collection technologies in the 21 st century,Wireless Sensor Network(WSN) technology provides the basic support for human to obtain various target information in the physical world.The traditional WSN is limited by the battery powered node's endurance,so it is difficult to run effectively for a long time,which limits the wide application of WSN.In recent years,researchers have proposed Energy Harvesting Wireless Sensor Network(EHWSN) technology to solve the problems caused by energy limitation in traditional WSN.The sensors in EHWSN can harvest the wireless,thermal,optical and other energy in the surrounding space,realize self charging and long-term work thus eliminating the limitation of battery supply.Compared with the traditional WSN,EHWSN has the advantages of long network life,low maintenance cost and less environmental pollution.It has been applied in environmental condition monitoring,industrial automation control,smart home and other fields,showing a broad application prospect.However,EHWSN technology was born in a short time and is not mature at present.There are some key technologies that need to be further studied and broken through.In particular,the existing communication protocols have the problems of high deployment cost,poor comprehensive performance and incompatibility with backscatter communication technology,which hinder the application and popularization of EHWSN.Therefore,exploring the design mechanism and implementation method of high-quality EHWSN communication protocol has important scientific significance and engineering application requirements.Aiming at the technical problems of EHWSN communication protocol,based on the in-depth study of mechanisms and technologies such as random relay strategy,meta-heuristic algorithm,fuzzy logic,backscatter communication and probabilistic polling,this thesis proposes a low-cost fixed clustering protocol based on random relay strategy,two high-performance dynamic clustering protocols based on meta-heuristic algorithm and fuzzy logic separately,and a high performance MAC protocol for EHWSN via Backscattering(BEHWSN).Theoretical analysis and experimental results show that the proposed protocols can solve the defects of existing protocols,and have broad application prospects.The main research contents and contributions of this thesis are as follows:1.Based on the research of random relay strategy and optimal inter-cluster communication,a new fixed clustering protocol RRCEH(Random Relay based Clustering Protocol for Energy Harvesting Wireless Sensor Networks),which is based on the random relay strategy,is proposed to solve the existing problems of low transmission efficiency,high cluster head energy consumption and expensive protocol deployment cost,which current fixed clustering protocols will suffer when working in EHWSN scenario.RRCEH adopts the random relay matrix to control the data uploading of cluster heads.For different cluster head configuration scenarios,based on the precise modeling of energy consumption of the cluster head,through optimizing the random relay matrix,RRCEH can reduce total demand of the energy harvesting capacity of cluster heads,and thus effectively save the deployment cost of the protocol.The performance verification results show that when the network area number and data compression rate are relatively high,in the two cluster head configuration scenarios,RRCEH can reduce the needed energy harvesting capacity of cluster heads by more than 30% on the basis of existing protocols.2.Based on the research of meta-heuristic algorithm and fuzzy logic technology,to solve the existing dynamic clustering protocols' problems of low data delivery rate,large performance diversities between sensors,and high sleep rate of cluster heads when working in EHWSN,this thesis proposes a centralized meta-heuristic based dynamic clustering protocol MACEH(Meta-Heuristic Approach based Clustering Protocol for Eenrgy Harvesting Wirless Sensor Networks) and a distributed fuzzy logic based dynamic clustering protocol FLCEH(Fuzzy Logic based Clustering Protocol for Energy Harvesting Wireless Sensor Networks).In order to make full use of the energy collected by the network and balance the performance between sensors with different energy harvesting rates,both MACEH and FLCEH take the remaining energy and energy harvesting rate as the important reference factors when clustering,and use harmony search algorithm and fuzzy logic technology respectively to make the clustering decision based on global and local information.In order to ensure the rationality of inter-cluster routing,MACEH and FLCEH use the Max-Min Ant Colony Algorithm and a greedy algorithm based on location and energy state separately to help cluster heads to carry out path planning.On the basis of reducing the energy consumption of the inter-cluster communication,cluster heads with higher energy states are made to undertake more relay tasks,so as to make full use of the remaining energy and energy harvesting capacity of the relevant cluster heads.Meanwhile,MACHE and FLCEH use the in-round routing update strategy and the cluster head relay strategy respectively to deal with the cluster dormancy caused by the exhaustion of the cluster heads' energy,and maintain the energy efficiency of inter-cluster communication.The simulation results show that MACEH and FLCEH can improve the data delivery ratio by more than 30% on the basis of their respective reference protocols under high load scenarios.The throughput difference coefficients of sensors with different energy harvesting rates in the two protocols are only 0.0236 and 0.0565.In addition,both protocols can effectively reduce the probability of cluster head dormancy and ensure the monitoring quality of the coverage area.3.Based on researching the mechanism of backscattering communication and the multi-rate probabilistic polling,to solve the problem that the existing MAC protocol is not compatible with the backscatter communication technology,this thesis proposes a MAC protocol designed for BEHWSN named Multi-rate Polling.Multi-rate polling is based on probabilistic polling.The data center broadcastings the related polling command to lead the data upload of sensors,which solves the problem that the backscatter node can not support complex functions such as carrier sense and idle signal detection.Multi-rate polling allocates different uplink data rates for nodes according to the channel characteristics of nodes and the data center,so that the ability of backscattering channel can be fully utilized to improve the network throughput on the basis of ensuring the successful data reception by the data center.At the same time,Multi-rate Polling can control the channel access priority of nodes with different data rates through parameter adjustment.Therefore,it can realize various levels of the throughput and fairness trade-off according to the QoS(Quality of Service)demand of different applications.The simulation results show that Multi-rate Polling can effectively cover all sensors in the network and has good scalability.In specific scenarios,Multi-rate Polling can improve the throughput by 104.8% under the same fairness index and can improve the fairness by 118% under the same throughput index compared with the reference protocol,showing good comprehensive performance.