Research On Energy Efficiency Technologies Of Wireless Sensor Network

With the rapid development of the new fields such as Internet of Things,CyberPhysical system and so on,the demanding of information acquisition is increasingly.Hence the Wireless Sensor Networks(WSNs)are becoming widely used.Due to the limited resources of the computing power,battery,and communication capacity of sensor nodes in a large scale,it is a challenge to utilize the energy of sensor nodes efficiently.Aiming at the problems of short lifetime of network,low energy utilization and large amount of communication in the application of WSNs,this paper has a research on energy efficiency technologies of WSN and proposes a series of solutions:1.To overcome the shortcomings of clustering routing algorithm in WSN,such as short lifetime of network and unbalanced communication load,an efficient routing algorithm called K-means++ Fuzzy Routing Algorithm(KFRA)is proposed,which clusters nodes based on K-means++ clustering algorithm,with a designed fuzzy logic system for distributed cluster head selection.The simulation results illustrate that,compared with the existing low consumption WSN clustering routing algorithm,the load between nodes is balanced and the lifetime of network is prolonged greatly.2.Genetic Adaptive Energy-Balanced Routing Protocol(GAEBRP)is proposed which can be used in different size or deployment of network sizes and for different network optimization objective.With the proposed protocol,sink node uses a designed genetic algorithm to obtain the optimal fuzzy rules for a specific network or a specific optimization objective.Finally,the optimal fuzzy rules are deployed to each node.Our GAEBRP has better generality for application without setting the fuzzy rules manually than other methods.The simulation results show that the network lifetime of GAEBRP proposed in this paper is longer than other low consumption WSN clustering routing protocols under different network sizes.3.A strategy of Sensing Compresssed and online Reconstruction while learning(SCo Rel)is proposed,aiming at the problem of large amount of data and high energy consumption when WSNs are used for data acquisition.Principal Component Analysis(PCA)and Compressed Sensing(CS)are introduced.The strategy can dynamically adjust the relevant parameters for different types of collected data,and obtain accurate global data with less communication cost when the statistical information of target data type is unknown.4.An alternating diffusion strategy(AD-LMS)is proposed for distributed parameter estimation in WSN in order to further reduce the communication load of WSNs.The operation steps and flow charts of the strategy are illustrated.The simulation results show that the AD-LMS strategy can greatly reduce the network traffic with little loss of network estimation performance.5.A Compressed Combined Reconstruction Adaptive(CCRA)strategy for sparse parameter estimation is proposed,in view of the fact that the unknown parameters estimated in most applications are sparse.The operation steps of the strategy are shown,and the convergence of the CCRA strategy is proved.At the same time,when the target parameters are sparse,CCRA strategy reduces much more communication load than that in other low-load strategies under the condition that the performance of network estimation is not affected at all.