Data Collecting Design And Performance Optimization In Wireless Powered Sensor Network

Wireless powered sensor network(WPSN)utilizes wireless power transfer(WPT)technique to provide stable energy supply for sensor nodes,which enables WPSN to observe physical quantities in real-time for human activities.Recently,Internet of Things(IoT)has been widely used for smart city and industry 4.0,which requires huge numbers of sensor nodes to observe interesting physical quantities.It brings a great challenge for WPSNs to collect the huge amounts of observed data from distributed nodes timely due to the low spectrum efficiency of data collecting schemes.Moreover,the energy transfer and data transmission in WPSN are influenced by each other,which brings more difficulties for data collecting.Based on above discussion,the design of data collecting scheme in WPSN has attracted an extensive attention from academia and industry.This dissertation focuses on designing schemes for WPSNs to collect huge amounts of observed data from distributed sensor nodes effectively.This work mainly contains three parts,which are the adaptive data collecting scheme in WPSN and the performance analysis for asynchronous multi-cluster WPSN,the data collecting scheme based on compressed sensing(CS),and the data collecting scheme for WPSNs observing multiple physical quantities.The main contributions of this dissertation are given as follows.(1)First,we studied the data collecting scheme for WPSNs observing one physical quantity by integrating energy transfer and data transmission.Considering the cou-pling relation between energy transfer and data transmission,with energy beamforming based on time division multiple access(TDMA)scheme,we propose an adaptive data collecting scheme.For the best performance of data collecting,we study an optimization problem of the time allocation of energy transfer and data transmission and derive the closed-form optimal solution.To evaluate the performance of WPSN with optimal solution,under the non-line of sight(NLOS)and line of sight(LOS)energy channels,we derive the outage probability and the diversity order for nodes.Furthermore,we study the interference problem of energy and data stream for asynchronous multi-cluster WPSN and analyze its performance,where sensor nodes are randomly deployed to observe multiple physical quantities.With the random geometry,we utilize Poisson Point Process(PPP)to model the location of node clusters and derive the outage probability to show performance.It reveals that the relation of communication performance to the transfer power and the cluster density.Finally,by Monte-Carlo numerical simulation,the proposed scheme is proved to be effective and the validity of performance analysis for asynchronous multi-cluster WPSN is also verified.(2)In the situation of WPSN observing one single physical quantity,we utilize the sparsity of observed data among distributed nodes to achieve an efficient and real-time observation.First of all,we exploit the coherent multiple access channel(Co-MAC)as the CS measurement matrix and achieve the CS measurement at the fusion center(FC)by coherent transmission,which accelerates data collecting by avoiding the CS coding of distributed nodes.With the derived data collecting rate,we compare the proposed scheme with other schemes,which indicates its superior performance.To maximize the data collecting rate,under given requirements of recovery error,we optimize the time allocation of energy transfer and data transmission by iterative algorithm.However,the recovery performance of the proposed scheme rapidly degrades in large-scale WPSNs.To overcome this shortage,we design a dynamic clustering and collaborative data collecting scheme based on CS.For optimal recovery performance,we study the optimization problem of power allocation and node clustering,and provide a closed-form suboptimal solution.Finally,we provide the Monte-Carlo numerical simulation to demonstrate the superior performance of the proposed schemes.(3)To observe multiple physical quantities,there are two node deployment strategies for WPSNs,which are concertized and distributed node deployments.In the WPSN based on concertized node deployment,we design an efficient data collecting scheme based on sparse observation and coding,where all nodes are required to observe and encode physical quantities adaptively.Since nodes observe multiple physical quantities,the power consumption of nodes increases rapidly,which brings a great challenge.To cope with this challenge,we study a joint optimization problem of observation matrix and coding matrix to minimize network power consumption,under the given recovery performance requirements.In the WPSN based on distributed node deployment,we exploit the joint sparsity of observed data and propose an efficient data collecting scheme based on joint CS,where different node clusters cooperatively achieve joint CS measurement at the FC by coherent transmissions.To achieve the optimal performance,we study the joint optimization of the time allocation of node clusters and the CS measurement allocation of common-component to maximize rate and provide a closed-form optimal solution.Finally,we provide the Monte-Carlo numerical simulation to demonstrate the superior performance of the proposed schemes.