Research On Optimization Design Of High-Performance Wireless Relay Communication Systems For The Internet Of Things

The Internet of Things(IoT)has been honored as the third revolutionary wave within the global information industries.The increasingly abundant IoT application scenarios produce new challenges for the development of wireless communication technology concerning the coverage capability of networks,the reliability of signal transmission,the efficient use of spectrum and energy resources,etc.In order to effectively alleviate the contradiction between the practical demand for the wide-area distribution of massive IoT devices and the limited propagation distance of wireless signals in the complicated communication environment,this dissertation studies two types of wireless relay devices,which can be used to extend the coverage range of the IoT.The former is the half-duplex(HD)amplify-and-forward(AF)multiple-input multiple-output(MIMO)relay transmission nodes that have fast signal processing speed and low hardware implementation complexity.The latter is the reconfigurable intelligent surfaces(RISs)that are able to carry out the full-duplex(FD)passive beamforming for the incident signals.The research of this dissertation regarding the optimization design of high-performance wireless relay communication systems for the IoT has mainly completed the following four aspects of innovative works.Firstly,this dissertation studies the problem on the joint optimization of the parameters within a two-hop HD AF MIMO relay system,which is constituted by multiple(information)source nodes,multiple parallel relay nodes,and an(information)destination node using the nonlinear minimal mean-squared error(MMSE)-decision feedback equalization(DFE)signal receiver.By following the mean-squared error(MSE)minimization design criterion and the block coordinate descent(BCD)method of Gauss-Seidel type,as well as taking advantage of the direct links,this study designs an iterative algorithm for the joint optimization of the precoding matrices at all the source nodes,the amplifying matrices at all the relay nodes,as well as the decision feed-forward and the decision feedback matrix at the destination node.This algorithm not only can support different numbers of parallel relay nodes to integrate their communication resources,such as energy,antennas,etc.,but also allows utilizing the signals from direct links,which have low power strength,to acquire extra diversity gains.Hence,it can be more flexible for the IoT systems to collect data from multiple source nodes through long-distance two-hop relay links.Simulation results show that,in comparison to the linear receiver-based algorithm,the new algorithm owns better MSE and bit-error-rate(BER)performance.In the meantime,as the power strength of the signals in direct links increases,the algorithm developed here can obtain further performance improvement.Secondly,this dissertation studies the problem for the jointly distributed optimization of the parameters within a multi-hop HD AF MIMO relay system,which is made up of multiple source nodes,multiple cascading relay transmission nodes,and a destination node with MMSE-DFE receiver.Through decomposing the amplifying matrix at each relay node,this study proves that those MSE matrices obtained by estimating the original signal vector at the destination node and each relay node,respectively,can present a cascading formula construction.Hence,the nonconvex objective optimization problem can be decomposed into several less difficult subproblems in the moderately high signal-to-noise ratio environment,thereby enabling the optimization of the precoding matrices at all the source nodes and the amplifying matrix at every relay node to be separated from each other.This study at first conducts the iterative optimization of the precoding matrices at all the source nodes and the decision feedback matrix at the destination node,and then derives a closed-form water-filling solution of the subproblem corresponding to the amplifying matrix at every relay node.Simulation results show that,compared with the other algorithms which were also designed by means of decomposing the original problems,the new algorithms possess superior MSE and BER performance.In addition,every relay node just needs to know the accumulated system information and the local channel state information(CSI)so as to complete the optimization of its amplifying matrix.By comparison to the centralized algorithms,such distributed algorithms split the quantity of calculations among all the relay nodes and the destination node,thereby relieving the calculating pressure on a single IoT device,and meanwhile,they avoid a part of overheads for the exchange of control information among the IoT devices,which is generated from the processes of gathering the CSI and delivering the optimized system parameters in the centralized algorithms.With the new algorithms developed here,a multi-hop relay system is able to quickly carry out the very long-distance data backhaul for multiple source nodes.Thirdly,this dissertation studies the problem concerning the joint optimization of the parameters in an analogue network coding(ANC)-based two-way communication HD AF MIMO relay system,which has multiple parallel relay nodes and two source/destination nodes that transmit signals to and receive signals from each other and use the MMSE-DFE receivers.Hereafter,these source/destination nodes will be directly called the source nodes.Under the imperfect CSI,this study has designed robust algorithms for the jointly iterative optimization of the precoding matrices adopted by two source nodes when they transmit signals,the amplifying matrices at all the relay nodes,and the permutation matrices for the detection orders of all the data streams as well as the decision feed-forward and the decision feedback matrices,which are employed by two source nodes when they receive signals.Via the ANC technique,the two-way relay transmission processes,which usually require four time slots,now just need two time slots,thereby reducing the consumption of communication resources in the long-distance data exchange.For mitigating the phenomenon of error propagation within the nonlinear receivers,this study also optimizes the detection orders of all the data streams.Besides,for the possible situation where the practical IoT devices can not obtain sufficiently accurate CSI,the robust system optimization algorithms are capable of handling the uncertainties of CSI,to a certain extent.Simulation results show that,compared with the linear receiverbased existing algorithm,the nonlinear receiver-based new algorithms possess superior MSE and BER performance,and meanwhile,they can present good robustness in terms of dealing with the imperfect CSI.Fourthly,this dissertation studies the problem regarding the joint optimization of the parameters in an uplink communication MIMO relay system,which consists of multiple source nodes,multiple RISs that act as FD relay devices,and a destination node that utilizes the linear MMSE or the nonlinear MMSE-DFE receiver.With the "smart radio environment" constructed by multiple RISs,this system can be applied to achieve the uplink data transmission of multiple IoT devices.In circumstances where the signal transmission power of each source node is restricted,this study designs the iterative algorithms for the joint optimization of the precoding matrices at all the source nodes,all the continuous or discrete phase shift parameters of every RIS,and the linear receiving matrix or the decision feed-forward and the decision feedback matrix at the destination node.Here,those massive low-cost adjustable elements that constitute each of the RISs can generate continuous or discrete phase shifts for their incident signals.The case of continuous phase shifts presents the upper bound of system performance,while the case of discrete ones is closer to the actual hardware conditions.Simulation results show that,when there is severe signal attenuation in the direct links between all the source nodes and the destination node,through introducing the indirect "source-RIS-destination" links,both the MSE and the BER performance of the MIMO systems can be significantly improved,thereby revealing the capability of multiple RISs for enhancing the uplink coverage via their passive FD relay transmission.Additionally,the new algorithm corresponding to the nonlinear receiver and continuous RIS phase shifts possesses the best performance.