Optimal Design Of Energy Efficient Multi-antenna Simultaneous Wireless Information And Power Transfer Networks

Internet of Things(IoT)is the natural evolutionary step of the Internet revolution,which spans multiple technology domains from data sensing and processing to networking and data analytics.With the development of the 5th Generation Mobile Communication System(5G),IoT is building a worldwide infrastructure that will influence all facets of human life.Improving the performance behavior of IoT especially in communication and networking is the fundamental problem in developing IoT.To tackle the main challenges in communication and networking of IoT including power management,high power consumption and network security,this dissertation inherits the advantages of simultaneous wireless information and power transfer(SWIPT),multi-antenna technology and physical-layer security,and studies the transmission designs for general single-cell and multi-cell IoT networks aiming to achieve green communication.Both performance limit and practical algorithm frameworks are presented following by theoretical analyses and simulation verification.The novelties and contributions of the dissertation include:1)The secure transmit design for multiple-input single-output(MISO)/ multiple-input multiple-output(MIMO)SWIPT system is investigated under the non-linear energy harvesting(EH)model,where a transmitter sends confidential information and transfers energy to multiple information receivers(IRs)and EH receivers(ERs),respectively,in presence of multiple eavesdroppers(Eves).To prevent the confidential information leakage,multiple artificial redundant signals(MARSs)are embedded into the transmit signals.The goal is to minimize the total transmit power by jointly optimizing transmit beamforming vectors and the covariance matrixes of MARSs such that the minimal information rate and EH requirements at IRs and ERs are guaranteed while making the received signal-to-Interference ratio(SINR)at ERs and Eves lower than their information decoding thresholds.Both the non-robust and the robust designs are studied.For the non-robust design,the optimal solution is derived.For the robust design,an approximate optimal solution is obtained by using Gaussian randomization procedure.Simulation results show that compared with traditional non-MARS aided beamforming design,the proposed design is superior in terms of the total required transmit power.It also shows that employing the non-linear EH model can avoid false output power at the ERs and/or save power at the transmitter.2)The energy-efficient design for MISO SWIPT system is investigated,where one transmitter serves multiple authorized receivers in presence of several potential Eves(idle EH receivers).To prevent the information interception by Eves,artificial noise(AN)is embedded into the transmit signals.The non-linear EH model is adopted and a novel power-splitting(PS)EH receiver architecture is proposed.Stochastic uncertainty channel model(SUM)is considered for the idle EH receivers due to outdated channel feedback.A global energy efficiency(GEE)maximization problem is formulated by jointly optimizing the transmit beamforming vectors,the AN covariance matrix and the PS ratios,under the minimal rate and secure transmission constraints of authorized receivers,the EH requirement of idle EH receivers and the total available power constraint at the transmitter.Since the problem is non-convex with no known solution,it is solved based on the following solution framework.Firstly,the PS ratios are optimized by using bisection method and successive convex approximation(SCA)and then,the transmit beamforming vectors and the AN covariance matrix are jointly optimized by using a Dinkelbach's Algorithm based method where SCA is applied to solve its inner subproblem.It is theoretically proved that by involving AN,the system GEE can be improved.Numerous results show that system GEE first increases and then keeps unchanged with the increment of the total available power,but it first keeps unchanged and then decreases with the increment of the minimal rate requirement.It is also observed that compared with traditional EH receiver architecture and linear EH model,the proposed PS EH receiver architecture is able to achieve higher GEE and avoid false output power at idle EH receivers.3)The secrecy energy efficiency(SEE)is investigated for MISO PS SWIPT networks in the presence of multiple passive Eves,where the non-linear EH model and the dual-layer PS receiver architecture are employed.With only channel distribution information(CDI)of Eves known and the AN embedded into the transmit signals at the transmitter,a SEE maximization problem is formulated under constraints of the minimal rate and EH requirements of legitimate receivers and the power budget at the transmitter.To tackle the difficulty caused by the fractional objective function and the probability constraints in solving the non-convex problem,the second-layer PS ratios are firstly optimized by a bisection and sum-of-ratios maximization methods based algorithm,and then the transmit beamforming vectors,the AN covariance matrix and the first-layer PS ratios are jointly optimized by using SCA and Dinkelbach's methods.The proposed solution approach is theoretically proved to converge to a stationary point of the semi-definition relaxation(SDR)form of the considered problem,which is further shown to be the optimal one.Numerical results show that the proposed design achieves the highest SEE over traditional power minimization and secrecy rate maximization designs.Besides,the dual-layer PS receiver architecture is able to improve the EH efficiency and system SEE.4)The AN aided multi-cell coordinated beamforming(MCBF)for secure SWIPT is investigated in both centralized and distributed manners.The proposed transmit design is formulated into a power-minimization problem to guarantee the authorized receivers' information and EH requirements while avoiding the information interception by unauthorized receivers.PS receiver architecture and the non-linear EH model are employed.Both perfect and imperfect channel state information(CSI)cases are considered.For the perfect CSI case,the non-robust design is presented by applying SDR.When no receiver harvests energy,the global optimum is guaranteed and when some receivers harvest energy,approximate global optimum is achieved.For the imperfect CSI case,the worst-case robust design under the deterministic uncertainty channel model(DUM)is studied where a solution approach based on SDR and S-procedure is proposed,and the statistically robust design under the SUM is also studied,where an upper bound to the global optimum is obtained by using SDR and Bernstein-type inequality.A distributed AN-aided MCBF design framework is further proposed by using alternating direction method of multipliers(ADMM)for the non-robust,worst-case robust and statistically robust designs,with which each transmitter is able to optimize its own transmit design with the local CSI.Simulation results show that the proposed distributed algorithm converges to the optimal results obtained by the centralized one.