| Multiple-antenna technology is one of the fundamental blocks in modern wireless communication systems,which can significantly improve spectral efficiency.In order to achieve the maximum gain of a multipleantenna system,each antenna needs to be fed with a dedicated radiofrequency(RF)chain.Compared with antennas,RF chains are costly,power-hungry,and bulky.As the antenna number increases,equipping each antenna with a dedicated RF chain will challenge the system with hardware cost,power consumption,and physical implementation.Low RFcomplexity technologies could handle the above difficulties.The commonly-used low RF-complexity technologies can be divided into two categories:RF Reduction and RF Multiplexing.RF Reduction refers to reducing the number of RF chains such that it is less than the number of antennas,and typical RF Reduction techniques include antenna selection and phase shifter-based analog precoding.RF Multiplexing refers to integrating the RF resources used to achieve different wireless functionalities(e.g.,communications and sensing)such that different functionalities share the same hardware resources,thereby reducing the RF complexity.Typical RF Multiplexing technologies include integrated sensing and communications(ISAC).In order to reduce the RF complexity of multiple-antenna systems,this thesis investigates wireless transmission in low RF-complexity multiple-antenna systems.1.Antenna selection-based low RF-complexity systems.For multiple-input multiple-output(MIMO)multiple-access channel with Gaussian inputs,the ergodic capacity achieved by receive antenna selection under Nakagami fading is analyzed.A general expression for the upper bound of the ergodic capacity is derived,and it is proved that the ergodic capacity achieved by any antenna selection algorithms scales with the number of base station(BS)antennas no faster than double logarithmically.For MIMO multiple access channels with Gaussian input,a statistical channel state information(CSI)-based joint transmit precoding and receive antenna selection method is proposed,and numerical results suggest that our proposed algorithm can significantly improve the average sum rate.For MIMO wiretap channel with Gaussian inputs,a lowcomplexity joint antenna selection and secure precoding method is proposed,and numerical results show that the proposed algorithm can significantly improve the secure transmission rate of the system.For the multiple-input single-output(MISO)channel with finite-alphabet inputs,an approximated expression for the ergodic mutual information(EMI)achieved by antenna selection is derived,based on which the spectral efficiency and energy efficiency of the system are analyzed.For MIMO wiretap channel with finite-alphabet inputs,an analytical framework for the secrecy performance achieved by finite-alphabet inputs is established,based on which closed-form approximated expressions for the secrecy outage probability and average secrecy rate are derived,as well as their approximations in the high signal-to-noise ratio(SNR)regime.Moreover,the secrecy diversity order and secrecy array gain achieved by finitealphabet inputs are analyzed.2.Phase shifters-based low RF-complexity systems.For low RF-complexity systems that are based on the conventional phase-shifter network(the phase shifters are deployed inside the transceiver and directly connected to the antennas and RF chains),the achieved mutual information of the system is analyzed and optimized.For MIMO multiple access channels with Gaussian inputs,a joint user-side analog precoding and BS-side beam selection method is proposed with polynomial complexity,and numerical results show that the proposed method can significantly improve the system sum rate.For the conventional phase-shifter network-based MIMO channel with finitealphabet inputs,it is proved that when considering statistical CSI,the analog precoding introduced by the phase shifters only has an influence on the array gain of the EMI instead of its diversity gain.The statistical CSI is utilized to design the analog precoding,and numerical results show that the proposed scheme can significantly improve the system EMI.For low RF-complexity systems that are based on the novel phase-shifter network(i.e.,RIS,reconfigurable intelligent surface),the achieved mutual information of the system is analyzed and optimized.For RIS-assisted single-user MIMO channel with Gaussian inputs,closed-form expressions for the upper and lower bounds of the EMI are derived under Rician fading,and it is demonstrated that the deployment of RIS does not affect the multiplexing gain of the system.For RIS-assisted MIMO multiple access channel with Gaussian inputs,a joint BS-side antenna selection,RIS-side discretely phase-shifts precoding,and user-side transmit precoding method is proposed with polynomial complexity,and numerical results show that the proposed scheme can significantly improve the system sum rate.For RIS-assisted single-user MIMO channel with finite-alphabet inputs,highSNR asymptotic expressions of the EMI are derived under the Rayleigh product fading,and the effect of the phase-shifts of RIS on the diversity order and array gain is discussed.In addition,an average channel gainbased phase-shifts design method is proposed to improve the EMI,whose effectiveness is verified by numerical simulations.3.ISAC-based low RF-complexity systems.For IS AC systems,a mutual information-based IS AC framework is proposed,and the relationship between the sensing MI and other distortion metrics is built using the information rate-distortion theory,which demonstrates that the sensing mutual information provides a universal lower bound for distortion metrics of sense.Based on this framework,the communication and sensing performance of uplink and downlink ISAC systems are analyzed.The high-SNR asymptotic expressions for the communication and sensing rates are derived for both systems.Based on this,the diversity order,high-SNR slope,and high-SNR power offset are unveiled.The achievable sensing-communication rate region is further characterized to define the performance limit of ISAC.Theoretical analyses and numerical results show that ISAC can provide more degrees of freedom for communications and sensing than conventional separated sensing and communications(SSAC)techniques.The sensingcommunication rate region of SSAC is totally included in the rate region of IS AC. |
