Channel Estimation And Passive Beamforming For Reconfigurable Intelligent Surface-assisted Communications

The growing energy consumption has become one of the crucial problems in nextgeneration wireless networks.Recently,reconfigurable intelligent surface(RIS)is envisioned to be a potential technology to solve this issue due to its low-cost,energy-efficient,programmable,and easy-to-deploy advantages.Specifically,RIS is an artificial metasurface composed of a large number of electromagnetic elements,each of which is capable of dynamically manipulating the amplitude,phase,polarization,and frequency of the electromagnetic response,thus creating a smart and reconfigurable wireless environment.Channel estimation and passive beamforming remain two crucial challenges in RIS-assisted communication systems,which have been discussed in some open literature.However,on the one hand,existing schemes generally adopt constant-enveloped pilot symbols and are mainly oriented for narrowband communication systems,thus limiting the RIS performance.On the other hand,they suffer from high training overhead and high computational complexity for large-scale RIS deployment.To address these problems,we systematically study RIS-assisted communication system design.Specifically,we discuss the pilot power allocation in channel estimation,passive beamforming in wideband systems,and low-complexity channel estimation and passive beamforming design.The main contributions of this dissertation are as follows:1.For RIS-assisted single-input single-output(SISO)communication systems,we formulate the relationship between ergodic achievable rate and pilot power under imperfect channel state information(CSI),where two cases with single RIS and multiple RISs are considered.By maximizing the ergodic achievable rate under imperfect CSI,we propose the optimal pilot power allocation strategy.Both theoretical analysis and simulation results demonstrate that for channel estimation of RIS-assisted SISO communication systems with limited total pilot power,it is suggested to allocate more pilot power to estimate the weak channel,thus improving the ergodic achievable rate under imperfect CSI.In addition,theoretical analysis and simulation experiments also evaluate the peak-to-average power ratio of pilot signals under the proposed scheme.2.For RIS-assisted orthogonal frequency division multiplexing(OFDM)communication systems,we firstly introduce the delay adjustable RIS.Following this,we formulate the achievable rate maximization problem upon jointly optimizing transmit power allocation,reflection phase shift,and reflection delay.In order to address this problem,we propose an alternate optimization algorithm.To reduce the computational complexity,we further propose a heuristic algorithm and theoretically analyze the achievable rate of RISassisted OFDM communication systems.Simulation results demonstrate that compared to conventional RIS,the delay adjustable RIS relying on proposed algorithms could significantly improve the achievable rate of RIS-assisted OFDM communication systems.In addition,we also evaluate the effects of imperfect CSI and power attenuation caused by practical devices on the achievable rate of OFDM communication systems.3.For RIS-assisted multi-user multiple-input single-output(MISO)communication systems,we propose a channel estimation and passive beamforming framework based on the reflection phase shift codebook,where the continuous reflection phase shift and the optimization objective of minimizing the downlink transmit power are considered.Furthermore,we propose two methods for generating the continuous reflection phase shift codebook,and analyze the channel gain between the base station and the user.Besides,the effects of imperfect CSI on the existing modularized scheme and the reflection phase shift codebook-based scheme are also analyzed.Simulation results show that when imperfect CSI is considered,compared to the existing modularized scheme,the proposed scheme based on reflection phase shift codebook has advantages in complexity,training overhead as well as control signaling overhead and also reduces the transmit power.4.For RIS-assisted point-to-point multiple-input multiple-output(MIMO)communication systems,we improve the channel estimation and passive beamforming design scheme based on the reflection phase shift codebook.Specifically,we propose two methods for generating the discrete reflection phase shift codebook,and analyze the channel gain between the base station and the user.In addition,we analyze the optimal training overhead to maximize the effective achievable rate of RIS-assisted MISO communication system under a given value of channel coherence time.Simulation results demonstrate that compared to the existing modularized design scheme,the proposed scheme based on the reflection phase shift codebook has stronger robustness to channel estimation errors and reflection phase shift quantization errors.When considering imperfect CSI,the proposed scheme is superior in terms of achievable rate,complexity,and training overhead.