Calibration Of Reciprocity Mismatch In Wireless Communication Systems

In Time-Division Duplexing(TDD)wireless communication systems,the base station(BS)exploits channel reciprocity to estimate the downlink channel state information(CSI),which can reduce the overhead of obtaining CSI for large-scale antenna systems.However,the estimated channel response consists of not only the wireless channel response but also the RF chain response.Since the transmit and receive RF chains contain different RF components,their responses are typically asymmetric,which causes the uplink-downlink channel reciprocity mismatch.In comparison to previous wireless systems,the fifth generation(5G)&beyond mobile communication systems will suffer from a greater degree of the reciprocity mismatch.On the one hand,the application of millimeter-wave(mmWave)will lead to more prominent non-ideal RF devices,which can result in more serious asymmetry of RF chains.On the other hand,the implementation of large-scale antenna systems needs to reduce the power consumption and cost of each RF component,which will lead to more severe RF chain asymmetry.Channel reciprocity mismatch can seriously degrade system performance,and thus,the research on the reciprocity mismatch plays an important role in the development and deployment of 5G&beyond systems.This dissertation focuses on the mathematical modeling,impact analysis,and calibration design of the uplink-downlink channel reciprocity mismatch in TDD large-scale antenna systems.Motivated by the developments of the RF response characteristics,the transceiver structure,and the communication system structure,we study the modeling,impact,and calibration of the nonlinear reciprocity mismatch in large-scale antenna systems,the reciprocity mismatch in mm Wave hybrid beamforming(HBF)systems,and the reciprocity mismatch in relaying systems.Specifically,the main contributions of this dissertation can be summarized in the following three parts.(1)Study the nonlinear reciprocity mismatch in large-scale antenna systems.When considering the RF chain response as a nonlinear function of the transmit signal,the reciprocity mismatch coefficient is also modeled as a nonlinear function of the transmit signal.To study the impact of the nonlinear reciprocity mismatch on the system performance,we apply the random matrix theory and the law of large numbers to derive the closed-form expression of the ergodic achievable rate,and then analyze the system performance loss caused by the nonlinear mismatch at the BS and user equipment(UE)side,respectively.The theoretical results reveal that the reciprocity mismatch at the BS side causes severe system performance degradation,while the impact of the reciprocity mismatch at the UE side is slight.Therefore,based on the theoretical results,a calibration approach is proposed to mitigate the performance loss caused by the nonlinear reciprocity mismatch at the BS side.During the reciprocity calibration,the polynomial function is applied to approximate the nonlinear mismatch ceofficient function,which simplifies the acquisition of the nonlinear function to estimate the polynomial coefficients.The polynomial coefficients can be estimated by the over-the-air training approach.After that,the optimal nonlinear calibration coefficients are computed by maximizing the downlink ergodic achievable rate.(2)Design the reciprocity calibration for mmWave-HBF large-scale antenna systems.In HBF systems,the channel reciprocity mismatch is caused by the asymmetries of digital RF chains and analog RF chains,which are coupled via a phase-shifter network.To reduce the overhead and computational complexity of the reciprocity calibration in HBF systems,we exploit the fully-connected characteristic of the phase-shifter network to virtually decouple the reciprocity calibration between the digital and analog chains by specifically designing beamforming matrices.Based on the decoupling,the entire reciprocity calibration problem of HBF systems can be equivalently decomposed into the reciprocity calibration problems of digital chains and analog chains.The mismatch coefficients of digital chains can be estimated by deriving the closed-form solution of the digital-chain calibration problem,while the mismatch coefficients of the analog chains can be solved by computing the numerical solution of the analogchain calibration problem.To solve the calibration problem efficiently,an alternating optimization-based algorithm is proposed to jointly estimate the mismatch coefficients and the parameters of the mm Wave channel.Finally,to measure the performance of the proposed reciprocity calibration,the Cramer-Rao lower bound of the mismatch coefficients is derived and reveals that the estimation error decreases with the power and length of calibration signals.(3)Study the reciprocity mismatch in multi-antenna relaying systems.The relaying system consists of sources,relays,and destinations,and the asymmetry of the RF chains in these three kinds of devices can lead to the channel reciprocity mismatch.To study the impact of the reciprocity mismatch on the relaying system performance,we first apply the random matrix theory and the law of large numbers to derive the closedform expression of the ergodic achievable rate,and analyze the system performance loss caused by the reciprocity mismatch at the source side,relay side,and destination side,respectively.The theoretical results reveal that the reciprocity mismatch at the relay side causes severe performance degradation,while the impact of the reciprocity mismatch at the source and destination side is quite slight.To mitigate the harmful impact of the reciprocity mismatch on the system performance,we propose a reciprocity calibration approach for relays.To avoid gathering CSI during the reciprocity calibration,the calibration process is divided into the intra-calibration and inter-calibration.Specifically,each relay first completes intra-calibration independently,which does not require the channel feedback.While the inter-calibration applies beamforming and multi-node concurrent transmission to avoid feeding back CSI and to improve the signal-to-noise ratio of calibration signals.Finally,to investigate the efficacy of the reciprocity calibration on the relaying system,the ergodic achievable rate is derived after the reciprocity calibration.Theoretical and simulation results indicate that the reciprocity calibration can significantly improve the relaying system performance.