Estimation And Compensation Of I/Q Imbalance In TDD MIMO-OFDM Systems

As an effective way to achieve high data rates, MIMO-OFDM technique has beenwidely studied and employed in new-generation of wireless communication systems. Ina TDD MIMO-OFDM system, base station can obtain downlink channel stateinformation (CSI) through channel reciprocity and utilize it for improving downlinkcapacity. However, I/Q imbalance caused by direct-conversion limits the applicability ofchannel reciprocity, thus performance of downlink may be severely degraded ifdownlink transmission schemes are designed based on uplink measurements only.This thesis analyzes the impact of I/Q imbalance on channel reciprocity in TDDMIMO-OFDM systems. On the basis, two pre-compensation algorithms for I/Qimbalance are proposed, and a novel approach to estimate I/Q imbalance parameters ispresented. These methods are simulated in a2×2TDD MIMO-OFDM system. Themain contributions are as follows:1. The equivalent model of TDD MIMO-OFDM systems with I/Q imbalance isderived and the corresponding system capacity is given. The lose of channelreciprocity due to I/Q imbalance is analyzed in terms of system capacity.2. Two pre-compensation algorithms are proposed to compensate for the loss ofchannel reciprocity due to I/Q imbalance: one requires I/Q imbalanceparameters, and the other does not. Simulation results show that the proposedpre-compensation algorithms can effectively compensate channelnon-reciprocity due to I/Q imbalance, and therefore significantly improve thesystem performance.3. A novel I/Q imbalance parameters estimation method based on Guo’salgorithm is proposed. The loss of channel reciprocity due to I/Q imbalance iscompensated with estimated parameters. Simulation results show that theproposed method can accurately estimate I/Q imbalance parameters, andtherefore compensate channel non-reciprocity due to I/Q imbalance.