Retrieving Channel Reciprocity for Coordinated Multi-Point Transmission with Joint Processing

In time division duplex coordinated multi-point transmission with joint processing (CoMP-JP) systems, the uplink-downlink channels are no longer reciprocal. Due to the difficulty of antenna calibration among the coordinated base stations (BSs), practical downlink channels differ from uplink channels by multiplicative ambiguity factors, which lead to severe performance degradation. In this paper, we propose an inter-BS antenna calibration strategy to facilitate downlink CoMP-JP transmission, whose basic idea is to estimate the ambiguity factors. To establish the observation equations for estimation, an extra uplink training frame except for the regular uplink and downlink frames is introduced, and existing signalling framework of limited feedback can also be reused. To improve the estimation performance, we can either select multiple users or employ multiple frames of one user to assist the calibration. After establishing the observation equations respectively with uplink training or limited feedback, the weighted least square criterion is used for estimation. We proceed to analyze and compare the mean square errors of the estimators, and provide a principle to select the users for assisting calibration. Simulation results show that the channel reciprocity is largely retrieved by the proposed antenna calibration strategy, which provides substantial throughput gain over the CoMP-JP systems without inter-BS calibration.