Downlink-Uplink Collaborative Channel Estimation for TDD Massive MIMO Systems

Channel estimation (CE) is a critical component in massive multiple-input multiple-output (MIMO) communication systems, while existing CE methods require a large amount of training overhead and suffer from limited estimation accuracy. In this paper, we consider the CE problem for time-division duplex (TDD) massive MIMO systems, where downlink (DL) and uplink (UL) channels exhibit good reciprocity. To fully exploit the channel reciprocity, we design a DL-UL collaborative channel sounding scheme, where a small number of transmit antennas are activated to reduce the training overhead. By integrating DL and UL channel measurements with distinctive signal-to-noise ratios into two data-fitting terms, we formulate the CE problem as a downlink-uplink collaborative atomic norm minimization (DUCANM) problem and propose a partially decoupled atomic norm minimization formulation to solve it effectively. Numerical simulations demonstrate the superiority of our proposed method in terms of CE accuracy and training overhead.