Adaptive Data-Aided Time-Varying Channel Tracking for Massive MIMO Systems

The time varying nature of the wireless propagation channel causes a mismatch between the true channel at the time of data transmission and its available estimate based on previously received pilot symbols, and is known to impair the performance of the massive multiple input multiple output (MIMO) systems. In this paper, we develop and evaluate adaptive data aided channel tracking and data detection algorithms to counter the effects of channel aging for uplink and downlink massive MIMO systems. We first present a recursive least squares (RLS) algorithm for tracking the matrix uplink channel at the base station (BS), and derive bounds on its MSE performance. We also derive a linear complexity stochastic gradient descent (SGD) algorithm for tracking the uplink channel, along with its performance bounds. Following this, we develop RLS and SGD based algorithms for tracking the scalar effective downlink channel at each UE, and derive their performance guarantees. Finally, via Monte Carlo simulations, we validate the efficacy of the algorithms in terms of their mean squared error performance, and demonstrate the gains achievable by channel tracking in the form of the improvement in the symbol error rates.