SAGE algorithm based MAP channel estimation for multi-cell massive MIMO systems

Abstract

This paper represents an efficient space-alternating generalized expectation-maximization (SAGE) algorithm based maximum a posteriori (MAP) channel estimation method for multi-cell massive multiple input multiple output (MIMO) systems. MAP channel estimation method requires conjugate transpose of a t × k pilot matrix where t is the number of pilot symbols per user and k is the number of single antenna users. Conjugate transpose of the large-size matrix increases computational complexity. The proposed method estimates the channel iteratively and converges to the same mean square error (MSE) performance of the MAP estimator with the increasing number of iterations. Consequently, the proposed method with low-rank approximation avoids conjugate transpose of the large-size matrix and hence reduces the computational complexity significantly.