Computationally Efficient Maximum Likelihood Channel Estimation for Coarsely Quantized Massive MIMO Systems

We consider computationally efficient maximum likelihood (ML) channel estimation for massive multiple-input multiple-output (MIMO) systems using coarsely quantized measurements obtained from low-resolution analog-to-digital converters (ADCs) at the receivers. We first devise a computationally efficient ML estimator (referred to as CQML) by using the cyclic optimization and majorization-minimization (MM) techniques. Then, we show the connections between CQML and the conventional unquantized ML estimator. Numerical examples are provided to demonstrate the effectiveness and computational efficiency of the proposed channel estimation algorithm.

Automated summary

This paper presents a computationally efficient maximum likelihood (ML) channel estimation method for massive MIMO systems, using coarsely quantized measurements obtained from low-resolution ADCs at the receivers. A computationally efficient ML estimator (CQML) is devised using cyclic optimization and majorization-minimization techniques, and the connections between CQML and the conventional unquantized ML estimator are shown. Numerical examples are provided to demonstrate the effectiveness and computational efficiency of the proposed channel estimation algorithm.