Active User Detection and Channel Estimation for Spatial-based Random Access in Crowded Massive MIMO Systems via Blind Super-resolution

TL;DR

This paper introduces a blind super-resolution method leveraging angular domain sparsity and clustering to detect active users and estimate channels in crowded massive MIMO systems, significantly reducing training overhead.

Contribution

It proposes a novel off-grid atomic norm minimization and clustering framework for joint user detection, channel estimation, and data recovery in grant-free massive MIMO systems.

Findings

Effective AoA detection and data recovery demonstrated in simulations

Reduces training overhead by exploiting sparsity and sporadic user activity

Accurate channel estimation achieved with limited observations

Abstract

This work presents a novel framework for random access in crowded scenarios of multiple-input multiple-output(MIMO) systems. A multi-antenna base station (BS) and multiple single-antenna users are considered in these systems. A huge portion of the system resources is dedicated as orthogonal pilots for accurate channel estimation which imposes a huge training overhead. This overhead can be highly mitigated by exploiting intrinsic angular domain sparsity of massive MIMO channels and the sporadic traffic of users, i.e., few number of users are active to sent or receive data in each coherence interval. In fact, the angles of arrivals (AoAs) coming from active users are continuous parameters and can take any arbitrary values. Besides, the AoAs corresponding to each active user are alongside each other forming a specific cluster. This work revolves around exploiting these features. Specifically, a blind clustering-based algorithm is proposed that not only recovers the transmitted data by users in grant free random access and primary pilots in random access blocks of coherent transmission, but also provides accurate channel estimation. Our approach is based on transforming the unknown variables into a higher dimensional space with matrix variables. An off-grid atomic norm minimization is then proposed to obtain the unknown matrix from only a few observed arrays at the BS. Then, a clustering-based approach is employed to identify which AoAs correspond to which active users. After identifying active users and their AoAs, an alternating-based approach is performed to obtain the channels and data or primary pilots of active users. Simulation results demonstrate the effectiveness of our approach in AoA detection as well as data recovery.