Multiple random access for massive MIMO framework: A unified Compressive Sensing based approach

Abstract

Recently, massive Multiple-Input Multiple-Output (MIMO) has become one of the key technologies to handle the emerging demands of future 5G wireless networks due to the capability of improving both spectral efficiency (SE) and energy efficiency (EE). Although, several practical issues such as pilot contamination, optimal spectrum utilization, computational complexity and pilot overhead still need to get more attentions. To address the above challenges, this paper views the massive MIMO system as a Multiple Random Access (MRA) problem and introduces a unified interference management framework based on compressive sampling to enhance its performance. Due to the sporadic characteristics of such a network and considering channel reciprocity, a novel uplink transmission scheme is developed which only permits active users to send pilot sequences. Then, an efficient Joint Sparse Recovery problem solver is adopted that enables Base Station (BS) to simultaneously perform user identification, channel estimation and data decoding in a one-shot paradigm. Consequently, two closed-form expressions are obtained for the maximum allowable sparsity level of uplink transmission and minimum channel gain of the proposed approach. Furthermore, sufficient constraint in order to achieve the maximum channel capacity and corresponding maximum throughput are determined as a function of system parameters. Numerical simulations illustrate the effectiveness of suggested approach in terms of spectral efficiency, pilot overhead and implementation costs, even for crowded scenarios where the sparsity constraint is not satisfied adequately.