Diverse Traffic Demands Oriented Multi-User Detection for Grant-Free Massive MTC Networks

The diverse time-varying transmission demands cause significant challenges in the grant-free based multi-user detection (MUD) scheme design for massive machine-type communications (mMTC) networks. In this paper, we develop a multistate Markov model to characterize the diverse time-varying traffic demands, where the temporal correlation of the user activity and the data length diversity are considered simultaneously. Based on the developed Markov model, a diverse traffic demands oriented MUD scheme is proposed to realize the efficient joint user activity and data detection. Specifically, we first construct the block sparse structure for the transmitted signal to fully exploit the structured sparsity of the data matrix. Then, we convert the MUD into a maximum a posteriori probability (MAP) problem such that the block sparsity of the transmitted signal and the temporal correlation and data length diversity provided by the established Markov model can be efficiently exploited. Moreover, we further develop an intra-block pruning aided Bayesian block orthogonal matching pursuit (IBPA-BBOMP) algorithm such that the formulated MAP problem is efficiently solved. Simulation results show that the proposed scheme can achieve a substantial performance gain over existing methods.

I. Introduction

Massive machine-type communications (mMTC) is one of the three generic scenarios in the fifth-generation (5G) mobile networks and will also play the important role in the future sixth-generation (6G) systems [1]. Traditional grant-based random access schemes for mMTC networks cause excessive signaling overheads and low resource utilization efficiency due to the features of massive connections, sporadic transmissions, and small-sized packets [2]. To address the above problems, the grant-free access technology, which allows the devices to transmit data directly to the base station (BS) without scheduling, has been proposed and attracted much research attention [3].