Joint Active Device and Data Detection for Massive MTC Relying on Spatial Modulation

Abstract

The Internet of Things promises the massive connectivity of everything ubiquitously, whereas enabling massive machine-type communications (mMTC) is one of the paramount hurdles. To this end, a new paradigm is conceived for mMTC by employing spatial modulation at the devices for enhancing throughput and massive multi-input multi-output at the base station with improved detection performance. However, the associated massive access poses the intractable active device and data detection challenge. In this paper, we formulate the massive access problem as sparse signal recovery problem by exploiting the sporadic traffic of mMTC. To solve this problem, we propose a joint structured approximate message passing (JS-AMP) algorithm for joint active device and data detection. Specifically, we use AMP to decouple the superimposed received signal into uncoupled scalar elements. Furthermore, to achieve enhanced data detection performance, we compute the posterior estimation of each scalar element by using the structured sparsity of spatial modulated symbols. Moreover, we estimate the device activity by exploiting expectation maximization, where the block sparsity of successive time slots is considered for improving performance. Finally, simulation results demonstrate that the proposed solution exhibits a significant performance gain over the state-of- the-art solutions.