Sparse Bayesian learning based user detection and channel estimation for SCMA uplink systems

Sparse code multiple access (SCMA), as a new non-orthogonal multiple-access technique, is capable to achieve massive connectivity and grant-free transmission in wireless radio access. In order to support decoding of signal from simultaneous accessing users, active users detection and channel estimation is needed, which requires significant pilot overhead, especially in channels with relative small coherent frequency / time. In this paper, an algorithm based on the framework of sparse Bayesian learning (SBL) is proposed to reduce the requirement of pilots overhead. The effectiveness of the proposal is described and analyzed, meanwhile the complexity is evaluated. Numerical simulations are performed to substantiate the performance of the proposed algorithm.