Sparse pilot matrix design for uplink grant-free SCMA system

Abstract

The existing active user detection for uplink grant-free SCMA adopts a detection method of 4G competitive access. The pilot sequence used for the detection are ZC sequences which occupy many resource elements. In the mass machine type communication scenario, the uplink access has the characteristics of sparsity, so the user's activity can be estimated by using compressed sensing. In this paper, a sparse block circulant matrix is used as the pilot matrix (observation matrix). Combined with the OMP algorithm, the active user detection in uplink grant-free SCMA is realized under the condition of unknown active user sparsity. The simulation results show that the active user detection scheme proposed in this paper has comparable performance to the existing detection methods under the premise of reducing pilot overhead.