Capacity for downlink massive MIMO MU-SCMA system

Abstract

Sparse code multiple access (SCMA) is a new multiple access technique with reusing non-orthogonal code domain resource. It has massive connectivity serving a large number of users and improves spectral efficiency of radio access. SCMA directly maps the information bits of different users to sparse complex codewords selected from the predefined codebooks, and each codeword represents a transmission layer. SCMA can be combined with MIMO technique to further improve the spectral efficiency. This combined technique is called code-space-multiplexing (CSM) which is the extension of the spatial multiplexing. Under downlink massive MIMO MU-SCMA (massive-CSM) system, the base station, equipped with massive transmit antennas, transmits SCMA codewords for each user. In this paper, by applying random matrix theory, we derive an asymptotic equivalent (AE) of signal to interference plus noise ratio (SINR) as the number of transmit antennas approaches infinity. In addition, some asymptotic results of SINR for two special cases can be obtained in terms of large-scale fading coefficients, sparse parameter and the number of resource blocks. Simulation results show massive-CSM can improve the capacity of MU-SCMA system and has a substantial advantage in terms of higher sum rate performance compared with Massive MIMO MU-OFDMA system. As the number of transmit antennas increases, our analytic expression of SINR closely approaches the theoretical result.