Iterative Decoding for SCMA Systems Using Log-MPA with Feedback LDPC Decoding

Abstract

Sparse Code Multiple Access (SCMA) is a technique to improve spectral efficiency by sharing the same spectral resources among different users. As a result, the receiver has to cope with excess interference. A channel encoder is used to aid receiver to correct the data stream. In this paper, we discuss the use of an LDPC encoder for channel coding on each user’s traffic, followed by an SCMA encoder. SCMA encoder is a potential candidate for fifth generation (5G) mobile communication and different codebooks have already been proposed. In this paper we choose specific codewords extracted from permutation of QPSK modulation and rotation. The receiver is based on Log Domain Message Passing Algorithm (Log-MPA) for SCMA. The novelty of our paper lies in the receiver scheme with exchange of extrinsic information between SCMA decoder and LDPC decoder. Feedback from LDPC decoder provides extrinsic information to SCMA decoder, thus enabling creation of a super-graph between LDPC and SCMA decoder. Numerical results depict performance gain in terms of Bit Error Rate (BER) and complexity over conventional LDPC coded SCMA systems.