Research On High-performance Detection Algorithm For Sparse Code Multiple Access(SCMA) System

The fifth generation(5G)wireless networks are expected to support massive connections with a large number of devices.The current 4G networks cannot efficiently support massive connections and two major issues lie in the low spectral efficiency and the high cost of the control signaling overhead.To improve the spectral efficiency,the non-orthogonal multiple access(NOMA)technology is proposed and to eliminate the grant signaling overhead,grant-free multiple access becomes an attractive solution.Sparse code multiple access(SCMA),a code-domain NOMA scheme,aims to achieve massive connections and grant-free transmission and it is a promising NOMA scheme for the 5G system.In this paper,we aim to study the detection algorithm for SCMA system in the uplink grant-based and grant-free transmission scenario,respectively.In the uplink grant-based transmission scenario,based on the sphere decoding(SD),we design a high-performance receiver for uncoded SCMA system.With binary expansion of SCMA codebooks and rank deficient matrix compensation schemes,we successfully introduce the SD method into SCMA system.Regarding the coded SCMA system,we propose a list SD method,where the existing lattice points in the list are employed to set the original radius so as to further reduce the complexity.In the uplink grant-based transmission scenario,we propose a blind detection scheme based on a novel sparsity-inspired SD(SI-SD)algorithm.Considering the sparsity features in the grant-free SCMA transmission,we establish its mathematical model where the degree of sparsity is characterized by a transmission probability parameter,with which the proposed SI-SD algorithm will achieve the maximum a posteriori(MAP)detection.Based on the sparsity,we exploit the constrained SD and derive a reduced metric,which help reducing the complexity efficiently.Simulation results show that,in the uplink grant-based transmission scenario,our proposed method provides much lower complexity than the maximum likelihood(ML)detection without sacrifice of the optimal performance and provides notable gain compared to the most popular message passing algorithm.In the SCMA uplink grant-free transmission scenario,the proposed SI-SD algorithm can achieve the MAP detection and compared with the conventional SD,it provides notable gain for both the performance and the complexity.