Low-Overhead Evaluation of Multiuser Detection Performance for Physical-Layer Multiple Access Systems

In order to facilitate massive connectivity in fifth-generation (5G) systems and make full use of advanced coding schemes, designing effective multiuser detection algorithms is necessary for mitigating the interference in multiple access systems. To evaluate the performance of different detection algorithms, the conventional method is implementing the Monte Carlo (MC) simulation to estimate the detection error rate. In this paper, we present a novel simulation scheme based on adaptive importance-sampling (AIS) theory, which accelerates the simulation speed for estimating the extremely low detection error rate in multiple access systems. Specifically, by restricting the generation of random codewords to the joint Gaussian distribution biased with scaling parameters, two algorithms are proposed to determine the optimal biased parameters such that the estimated variance or the cross-entropy resulted in the AIS simulation is minimized respectively. Our proposed simulation scheme is compared with the standard MC simulation in the performance evaluation of message passing algorithm (MPA) for uplink sparse code multiple access (SCMA) system. Numerical results show that our proposed scheme provides a feasible estimation of extremely low detection error rate and achieves significant performance gain with reduced simulation overhead.

Abstract licence: CC BY-NC-ND

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.