A Comprehensive Technique to Design SCMA Codebooks

Abstract

Sparse code multiple access (SCMA) is a potential candidate for improving spectral efficiency through overloading over resources. This letter proposes a comprehensive approach to achieve near-optimal SCMA codebooks designs for 150% and 200% overloading factors under various channel environments. This approach is based on a novel diversity analysis framework presented in this letter. Following this diversity analysis for the Rayleigh fading channel, we formulate a new key performance indicator (KPI) of the average symbol error rate (SER). The SER performance of an SCMA codebook is sensitive to KPIs. Thus, a suitable selection of the KPIs can improve performance. The proposed codebook design technique involves constellation rotation and dimensional permutation to optimize the KPIs. The proposed method can yield an SCMA codebook, which works equally well for uplink (UL) and downlink (DL) Rayleigh fading channels. This aspect of the proposed method eliminates the requirement of different designs for UL and DL. The SER is calculated through extensive Monte-Carlo simulations for all the proposed codebooks (with varying factors of overloading and alphabet sizes). The proposed codebooks outperform the state-of-the-art SCMA codebooks.