An efficient codebook design approach based on salp swarm algorithm for the downlink SCMA system
Abstract
The codebooks are the key components of the sparse code multiple access (SCMA) system. In order to distinguish the multiple users from one another, a distinct codebook is assigned to each user. Apart from this, the codebook design has a direct effect on the SCMA performance. In other words, a properly designed codebook set leads to a performance boost in the SCMA system. For this reason, it is crucial to find the most appropriate codebook set to maximize the system performance. With that purpose in mind, a salp swarm algorithm (SSA)-based codebook design scheme has been proposed for the downlink SCMA system in this paper. In the proposed scheme, the codebooks are directly optimized by a recently developed robust algorithm called SSA to acquire the ideal codebook set for the related system. Most of the codebook design procedures available in the literature consider some key performance indicators like minimum product distance (MPD) and minimum Euclidean distance (MED) of the superimposed codewords while designing the codebook set. However, the system performance represented by the average bit error rate (BER) of the multiple users cannot only be determined by the distance properties of the superimposed codewords but also affected by the symbol detector by which the information bits are recovered using the designed codebook set at the receiver side. For this reason, in order to take the whole system into account without concentrating solely on the distance characteristics of the codewords, the average BER value of the SCMA scheme is employed as a fitness function of SSA to be minimized while optimizing the codebook set. In the simulation study, the BER achievement of the proposed SSA-based codebook design strategy is compared to those of varied methods existing in the literature. According to the simulation results, the codebook set produced by the suggested strategy makes the downlink SCMA system to reach better BER levels compared to the other methods considered in this paper.
© 2026 Şakir Şimşir, published by Slovak University of Technology in Bratislava
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.