DBSCAN-based optimization of SCMA constellation operators

In large-scale IoT (Internet of Things) device access scenarios, traditional orthogonal multiple access systems struggle to effectively handle the concurrent access of multiple devices and meet low-power communication requirements. Sparse Code Multiple Access (SCMA), as a non-orthogonal multiple access technique, improves spectrum utilization and system capacity under limited spectrum resources by designing sparse codebooks. The paper focuses on the impact of constellation phase rotation on the system when designing SCMA codebooks. Then, with the goal of maximizing the relative distance of constellation points on the synthetic constellation map, we incorporate evaluation factors in the design stage of the parent constellation and give a method to find the optimal rotation angle by using DBSCAN (Density-Based Spatial Clustering of Applications with Noise) clustering of the evaluation factors. This strategy not only ensures the sparsity of the codebook, but also reduces the mutual interference between the codebooks, so as to improve the constellation diagram layout of the codebook, and increase the system's effectiveness in dealing with the wrong code as well as the effect of detection. The results of the simulation reveal that the proposed approach can achieve lower BER and better system performance than the traditional SCMA codebook design under different SNR conditions. It meets the demand for large-scale and low-power consumption connectivity in the IoT.