基于Rtree的稀疏码多址接入系统的译码算法

Decoding Algorithm for Sparse Code Multiple Access System Based on Rtree

在非正交多址接入(NOMA)技术中,稀疏码分多址接入(SCMA)由于码本的稀疏性,通常可以在接收端采用消息传递算法(MPA)实现多用户检测,但其计算复杂度依然较高。利用Rtree快速高效搜索的特点,提出了一种Rtree-MAP的SCMA译码算法。具体做法是将所有合成星座点生成多维Rtree空间数据结构,然后将接收信号利用Rtree快速检索得到发送端对应的用户数据信息。仿真表明,提出的Rtree-MAP译码算法在误码性能和译码速度上都优于MPA译码算法。

In non-orthogonal multiple access (NOMA) technology, sparse code multiple access (SCMA) can usually use message passing algorithm (MPA) to achieve multi-user detection at the receiver due to the sparsity of the codebook, but its computational complexity is still high. A SCMA decoding algorithm for Rtree-MAP is proposed by taking advantage of the fast and efficient search of Rtree. The specific approach is to generate a multi-dimensional Rtree spatial data structure for all synthesized constellation points, and then use Rtree to quickly retrieve the user data information corresponding to the sending end from the received signal. Simulations show that the proposed Rtree-MAP decoding algorithm outperforms the MPA decoding algorithm in terms of both BER performance and decoding speed.