SCMA译码器研究与实现

发布时间：2018-11-07 16:24

【摘要】：移动通信经过三十年的快速发展,逐渐演化成为社会发展必不可少的工具。面对着新兴移动业务的流量需求,第五代移动通信系统应运而生。在憧憬着万物互联的新时代的同时,未来5G技术也面临着新的发展和挑战。多址接入技术是无线通信物理层最核心技术之一,传统的正交多址接入技术已经不能满足5G中大容量,海量接入等需求,稀疏多址码接入技术(SCMA)应5G需求设计产生。SCMA使用稀疏码本实现过载,有效的提高了频谱利用率,且可以通过设计良好的多维星座图和SCMA用户码本来提高系统性能。在接收端通过消息传递算法(MPA)进行译码,对译码算法进行改进也可降低多用户检测算法的复杂度。本文首先对SCMA的基本架构和基础原理进行了研究,介绍了SCMA的系统模型和编码原理,其中重点研究了三种SCMA码本设计方法,包括基于方形QAM星座图码本设计、基于星形QAM星座图码本设计以及基于星座旋转交织码本设计方式,并对不同生成码本在相同仿真条件下进行性能仿真和比较。接下来研究了SCMA译码算法,由于传统的消息检测算法涉及到大量的非线性指数运算,在其基础上提出了低复杂度的Max-Log-MPA译码算法,在1/2Turbo码率下仅有0.5dB左右的性能损失。本设计还在基于Turbo迭代联合译码下,提出了基于MPA(Max-Log-MPA)与Turbo的联合迭代多用户译码算法,并完成了其性能仿真。最后在基于本文理论基础上,在FPGA上完成了基于Max-Log-MPA检测的SCMA译码的系统验证与硬件实现,硬件实测结果与软件性能仿真结果一致,且在SCMA译码模块工作在240MHz时钟频率下,译码模块的吞吐率可以达到720Mbps。
[Abstract]:After three decades of rapid development, mobile communication has gradually evolved into an indispensable tool for social development. Facing the traffic demand of the new mobile service, the fifth generation mobile communication system emerges as the times require. At the same time, 5 G technology is facing new development and challenge. Multiple access technology is one of the core technologies in the physical layer of wireless communication. The traditional orthogonal multiple access technology can no longer meet the needs of large capacity and mass access in 5G. Sparse multiple access code (SCMA) technology should be designed for 5G requirement. SCMA uses sparse codebook to realize overload, which can effectively improve spectrum efficiency and improve system performance by designing good multi-dimensional constellation diagram and SCMA user code. The improved decoding algorithm can also reduce the complexity of the multiuser detection algorithm by decoding through the message passing algorithm (MPA) at the receiving end. In this paper, the basic structure and principle of SCMA are studied, and the system model and coding principle of SCMA are introduced. Three design methods of SCMA codebook are studied, including the design of codebook based on square QAM constellation. Based on star QAM constellation codebook design and constellation rotation interleaved codebook design method, the performance of different generated codebooks under the same simulation conditions is simulated and compared. Then the SCMA decoding algorithm is studied. Because the traditional message detection algorithm involves a large number of nonlinear exponential operations, a low complexity Max-Log-MPA decoding algorithm is proposed. Under 1/2Turbo code rate, only the performance loss of 0.5dB can be achieved. Based on Turbo iterative joint decoding, a joint iterative multiuser decoding algorithm based on MPA (Max-Log-MPA) and Turbo is proposed, and its performance simulation is completed. Finally, based on the theory of this paper, the system verification and hardware implementation of SCMA decoding based on Max-Log-MPA detection are completed on FPGA. The measured results of hardware are consistent with the simulation results of software performance. The throughput of the decoding module can reach 720Mbps when the SCMA decoding module works at the 240MHz clock frequency.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN929.5

【参考文献】