基带处理共性技术多核并行算法研究

发布时间：2018-05-10 10:22

本文选题：IMT-Advanced系统 + 基带信号处理　；参考：《电子科技大学》2012年硕士论文

【摘要】：随着现代通信技术的飞速发展,宽带无线移动数据业务中对传输速率、系统容量和网络性能的需求正在不断提升。为了应对这种局面,国际电信联盟(ITU)提出了第四代(4G)移动通信系统,即IMT-Advanced (International Mobile Telecommunications-Advanced)系统。IMT-Advanced系统引入了正交频分复用(OFDM)、多天线收发(MIMO)和协同多点传输(CoMP)等新技术,这对信号处理能力以及各个模块间的信息交互能力提出了更高的要求,因此有必要去研究开发支持大规模并行处理的多核数字信号处理架构并形成工具化的基带辅助设计系统和基带整体设计方案,以更好地适应未来不断发展的新技术,同时能够满足低成本、平滑升级和快速产业化的需求。因此,本文依托国家科技重大专项“面向IMT-Advanced新型基带处理共性技术研究”,展开了对IMT-Advanced系统中常用基带算法的多核并行处理研究,并设计实现了适合于多核并行算法的通用多核架构硬件验证平台。本文的研究工作主要有以下几点： 1)以多核CPU为硬件平台、Microsoft Visual Studio2008为软件环境,采用C语言搭建了基于OpenMP共享内存并行指导语句的多核并行处理仿真平台。 2)以Xilinx Vertex-5FPGA为硬件平台、EDK和ISE为开发工具,采用C语言和Verilog语言搭建了基于MicroBlaze可配置处理器软核的多核并行处理硬件验证平台。 3)在基于OpenMP和基于MicroBlaze的两种多核验证平台上,设计并实现了面向LTE-A中FFT算法的多核并行处理算法,对两种平台下的加速比性能进行测试。 4)在基于MicroBlaze的多核硬件验证平台上,设计实现了咬尾卷积编码的Viterbi译码和多相分解的FIR滤波器的多核并行算法,对加速比性能进行了测试。综上所述,本文对基带信号处理中常用算法的多核并行处理进行了研究,设计和实现了基于OpenMP和基于MicroBlaze的多核验证平台,并对FFT算法、Viterbi译码和FIR滤波器的多核并行算法进行了测试。测试结果验证了多核并行处理的可行性和有效性,揭示了并行效率与处理规模之间的关系。
[Abstract]:With the rapid development of modern communication technology, the demand for transmission rate, system capacity and network performance in broadband wireless mobile data services is increasing. In order to cope with this situation, the International Telecommunication Union (ITU) has proposed the fourth generation 4G) mobile communication system, I. e., the IMT-Advanced International Mobile Telecommunications Advanced system. The IMT-Advanced system introduces new technologies such as orthogonal Frequency Division Multiplexing (OFDM), Multi-Antenna Transceiver (MIMOM) and Cooperative Multiple-Point Transmission (CoMP). This puts forward higher requirements for the ability of signal processing and the ability of information exchange among various modules. Therefore, it is necessary to study and develop a multi-core digital signal processing architecture that supports large-scale parallel processing and to form a tool baseband aided design system and a baseband overall design scheme, so as to better adapt to the new and developing technologies in the future. At the same time can meet the need of low cost, smooth upgrade and rapid industrialization. Therefore, based on the national science and technology major project, "oriented to the IMT-Advanced new baseband processing commonality technology", this paper carried out the research on the multi-core parallel processing of the common baseband algorithms in IMT-Advanced system. The hardware verification platform of general multi-core architecture suitable for multi-core parallel algorithm is designed and implemented. The main work of this paper is as follows: 1) taking the multi-core CPU as the hardware platform and the Microsoft Visual Studio2008 as the software environment, a multi-core parallel processing simulation platform based on the OpenMP shared memory parallel instruction statement is built with C language. 2) using Xilinx Vertex-5FPGA as hardware platform and ISE as development tool, a multi-core parallel processing hardware verification platform based on MicroBlaze configurable processor soft core is built by using C language and Verilog language. 3) on two multi-core verification platforms based on OpenMP and MicroBlaze, a multi-core parallel processing algorithm for FFT algorithm in LTE-A is designed and implemented, and the speedup performance of the two platforms is tested. 4) on the platform of multi-core hardware verification based on MicroBlaze, we design and implement the parallel algorithm of Viterbi decoding and FIR filter based on bit-end convolutional coding, and test the speedup performance. To sum up, this paper studies the multi-core parallel processing of common algorithms in baseband signal processing, and designs and implements a multi-core verification platform based on OpenMP and MicroBlaze. The multi-core parallel algorithm of FFT decoding and FIR filter is tested. The test results verify the feasibility and effectiveness of multi-core parallel processing and reveal the relationship between parallel efficiency and processing scale.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TN911.7;TP338.6

【相似文献】