[关键词]
[摘要]
基于阵列信号处理的MUSIC算法,由于其超高的角分辨率和角精度,受到广泛的关注。但是其高运算复杂度,较大地限制了算法的硬件实现。为此,基于MUSIC算法的基本原理,文中分析了该算法在DSP (TI 1843)硬件系统实现中关键子步骤的运行时长。针对最为耗时和占用内存最大的空间谱构建和谱峰搜索子步骤,分别提出了便于工程实现的简化方法,并通过硬件平台移植验证了算法的可行性,对比发现该方法大大减少了运算时间和内存空间,具有较高实际应用价值。
[Key word]
[Abstract]
The MUSIC algorithm based on array signal processing is widely concerned due to its super-high angular resolution and accuracy. However, its high computational complexity greatly limits the hardware implementation of the algorithm. Therefore, based on the basic principle of MUSIC algorithm, this paper analyzes the running time of the key sub steps in the DSP (TI1843) hardware system. Aiming at the most time-consuming and memory-consuming sub-steps of spatial spectrum construction and spectral peak search, a simplified method for engineering implementation is proposed. The feasibility of the algorithm is verified through hardware platform transplantation. Comparative experiments show that this method greatly reduces computing time and memory space.
[中图分类号]
TN957
[基金项目]
国家发改委“互联网+”和数学经济专项资助项目(发改投资2019-928);安徽省新能源汽车暨智能网联汽车产业技术创新工程资助项目(皖政2017-110)
