摘要
雷达目标一维像识别是雷达目标识别中的一个重要方向,并且已经发展了多种算法,子空间法是其中一种有效的方法。本文侧重于子空间法中特征子空间的实现研究,介绍了利用FPGA实现算法的基本设计方法,讨论了CORDIC实现原理、数值精度以及利用FPGA实现CORDIC-SVD的方法并将其应用于特征子空间法的实现中,最后提出了实现奇异值分解的Systolic阵列结构。
Radar target recognition using range profiles plays an important role in radar target recognition and many algorithms are proposed. Subspace method is one of the efficient ways to implement it. This paper attempts to implement the Eigen - subspace method. We introduce the basic design method using FPGA to implement the arithmetic and discuss the CORDIC principle,precision of the numerical value and the implementation of the CORDIC- SVD in FPGAs and its applications in the Eigen - subspace method＇s implementation. Finally bring forward the Systolic array structure algorithm for the higher dimension SVD and CORDIC implement theory.
出处
《现代电子技术》
2007年第19期77-79,共3页
Modern Electronics Technique
