摘要
针对多信号分类(MUSIC)算法计算复杂度高,难以实时实现的特点,给出了适用于均匀圆阵的实数化预处理算法和实用的空间谱定义,并选择了适合硬件实现的特征值分解算法和排序算法;另外,基于某测向系统给出了MUSIC算法FPGA实现的总体结构和执行流程,并重点讨论了大矩阵特征值分解和空间谱计算的硬件结构设计.验证结果表明,该FPGA实现能够完成MUSIC算法的准确、快速计算.
Multiple signal classification （MUSIC） is a computationally complex algorithm that is hard to implement for real-time applications. First, for the case of uniform circular array （UCA）, the preprocessing algorithm making the following computations in real-value is presented and a practical spatial spectrum is defined. The algorithms for eigenvalue decomposition （EVD） and sorting suitable for hardware implementations are selected as well. Then the overall architecture and operation flow of a FPGA implementation of MUSIC on a direction finding （DF） system are proposed and the specific hardware architectures for the EVD of large matrices and the computation of spatial spectrum are discussed. Verification results show that this implementation with FPGA can compute MUSIC accurately with high speed.
出处
《北京理工大学学报》
EI
CAS
CSCD
北大核心
2010年第9期1107-1111,共5页
Transactions of Beijing Institute of Technology
关键词
MUSIC算法
测向
现场可编程门阵列(FPGA)
算法选择及优化
硬件实现结构
multiple signal classification （MUSIC）
direction finding （DF）
field programme gate array（FPGA）
algorithm selection and optimization
hardware implementation architecture
