海杂波FRFT谱的多重分形特性与目标检测

摘要

The paper mainly studies the multifractal property of sea clutter in the fractional Fourier transform ( FRFT) domain and its application to target detection within sea clutter. According to the scale property of the mathematical definition of FRFT, the FRFT spectrum of a self-similar process in some transform order does not have uniform self-similarity at each scale. In the light of this characteristic, the multifractal theory is introduced to analyze the self-similar structure of the sea clutter FRFT spectrum and the influencing factors of the multifractal parameters in FRFT domain are studied as well. By the verification of S-band and C-band radar data, it' s shown that the sea clutter FRFT spectrum is multifractal and the generalized Hurst exponent in FRFT domain has the ability of distinguishing target from sea clutter. Based on this, the generalized Hurst exponent in FRFT domain and the bi-parameter constant-false-alarm-rate detector are combined to design target detecting method in sea clutter and the performance is analyzed as well. The results show that the proposed method performs much better than the classic fractal detecting methods in time domain.%本文主要研究海杂波在分数阶傅里叶变换(FRFT)域所表现出的多重分形特性及其在海杂波目标检测中的应用.由FRFT数学定义的尺度性质可推得,自相似过程在某一变换阶数下的FRFT谱在各尺度下不具有统一的自相似特性.针对这一特性,本文将多重分形理论引入到对海杂波FRFT谱的自相似结构分析中并研究FRFT域多重分形参数的影响因素,经S波段和C波段雷达实测数据验证表明,海杂波FRFT谱具有多重分形特性且FRFT域广义Hurst指数对海杂波和目标具有良好的区分能力.在此基础上,本文利用FRFT域广义Hurst指数与双参数恒虚警检测器相结合设计海杂波中目标检测方法并分析检测性能,结果表明本文所提方法相比于经典的时域分形检测方法具有较明显地性能提升.