基于分步变门限孤立森林的MFR波形单元无监督提取

Unsupervised extraction of MFR waveform units based on step-by-step variable threshold isolated forests

摘要:

波形单元是多功能雷达(multi-function radar,MFR)发射信号的基本构成,其提取的准确性直接影响后续MFR行为意图的分析。针对非协作MFR截获信号在无先验信息情况下波形单元提取时效性不佳的问题,提出了一种基于分步变门限孤立森林的波形单元提取方法。对接收的MFR脉冲信号进行漏脉冲和伪脉冲的检验,在获得脉冲参数一阶差分数据后,采用改进的分步变门限孤立森林算法进行波形单元起始脉冲搜索,实现无监督波形单元提取。理论分析和实验表明,该方法在保证波形单元提取鲁棒性的同时,显著降低了计算复杂度。

Abstract:

The transmitted signal of Multi-Function Radar(MFR)consists of waveform unit, and the extraction accuracy of waveform unit directly affects the analysis of the subsequent behavior intent of the MFR. Aiming at the problem that the timing of waveform unit extraction is not efficient in the absence of prior information, this paper proposes a waveform unit extraction method based on step-by-step variable threshold isolated forest. Firstly, the received MFR pulse signal is connected to the leakage pulse and pseudo pulse test, and after obtaining the first-order differential data of the pulse parameters, the improved step-by-step variable threshold isolation forest algorithm is used to search for the initial pulse of the waveform unit to achieve unsupervised waveform unit extraction. Theoretical analysis and experiments show that the proposed method significantly reduces the computational complexity and ensures the robustness of waveform unit extraction.