基于机器学习的激光干扰数据排除方法

Laser interference data elimination method based on machine learning

为了提高激光数据检测中的抗干扰能力,进行激光干扰数据优化滤波,实现干扰数据排除,提出一种基于机器学习的激光干扰数据排除方法。对激光数据比特序列流进行随机化处理,采用二阶格型滤波检测方法进行激光干扰数据和有用数据的滤波盲分离,提取激光干扰数据的希尔伯特谱特征量,对提取的干扰数据特征量采用机器学习方法进行自适应加权分离,实现激光干扰数据的有效排除。仿真结果表明,采用该方法进行激光干扰数据排除的滤波性能较好,对干扰数据的准确检测概率较高,盲分离能力较强,提高激光数据的检测能力。

In order to improve the anti-interference ability in laser data detection,laser interference data optimization filtering is carried out to eliminate interference data. A laser interference data elimination method based on machine learning is proposed. The laser data bit sequence stream is randomized,and the second-order lattice filter detection method is used to separate the laser interference data and the useful data. The Hilbert spectral feature quantity of the laser interference data is extracted,and the extracted interference data features are extracted. The machine learning method is used for adaptive weighted separation to achieve effective elimination of laser interference data. The simulation results show that the filtering performance of laser interference data elimination is better,the probability of accurate detection of interference data is higher,the blind separation ability is stronger,and the detection capability of laser data is improved.