针对雷达如何实施智能抗欺骗干扰问题,提出了基于时频图像色度差异与深度随机森林的雷达智能抗欺骗干扰方法。首先,对雷达接收信号进行时频变换,经灰度处理后,将RGB三维数据信息转换为一维信息,利用全局阈值对原始时频图分割处理;其次,...针对雷达如何实施智能抗欺骗干扰问题,提出了基于时频图像色度差异与深度随机森林的雷达智能抗欺骗干扰方法。首先,对雷达接收信号进行时频变换,经灰度处理后,将RGB三维数据信息转换为一维信息,利用全局阈值对原始时频图分割处理;其次,对分割后的图片提取颜色分量特征,利用深度随机森林(deep random forest,DRF)进行干扰识别;最后,基于识别结果及真实信号的位置信息设计二维时频滤波,从而将干扰与噪声滤除。该算法不必对信号参数进行估计,计算复杂度低。仿真结果表明,干信比(JSR)大于3 dB和小于-2 dB时干扰识别率均能达到90%以上,而在干扰与信号功率相当时,可通过增加盒维数与信息熵等特征来进一步提高识别率。展开更多
Deception is widespread throughout the animal kingdom and various deceptive strategies are exemplified by social parasites. These are species of ants, bees and wasps that have evolved to invade, survive and reproduce ...Deception is widespread throughout the animal kingdom and various deceptive strategies are exemplified by social parasites. These are species of ants, bees and wasps that have evolved to invade, survive and reproduce within a host colony of another social species. This is achieved principally by chemical deception that tricks the host workers into treating the invading parasite as their own kin. Achieving levels of acceptance into typically hostile host colonies requires an amazing level of decep- tion as social insects have evolved complex species- and colony-specific recognition systems. This allows the detection of for- eigners, both hetero- and con-specific. Therefore, social parasitic ants not only have to overcome the unique species recognition profiles that each ant species produces, but also the subtle variations in theses profiles which generate the colony-specific profiles We present data on the level of chemical similarity between social parasites and their hosts in four different systems and then discuss these data in the wider context with previous studies, especially in respect to using multivariate statistical methods when looking for differences in these systems.展开更多
文摘针对雷达如何实施智能抗欺骗干扰问题,提出了基于时频图像色度差异与深度随机森林的雷达智能抗欺骗干扰方法。首先,对雷达接收信号进行时频变换,经灰度处理后,将RGB三维数据信息转换为一维信息,利用全局阈值对原始时频图分割处理;其次,对分割后的图片提取颜色分量特征,利用深度随机森林(deep random forest,DRF)进行干扰识别;最后,基于识别结果及真实信号的位置信息设计二维时频滤波,从而将干扰与噪声滤除。该算法不必对信号参数进行估计,计算复杂度低。仿真结果表明,干信比(JSR)大于3 dB和小于-2 dB时干扰识别率均能达到90%以上,而在干扰与信号功率相当时,可通过增加盒维数与信息熵等特征来进一步提高识别率。
文摘Deception is widespread throughout the animal kingdom and various deceptive strategies are exemplified by social parasites. These are species of ants, bees and wasps that have evolved to invade, survive and reproduce within a host colony of another social species. This is achieved principally by chemical deception that tricks the host workers into treating the invading parasite as their own kin. Achieving levels of acceptance into typically hostile host colonies requires an amazing level of decep- tion as social insects have evolved complex species- and colony-specific recognition systems. This allows the detection of for- eigners, both hetero- and con-specific. Therefore, social parasitic ants not only have to overcome the unique species recognition profiles that each ant species produces, but also the subtle variations in theses profiles which generate the colony-specific profiles We present data on the level of chemical similarity between social parasites and their hosts in four different systems and then discuss these data in the wider context with previous studies, especially in respect to using multivariate statistical methods when looking for differences in these systems.
