These coherent processors can fulfill arbitrary complex parameters . chapter 4 puts forward two deinterleaving methods based on probabilistic neural network ( pnn ) . the self - organization pnn deinterleaver is suitable for unknown emitters , while the rbpnn deinterleaver is suitable for known emitters 神经网络用于脉冲列去交错是国内外一直关注的解决方案,论文第四章讨论了基于概率的分类原理,提出了两种概率神经网络脉冲去交错器结构,分别适用于未知辐射源及具有先验信息辐射源两种情况。
In this dissertation , several technology problems of pulse trains deintrleaving algorithms are dealt with , they are presorting techniques based on coherent processor , probabilistic neural network deinterleavers , adaptive data association methods for pulse trains analysis and deinterleaving , signal processor designing issues . the research is focused on real time processing . the coherent processor is a crucial technique for real time presorting 本论文研究高密度复杂信号下的脉冲列去交错技术的若干问题,包括基于关联比较器的信号预分选技术研究;概率神经网络脉冲去交错器的研究与设计;卡尔曼滤波和概率数据关联方法用于脉冲列分析和去交错;雷达截获系统信号处理器设计等等。