论文部分内容阅读
在现代通信对抗中,由于电磁环境特别复杂及信号调制方式的多样性,要知道所发射信号的先验知识几乎不可能,这给通信对抗带来极大的困难。为了解决在复杂多信号情况下的这个难题,提出了一种新的盲识别技术,该技术使用独立分量分析(ICA)算法来盲识别原始信号且对所得的结果进行下一步的分别处理。首先介绍了ICA的基本原理:它使用差分负平均信息量的最大化逼近。基于此,ICA的一个目标函数和一种快速的ICA算法在本文中被提出。在深入分析该快速ICA算法的基础上,将其应用于卫星TT&C信号的盲识别上。仿真结果表明:在没有任何先验知识(例如:载波频率,信号带宽和调制方式)的情况下,原始的信号可以被很好地分离出来。这为下面步骤的信号处理建立了一定的基础,比如信号分析和识别,解调信号及证明其收敛和鲁棒性等。
In modern communications confrontation, due to the special complexity of the electromagnetic environment and the diversity of signal modulation methods, it is almost impossible to know the prior knowledge of the transmitted signals, which brings great difficulties to the communications confrontation. In order to solve this problem in the case of complex multi-signal, a new blind identification technique is proposed, which uses the Independent Component Analysis (ICA) algorithm to blindly identify the original signal and process the resulting results separately. First of all, the basic principle of ICA is introduced: It uses the maximum approximation of the differential negative average information. Based on this, an objective function of ICA and a fast ICA algorithm are proposed in this paper. Based on the deep analysis of the fast ICA algorithm, it is applied to the blind identification of satellite TT & C signals. The simulation results show that the original signal can be well separated without any prior knowledge (such as carrier frequency, signal bandwidth and modulation). This establishes the basis for signal processing in the following steps, such as signal analysis and identification, demodulation of signals, and demonstration of convergence and robustness.