RS理论在自组织调制识别技术中的应用研究

Study on application of the Rough Set theory in self-organized modulation recognition technology

通信信号处理中最主要的问题是调制样式自动识别,其实质就是通过对输入信号的判别和归类进而理解该信号的结构和属性。本文研究了将粗糙集理论应用于自组织调制识别中的技术。它利用粗糙集理论进行信号的属性约简,求出约简和核;神经网络利用最简决策表构建分类器和提取主要特征参数。二者结合实现了信号自动调制识别,这不仅简化了网络结构和缩短了训练时间,而且也提高了调制信号的分类精度和增强了它的识别效果,以便该技术更好地运用于频谱监测、信号自动接收、实施信号干扰、电子对抗等通信领域。

The uppermost problem in the field of communication signal processing is automatic recognition of modulations types, the essence of which is to understand the structure and attribute of the signal through distinguishing and classifying input signal. This paper researched a kind of technology that rough set theory could be applied to self-organized modulation and recognition. It used rough set theory for carrying on attribute reduction and finding the core of attribute reduction, and neural networks utilized the simplest table of decision to construct classifier and extract the main eharaeteristie parameters. The combine of the two aspects above realized automatic modulation and recognition of signal. This not only simplifies the structure of network and reduces the training time, but also improves the classification accuracy of modulating signal and enhances the recognition effect of it, which could be applicated better in communication fields including spectrum monitoring, signal receiver, signal interfere and electronic countermeasures.