一种改进的卫星通信信号识别方法研究

A Modified Recognition Method for Satellite Communication Signal

信号识别可为后续解析提供依据,在通信系统传输中具有重要作用。判识信号的调制类型,是实现通信信号正确接收解调的前提。针对复杂环境中卫星信号识别问题,提出了一种使用改进的高阶矩的方法,可更好地完成特征提取。在分类思路方面,采取前馈型多层神经网络优化分类器,且结合对卫星信道的深入分析,进一步提高抗噪性能和分类效率。该方法稳健性好,利于工程实现。最后,通过仿真试验验证了识别结果的有效性,且识别正确率高,性能较优。

Modulation pattern recognition can provide basis for subsequent analysis of signal parameter,and plays an important role in satellite system transmission.Identifying the modulation type of the signal is a prerequisite for the correct reception and demodulation of the communication signal.In order to solve the problem of modulation pattern recognition of satellite signal in complex environment,a modified hybrid high-order cumulant method is proposed,which can better complete feature extraction.In the aspect of classification idea,feedforward multi-layer neural network is used to optimize the classifier,and combined with the in-depth analysis of satellite channel,the anti noise performance and classification efficiency are further improved.The method has good robustness and is conducive to engineering realization.Finally,simulation experiments verify the validity of the recognition results,and the recognition rate is high and the performanc is better.