Morse报自动译码算法研究

摘要

在现代化军事通信领域,短波电报是恶劣条件下重要的应急通信手段,但现有的自动译码方法仍没达到实用化水平,依旧采用人工译码方式.人工译码方式是在确定信号频率的同时判定信号的持续时间,因此时频分析方法非常适用于Morse报的分析.提出一种基于时频分析理论,结合数字图像处理和机器学习方法的Morse报自动译码算法,旨在提高译码准确率.在点、划、间隔的识别问题中,采用了聚类算法,智能的判定元素的种类;在码文查表转换时,提出了快速索引查表算法,提高了查表效率.仿真实验表明,提出算法可以从强噪环境下较为准确的还原码文信息,对手工码的适应性强,在时变的噪声环境下具有鲁棒性.%In the field of modern military communications,shortwave telegraphy is an important means of emergency communication in harsh conditions.But the existing automatic decoding method did not reach the practical level,somanual decoding is still used.To improve the decoding accuracy,this paper proposes a Morse automatic decoding algorithm based on time-frequency analysis theory that combined with digital image processing and machine learning.The manual decoding method is to determine the signal frequency at the same time to determine the duration of the signal,so the time-frequency analysis method is very suitable for the analysis of Morse telegraph.In the problem the recognition of dot,dash and intervals,the clustering algorithm is used to make the intelligent decision about which kind of one element belongs to.To converse code to text,the search algorithm based on binary tree is used to improve the efficiency of table search.The simulation results show that the proposed algorithm can restore the code information from the strong noise environment,and it shows strong adaptability to the manual code and robustness in the time-varying noise environment.