Study Of Detection Techniques For Short-wave Telegraph Signal Based On Markov Random Field

Short wave telegraph communication is using Morse code message, it is an important task ofthe shortwave tactical emergency communication way. The existing detection and decoding of shortwave telegraph signals using artificial means, has yet to realize automatic detection. This topic froma Markov random field theory, studies the shortwave telegraph signal time-frequency joint detectiontechnology, aimed at resolving the shortwave telegraph signal under strong noise backgroundtime-frequency image segmentation problem, achieve shortwave telegraph signals dots and dashesdetection, then to achieve the automatic decoding of the short wave telegraph signals theory andtechnology foundation. In this paper, the main work includes:①conducting a comprehensive summary to existing shortwave telegraph signal automaticdetection technology, in-depth analysis of the basic characteristics of the short wave telegraphsignals, from time-frequency joint analysis of the time-frequency two-dimensional plane on itsautomatic detection technology.②Studying shortwave telegraph signal time-frequency analysis method, through theexperiment analysis and comparison the short-time Fourier transform (STFT) and S transform, thesimulation experimental results is obtained, easy to capture telegraph signals by using the method ofimage segmentation.③Studying Markov random field image segmentation theory,using Markov random fieldmethod segment shortwave telegraph signals time-frequency image, through the simulationexperiment dots and dashes Morse signal has been divided.④Based on the time-frequency images which have been segmented by Markov random fieldmethod, using the algorithm of automatic detection to detect dots and dashes Morse signal, to reducethe detection error, and the effectiveness of the algorithm is verified by simulation experiments.Simulation results show that the proposed short-wave telegraph signal detection method is ableto work under strong noise background and reliable,to the human ear is difficult to identify the lowSNR short-wave telegraph signals also have very good detection effect,is suitable for short-wavetelegraph signal under strong noise background of automatic detection.