A novel fingerprint feature extraction method for communication radiation source

摘要

In the complex and changeable war environment, how to intercept the enemy's communication signals and study the source individual fingerprint identification to obtain the information of the other party's communication equipment and weapon system is an important basis for understanding the enemy. Due to the subtle differences in hardware between each communication device, there are individual characteristics that differ from other devices. Fingerprint identification of communication radiation source, by studying the individual characteristics carried in the communication signal, can identify which communication equipment is coming from. Tracing the communication equipment, an important basis is provided for determining the communication system composition and strategic intention of the other party, and for the development of our action plan, which has important strategic significance in the field of communication confrontation. Based on this, this paper studies the bispectrum feature extraction method of communication emitter signal, constructs a classifier of communication emitter, and verifies the superiority of rectangular integral bispectrum method through experiments. The classifier is trained by using reduced dimension bispectrum feature vector of signal.