An Improved Communication Signal Recognition Algorithm Based on Extreme Learning Machine

In the modern information warfare, the requirements for the reliability and real-time performance of the communication signal recognition technology are getting more and more strict. Although a great number of studies have been conducted in the reliability of communication signal recognition, few studies have been performed in the speed of communication signal recognition. The purpose of this study is to explore an improved feature extraction methods based on extreme learning machine (ELM) which has the advantage of higher speed in communication signal recognition. The results of simulations show that the approach in this paper not only improves the speed of recognition and ensures a high reliability, but also reach an ideal recognition accuracy at a low SNR.