The application and improvement of Dempster-Shafer evidence theory in communication signal recognition

Communication signal´s sorting and recognition is an important problem for the electronic countermeasure procedure. Generally speaking, the essential problem of Communication signal´s sorting and recognition is a pattern recognition problem. Through many years developing, Dempster-Shafer evidence theory has become a very important pattern recognition method which is widely used in signal sorting and recognition. But, there are still many key problem needed to be worked out, such as how to get the Basic Probability Assignment Function (which is usually called BPAF in short) of evidence, how to fuse the conflict evidence, how to get more effective evidence fusion structure. So, in these thesis, first, the computation of Basic Probability Assignment Function based on grey associated analysis algorithm is provided, secondly, a traditional conflict evidence fusion method is introduced and a double-deck fusion structure is built based on the research achievement. Finally, two simulation experiments are designed to prove the research result. The final result shows that the Dempster-Shafer evidence theory has a well prospect in communication signal´s sorting and recognition.