Journal of Beijing University of Posts and Telecommunications

  • EI核心期刊

JOURNAL OF BEIJING UNIVERSITY OF POSTS AND TELECOM ›› 2011, Vol. 34 ›› Issue (1): 35-39.doi: 10.13190/jbupt.201101.35.zhoux

• Papers • Previous Articles     Next Articles

Application of Kernel Fisher Discriminant Analysis to  Digital Signal Classification

  

  • Received:2010-03-03 Revised:2010-07-02 Online:2011-02-28 Published:2011-02-28

Abstract:

In order to improve signal classification accuracy, a new classification method based on kernel Fisher discriminant analysis (KFDA) is given in the digital modulation signal classification. The higher order cumulants of the received signals are used as the classification vectors firstly, then the kernel thought is used to map the feature vector to the high dimensional feature space nonlinearly and linear Fisher discriminant analysis is applied to signal classification. The multiclass classifier is decomposed to multiple twokinds classifiers, and the classification steps of signal recognition based on kernel Fisher discriminant analysis are described in detail. It is concluded based on experiment that when proper parameters are selected, it will almost get the same classification accuracy as the support vector machine classifier. Meanwhile, support vector machine requires less time and computational complexity is lower because KFDA avoids solving the nonlinear optimization problem.

Key words: highorder cumulant, kernel Fisher discriminant analysis, kernel function, signal classification