Automatic Modulation Recognition of Digital Signals Based on Fisherface
Shanshan Jin, Yun Lin, Hui Wang · 2017
This paper focuses on the design of dimensionality reduction based on Fisherface. We propose to apply the Fisherface algorithm in face recognition to automatic modulation recognition, and combine it with cyclic spectrum and k nearest neighbor classifier to realize the correct recognition of 9 kinds of modulation signals. Fisherface is an improved algorithm based on Fisher linear discriminant analysis, which can effectively reduce the sample dimension. This paper discusses the design process and gives the simulation results. The results show that the Fisherface algorithm is effective in reducing the feature dimension of digital signal in automatic modulation recognition. This can also be used for security detection and recognition.