Radar Emitter Signal Recognition Based on PCA-ICS-SVM

Abstract:
In this paper, a principal component analysis (PCA) based algorithm is proposed, which combines with improved Cuckoo Search (ICS) and Support Vector Machine (SVM). Firstly, PCA is used to reduce the dimension of the eigenvalue, and then select the principal component with high contribution rate as the input vector of SVM for analysis. This paper also proposes a cuckoo algorithm which improves the search formula to find the optimal value of the key parameter penalty factor C and kernel function parameter g of SVM. The simulation results show that this method can obtain better recognition effect in the environment containing a certain amount of noise signals, which not only consumes a short time, but also has a higher recognition rate.