Radar Target Recognition based on KLLE and a KNRD Classifier

This paper presents a radar target recognition method using kernel locally linear embedding (KLLE) and a kernel-based nonlinear representative and discriminative (KNRD) classifier. Locally linear embedding (LLE) is one of the representative manifold learning algorithms for dimensionality reduction. In this paper, LLE is extended by using kernel technique, which gives rises to the KLLE algorithm. A KNRD classifier is a combined version of a kernel-based nonlinear representor (KNR) and a kernel-based nonlinear discriminaor (KND), two classifiers recently proposed for optimal representation and discrimination, respectively. KLLE is firstly utilized to reduce data dimension and extract features from a high resolution range profile (HRRP). Then, a KNRD classifier is employed for classification. Experimental results on measured profiles from three aircrafts indicate the relatively good recognition performance of the presented method. Key-Words: Radar target recognition, high resolution range profile, kernel locally linear embedding, kernelbased nonlinear representative and discriminative classifier