A Compound Statistical Model Based Radar HRRP Target Recognition

In radar HRRP based statistical target recognition, one of the most challenging tasks is how to accurately describe HRRP's statistical characteristics. Based on the scattering center model, range resolution cells are classified, in accordance with the number of predominant scatterers in each cell, into three statistical types. In order to model echoes of different types of resolution cells as the corresponding distribution forms, this paper develops a compound statistical model comprising two distribution forms, i.e. Gamma distribution and Gaussian mixture distribution, for target HRRP. Determination of the type of a resolution cell is achieved by using the rival penalized competitive learning (RPCL) algorithm. In the recognition experiments based on measured data, the proposed compound model not only has better recognition performance but also is more robust to noises than the two existing statistical models, i.e. Gaussian model and Gamma model.