Developing an innovative bimodal model to characterize the dynamic radar cross section of aircrafts

Abstract

Abstract The distribution of aircraft dynamic radar cross section (RCS) varies at different course short-cuts and heights and can behave as a bimodal shape. Classical models, such as the Weibull distribution, the chi-square distribution, and lognormal distribution, fit poorly with bimodal RCS distributions and are unable to accurately describe the statistical characteristics of aircraft dynamic RCS. In order to overcome the shortcoming of the classical models, we propose an exponential polynomial distribution model as the probability density function for studying the RCS statistical characteristics. The model parameters were acquired through the linear least-square method. Based on the RCS statistical data, the proposed distribution model was validated through simulation experiments. We calculated the fitting error between the new model and the RCS statistical distribution and determined the optimal exponential polynomial distribution model. We then computed the coefficients of the determination to evaluate the proposed model. The Kolmogorov–Smirnov test showed that the proposed approach yielded smaller values compared to conventional distribution models. For all the analyzed cases, the exponential polynomial model can fit very well against the RCS distribution data with bimodal shape, and the fitting effect is significantly better than the results of the three classical models. The advantages of the proposed model are further demonstrated by comparing it with the Gaussian mixture distribution model from fitting error, determination coefficient, goodness-of-fit, mean error and variance error. The results suggest that the exponential polynomial distribution model provides a more effective alternative in describing the fluctuation characteristics of aircraft dynamic RCS at the different course short-cuts and heights. Given the advantages exhibited by the exponential polynomial distribution model in this study, the methodology and findings in this study can be used to improve in early warning detection of radar targets and as a theoretical basis to enhance the overall performance of radars. 2009 Elsevier Ltd. All rights reserved.