A multiple-model prediction approach for sea clutter modeling

Accurate modeling of sea clutter is an important problem in remote sensing and radar signal processing applications. Due to a recent discovery that sea clutter, the electromagnetic wave backscatter from a sea surface, is chaotic rather than purely random, computational intelligence techniques such as neural networks have been applied to develop new models for sea clutter. In this paper, we propose using the multiple neural network model approach to construct a predictive model for sea clutter. The motivation comes from the observation that the sea usually has some unpredictable motions that result in impulsive events such as sea spikes. Although a single nonlinear model could describe the Bragg scattering reasonably as shown in the literature, it is usually incapable of capturing sea spikes motions. Therefore, target detection performance might be degraded when such a clutter model is employed. Using a multiple radial basis function (RBF) net predictor, we found that a sea clutter signal with different underlying dynamics from sea spikes to normal motions can be modeled accurately. The multiple model (MM) approach automatically assigns different RBF predictors to model sea spikes and other mechanisms like Bragg scattering. The proposed multiple RBF neural network uses the expectation-maximization algorithm and multistep prediction for training, and hence it is suitable for real-time signal processing. Using real-life radar clutter data collected at the east coast of Canada, the proposed MM approach is shown to be effective in isolating and characterizing various components of sea clutter and, therefore, provides a promising model for clutter suppression in radar detection.