A new design procedure for generating multiple clutter data by Arma Modelling

It is well established that a single source of clutter can be stochastically modelled as an Autoregressive (AR) process of relatively low order. However for Radar Target detection in presence of multiple sources of clutter the composite clutter has to be modelled as an ARMA process. In this paper a simple design procedure is given to generate Bimodal clutter data by modelling the composite process as ARMA(4,4). In the absence of real clutter data from a working Radar this data will be extremely useful in evaluating various whitening algorithms with FIR, Lattice and IIR configurations. The design parameters are found to give good flexibility in positioning the spectral peaks, the relative power levels etc.