Radar Code Design Framework with Whitened Autocorrelation Constraint

Transmitted signal waveform is among the most interesting factors affecting the radar system performance which is important to the development of cognitive radars. In this paper, the problem of adaptive waveform design is studied which specially addresses the waveforms that maximize the signal-to-disturbance ratio that finally leads to an increased probability of detection. The coded waveform is analysed in the class of linear inter-pulse coding and studied in the presence of coloured Gaussian disturbance. For this end, a new framework is introduced that leads to the design of an adaptive radar code in accordance with background distribution covariance matrix under a control on the region of available signal energy, achievable estimation accuracies and imposing a similarity constraint on the devised code periodic autocorrelation function. It is shown that the autocorrelation similarity constraint should be assessed after the whitening process in the optimum receiver. In more details, first waveform design is expressed as a non-convex quadratic optimization problem with a new autocorrelation (ACF) constraint in the whitened domain and then the optimization problem is solved by an innovative iterative algorithm named as GPP (Gradually Penalizing programming) with polynomial computational complexity. It is shown that the convergence of the algorithm is guaranteed and despite considering the suitability constraints of the ACF, improved detection performance can be achieved compared to other code synthesis methods.