1Design of Phase Codes for Radar Performance Optimization With a Similarity Constraint

This paper deals with the design of coded waveforms which optimize radar performances in the presence of colored Gaussian disturbance. We focus on the class of phase coded pulse trains and determine the radar code which approximately maximizes the detection performance under a similarity constraint with a pre-fixed radar code. This is tantamount to forcing a similarity between the ambiguity functions of the devised waveform and of the pulse train encoded with the pre-fixed sequence. We consider the cases of both continuous and finite phase alphabet, and formulate the code design in terms of a non-convex, NP-hard quadratic optimization problem. In order to approximate the optimal solutions, we propose techniques (with polynomial computational complexity) based on the method of Semidefinite Program (SDP) relaxation and randomization. Moreover, we also derive approximation bounds yielding a “measure of goodness ” of the devised algorithms. At the analysis stage, we assess the performance of the new encoding techniques both in terms of detection performance and ambiguity function, under different choices for the similarity parameter. We also show that the new algorithm