In the field of radar sensing, a paradigm shift is taking place from the traditional radar system to distributed sensing/radar networks. Comparing with the single radar system distributed radar networkshave a superior performance in detection and estimation of parameters of target. They also enhance the imaging and classification performance of sensing systems. As in distributed radar networks several radars operate at the same frequency band in a network and share the spectrum they may interfere each other, and accordingly the radar-to-radar interference becomes a major problem to be addressed. Therefore, the context aware design of radar signals for radar sensing networks to cope with the interference is pronounced. On the other hand Wavelet packet transform has recently emerged as a novel signal design technique with attractions such as good time-frequency resolution, low sidelobes, and the reconfigurability capability.
The wavelet approach is advantageous for the signal design of cognitive radar networks mainly because of its flexibility, lower sensitivity to distortion and interference, and better utilization of spectrum. For minimum interference to the adjacent radar bands, the wavelets should be maximally frequency selective. Commonly known wavelets are not frequency selective in nature and hence result in poor spectrum sharing performance. To alleviate this problem, we design a family of wavelets that are maximally frequency selective in nature. To this end, the design constraints are first enlisted. Then the problem, originally non-convex, is reformulated into a convex optimization problem and solved using Semi Definite Programming tools. Through simulation studies the benefits of the newly designed wavelets for a distributed radar sensing network are demonstrated.