The energy allocation of a transmit waveform ultimately dictates the effectiveness by which it extracts a target in a cluttered environment. The quantification of information present in the radar cross section offers notable advantages as a fitness function for the design of efficiently energy distributed waveforms for target identification. A robust method of suppression of both the temporal and spectral characteristics for target attenuating radar returns (clutter) is developed in this paper. By means of a priori knowledge of a target spectral response, a method of clutter mitigation for target identification using ultra-wide band (UWB) radar is developed. The robust design method takes after the Taguchi Method after and has seen growing use in biotechnology, statistics, and engineering as a method for both design and analysis. The Taguchi algorithm (TA) is created, based on an orthogonal matrix level design, in conjunction with the mutual information (MI) used as a criterion for convergence. This method efficiently allocates available resources within bins in which target spectral characteristics dominate those of which are undesired. As cognitive UWB radars constantly received clutter echoes and experience external noise sources, the mutual information is calculated adaptively during optimization between a transmit waveform given knowledge of the target, and the received waveform.