Optimization of quiescent response in partially adaptive beamformers

The adaptive weights utilize only a fraction of the available degrees of freedom in partially adaptive LCMV (linearly constrained minimum-invariance) beamformers. Two procedures are presented for incorporating unused degrees of freedom into the beamformer's quiescent response. In one case the quiescent response is chosen to approximate a desired quiescent response, and in the other case it is chosen to maximize the average interference cancellation over a set of likely interference scenarios. A sidelobe-canceling interpretation indicates that maximization of average interference cancellation results in quiescent weights that optimize the sum of the white noise gain and the error between the quiescent response and the class of responses which the adaptive branch can synthesize. Simulations illustrate the utility of the methods.>