Spatial array filtering employing a nonlinear constraint measure

An approach to improve the performance of arrays with a low order of sensors is examined. The motivation is to reduce the number of sensors, and hence the cost, required for satisfactory performance. Based on a review of commonly employed performance optimization methods, a trade-off approach is formulated which employs a signal propagation model and a nonlinear smoothness constraint model of the subspace of optimal processor weight vectors. Simulation results illustrate the behavior of the approach.