Fast realization of maximum likelihood angle estimation in jamming: Further results

Abstract

Angle estimation in the presence of jamming is an important function of adaptive array. A maximum likelihood (ML) angle estimator can obtain optimal angle estimation performance at the expense of high computational complexity. Using the low-rank property of the steering matrix consisting of steering vectors in the mainbeam, an arbitrary steering vector in the mainbeam can be decomposed as a product of a reduced-dimensional matrix and a low-order polynomial vector. Then, the derivative of the concentrated ML function can be well represented by four low-order real polynomials, and the extreme points of the ML function within the mainbeam can be determined by low-order real polynomial rooting. Compared to the previous real polynomial rooting technique, the computational complexity of the presented technique can be greatly reduced. Numerical examples are given to demonstrate the effectiveness of the presented technique.