Complex exponential signal angle estimation based on angle invariant combiner

Abstract

In order to achieve estimation performance limits, we often need to use computationally demanding estimation algorithms and/or signal information of higher order such as cumulants. Our goal is to reduce the computational complexity of angle estimation, and to achieve the Cramer-Rao estimation bound, and the maximum-likelihood signal-to-noise ratio threshold by using iterative estimation where the most computationally demanding processing is done as much as possible in the initialisation step, while in each iteration we require less complex processing. This is achieved by using the angle invariant combinations of signal autocorrelation samples for estimation.