Estimación de dirección de llegada basada en los métodos de optimización metaheurística mediante un único muestreo

Resumen

Over the last few decades, important technological developments have been introduced and applied in the field of Radar and Electronic Warfare. One of the challenges that has been the aim of a big effort by the scientific community is the estimation of the Direction Of Arrival (DOA) of a set of plane waves impinging on an array using fewer snapshots. The extreme case in which only one snapshot is available, commonly known as Single Snapshot DOA estimation, is still an open problem. This dissertation tackles that problem and proposes a method to solve it by making use of a powerful tool with applications in very diverse fields, that is, metaheuristic optimization methods. The proposed method for solving the Single Snapshot DOA estimation consists of building and then minimizing a function whose inputs are the directions of arrival of the incoming signals, which are to be estimated, and its output is a real number. The construction of this function makes use of the induced complex voltages read at the terminals of the antennas and its main characteristic is that, in the absence of noise, it has its absolute minimum in the point where the input matches the directions of arrival of the incoming waves. Moreover, in that case, the value of the absolute minimum is exactly zero. The potential of this function is observed all along the dissertation as it is used to estimate not only the directions of arrival of a set of incoming waves, but also their twodimensional directions of arrival along with their frequency, both in noisy and noiseless scenarios as well as in uniform and non-uniform arrays. A modification of this function is also presented to allow its usage in scenarios where the effects of mutual coupling and the directivity of the antennas of the array are not negligible. This modification, based on the use of numerical electromagnetic simulations, is fully consistent with the original construction of the function and permits its straightforward extrapolation to the two-dimensional direction of arrival and frequency estimation. The minimization of this function, which, as its construction, is one of the main points of the method, is accomplished by a metaheuristic optimization method. For that reason, a study on the application of five different metaheuristic optimization methods is included and their performance is compared using different criteria. All of the presented optimization methods are based on behaviours of phenomena present in nature: some of them are based on the evolutionary paradigm and some others are based on what is called Swarm Intelligence. For the evaluation of the performance of the proposed method, the Cram´er-Rao Bound has been obtained for the particular case of two-dimensional direction of arrival and frequency estimation using a Single Snapshot and also for the case when the mutual coupling is taken into account. In both cases, the estimation is efficient, from the statistical point of view