빔공간 기반 효율적인 신호 개수 추정 알고리즘 연구

Abstract

In a modern wireless communication system such as an array antenna based beamformer or Location Based Service (LBS) composed of a plurality of antenna elements, it is important to accurately estimate the number of signals. As representative signal number estimation algorithms, there are Akaike Information Criterion (AIC) and Minimum Description Length (MDL) which are information theoretic criterion model, and the number of signals is determined using a reference value that minimizes each criterion. Although, as the number of antenna elements is increased, the accuracy of estimating the number of signals is improved, the computational complexity of the signal number counting algorithm is dramatically increased, because values of criteria for entire antenna elements should be calculated for finding their minimum. In order to improve this problem, in this paper, we propose a beamspace based AIC and MDL algorithm that efficiently estimates the number of signals by reducing the dimension of an array antenna. The proposed algorithm reduces the dimension by multiplying a received signal by beamspace weight matrix and then estimates the number of signals included in the received signal using the signal number estimation algorithm AIC and MDL. In addition, the performance of the proposed algorithm is evaluated and analyzed through computer simulation.