DOA-estimation and source-localization in CR-networks using steerable 2-D IIR beam filters

Abstract

The application of multi-dimensional (MD) infinite impulse response (IIR) space-time beam filters in radio source localization in cognitive radio (CR) environments is investigated. Knowledge of the position of radio sources in a CR network leads to the detection of white spaces in the MD frequency domain, thereby creating more opportunistic links for the secondary users. The use of MD IIR beam filters is motivated by their very low computational complexity and small side lobe levels compared to digital phased arrays. As a proof-of-concept, the two dimensional (2-D) propagation scenario including at least two receiver stations and a data fusion station, which combines the direction of arrival (DOA) estimates from the two receiver stations, to yield a position estimate, is considered. Each receiver station employs a uniform linear array (ULA) of antennas and a steerable 2-D IIR beam filter and provides information pertaining to peak energy directions. First order 2-D IIR beam filters are shown to provide acceptable DOA estimates with a SNR of 6 dB. The peak energy direction information leads to both position and MD white space detection.