Efficient method of estimating Direction of Arrival (DOA) in communications systems.

Abstract

In wireless communications systems, estimation of Direction of Arrival (DOA) has been used both for military and commercial purposes. The signal whose DOA is being estimated, could be a signal that has been reflected from a moving or stationary object, or a signal that has been generated from unwanted or illegal transmitter. When combined with estimating time of arrival, it is also possible to pinpoint the location of a target in space. Localization in space can also be achieved by estimating DOA using two receiving nodes with capability of estimating DOA. The beamforming pattern in smart antenna system is adjusted to emphasize the desired signal and to minimize the interference signal. Therefore, DOA estimation algorithms are critical for estimating the Angle of Arrival (AOA) and beamforming in smart antennas. This dissertation investigates the performance, angular accuracy and resolution of the Minimum Variance Distortionless Response (MVDR), Multiple Signal Classification (MUSIC) and our proposed method Advanced Multiple Signal Classification (A-MUSIC) as DOA algorithms on both Non-Uniform Array (NLA) and Uniform Linear Array (ULA). DOA is critical in antenna design for emphasizing the desired signal and minimizing interference. The scarcity of radio spectrum has fuelled the migration of communication networks to higher frequencies. This has resulted into radio propagation challenges due to the adverse environmental elements otherwise unexperienced at lower frequencies. In rainfall-impacted environments, DOA estimation is greatly affected by signal attenuation and scattering at the higher frequencies. Therefore, new DOA algorithms cognisant of these factors need to be developed and the performance of the existing algorithms quantified. This work investigates the performance of the Conventional Minimum Variance Distortion-less Look (MVDL), Subspace DOA Estimation Methods of Multiple Signal Classification (MUSIC) and the developed hybrid DOA algorithm on a weather impacted wireless channel. The performance of the proposed Advanced-MUSIC (A-MUSIC) algorithm is compared to the conventional DOA estimation algorithms of Minimum Variance Distortionless Response (MVDR) and the Multiple Signal Classification (MUSIC) algorithms for both NLA and ULA antenna arrays. The developed simulation results show that A-MUSIC shows superior performance compared to the two other algorithms in terms of Signal Noise Ratio (SNR) and the number of antenna elements. The results show performance degradation in a rainfall impacted communication network with the developed algorithm showing better performance degradation.