Düzensiz anten dizileri ile iki boyutlu geliş açısı kestirimi

Wireless communication is gaining popularity with plethora of applications occasioning a crunch in bandwidth capacity. This has necessitated exploitation of higher frequency bands with higher spectral availability. However, higher frequencies results in higher data rates with higher user density leading to multipath fading and cross-channel interference which degrades bit error rate (BER). This research proposes techniques of addressing these challenges through estimation of direction of arrival (DOA) of user signals. Specifically, the research focuses on two-dimensional (2D) DOA estimation using arbitrary arrays. Description of adaptive antenna array with factors that influence DOA estimation is comprehensively covered. Consequently, for conducting DOA estimation, coprime array is considered using MUltiple SIgnal Classification (MUSIC) algorithm. Two one-dimensional (1D) DOA estimation methods; DEcompose and COMbine (DECOM) and Unfolded Coprime Linear Array (UCLA) are proposed followed by 2D DOA estimation methods, for which; Coprime L-shaped Array (CLSA) and Unfolded Coprime L-shaped Array (UCLSA) methods coupled with a novel low-complexity 2D MUSIC algorithm are considered. Finally, virtual array interpolation technique using nuclear norm minimization is used to interpolate holes in di erence co-array of coprime array to increase degrees of freedom (DoF) of the array without increasing physical sensor elements. It was established that UCLA performed better than DECOM. Equally, UCLSA outperformed CLSA. DECOM and CLSA consider each coprime sub-array separately thereby losing the intrinsic mutual information of the array, a characteristic that is preserved in UCLA and UCLSA. Virtual interpolated array exhibited greater DoF thereby resolving 2MN -N sources with only M+N -1 sensors.