Smart antennas have been proposed to improve the performance, coverage and capacity of wireless communications systems. Such antennas focus narrow beams towards the desired directions, providing more of the signal power in the transmission direction while reducing the interference in other directions. Smart antenna systems require the design of antennas that provide high gain, minimized side lobe level and uniform radiation pattern and capable of full scanning in azimuth. A good candidate for smart antenna systems is the circular array antenna for its compact structure and inherent symmetry that enables full-azimuth scanning with minimal changes in the side lobe level. However, design of antenna arrays faces several challenges and design issues, which have been addressed in this research work. This research proposes a design of circular array for smart antenna systems operating at 5.8 GHz. As the number of array elements increases, the resonant frequency shifts, and that poses a design challenge to maintain the desired resonance. Another challenge is the mutual coupling that occurs between array elements, which degrades the array performance. Inter-element spacing has been optimized at 0.71 ? to minimize mutual coupling and avoid the exhibition of grating lobes. Antenna Magus Software is used for the selection and synthesis of optimal array element, and CST Microwave Studio is used for design, verification and optimization of the circular array. The proposed circular antenna array is designed with minimal number of elements, which achieves a high gain of 15.7 dB and directivity of 17.2 dBi compared to 14.22 dB and 14.32 dBi of the reference array respectively. Mutual coupling is minimized to -27 dB and side lobe level is minimized to -17.6 dB compared to -4.5 dB of the reference array. The antenna array is capable of producing a uniform radiation pattern that is narrow towards the desired direction. This work finally achieves the design of inset-fed circular array, which outperforms the available literature and with least design complexity that is feasible for smart antenna systems.

Smart antenna refers to any type of antenna arrays combined with signal processor components which can adjust its own beam pattern in order to emphasize on the signal of interest and minimizing the interference signal. The accurate estimation of direction of interest which is also known as Direction of Arrival (DOA) of the incident signals is very significant to produce beam from antenna. There are several algorithms those have the ability in calculating the DOA of the incidents signals. In this thesis an Adaptive antenna is applied by using the most frequent used algorithm, Multiple Signal Classification (MUSIC) and Estimation of signal Parameters via Rotational Invariance Technique (ESPRIT) to obtain the Direction of Arrival (DOA) of any incident signals. Both algorithms have been implemented in Matlab 7.0 and a program was developed to investigate the effect of varying the number of samples and number of elements used in the algorithms with the existence of noise. The exact number of samples and elements used is the most important parameter in both algorithms in order to sustain the accuracy of the direction of arrival of the incident signals. The optimization was done by running the program in cellular mobile environments. The results show that the optimized number of elements is at 15 for MUSIC and 20 for ESPRIT. As for the number of sample used adequate to support the algorithm is 100 and more than 500 for MUSIC and ESPRIT respectively. The algorithm was tested for both noisy and Rayleigh Fading environment and found that it is capable to estimate the direction of arrival of the incident signals accurately with optimized number of elements and samples.