الفهرس Only 14 pages are availabe for public view
 |  | from | 164 |  |  |
| | | from | 164 | | |
Abstract
In the recent years, the number of cellular mobile phone users is increasing dramatically allover the world especially in developing countries. Among all standards, GSM is the leader with the largest market share. All expectations that said that the third generation standards including UMTS will take over the cellular market by 2002 are not realized. The real state nowadays in 2006 is that the second generation networks are still penetrating the market even in complement with third generation networks; and the number of standalone 3G operating networks stills relatively small. Thus, it is very important to support this increasing need for higher capacities not only for COMA based 3G networks but also for FDMA/IDMA based 2G networks, especially GSM. In this thesis, a practical solution for this problem using smart antenna technology is proposed.
For a smart antenna processor to be able to work in multi-standard environment that also may use different types of duplexing for the up and down links, it is very important to depend on parameters that are the same in all cases such as the received signal DOA. Also to be practically accepted, the computational complexity and execution time of any used adaptive signal processing algorithm are required to be as minimum as possible. Another important issue is that the number of signals impinging the base station antenna in any cellular system is large, thus the degree of freedom of any used algorithm’ should be as large as possible. The considered degree of freedom here is the number of coherent signals that can be simultaneously processed.
The main concern of this thesis is to propose a complete DOA-based smart antenna processor to be implemented at the base station of cellular networks with minimum computational requirements and highest possible degree of freedom. For these reasons, the proposed processor depends mainly on deterministic approaches for solving the DOA estimation problem, DOA tracking problem, and the adaptive beamforming problem. Deterministic approaches that depend on snapshot by snapshot processing are chosen to be the algorithms of interest due to its relatively small computational time, and due to its capability to deal with highly varying environments without any constraint on the statistics of the processed signal. These deterministic approaches are studied in details. Methods for both increasing degrees of freedom and for 20 generalization of these algorithms are also proposed. The complete proposed system is a multi-standard system, which means that it is suitable for mobile network operators providing complementary 2G/3G services. Deterministic solutions are also proposed for non-coherent interference cancellation for reduction of the number of signals that the processor simultaneously deals with. All used algorithms and the validity of the complete proposed system are tested using computer simulations.