الفهرس Only 14 pages are availabe for public view
 |  | from | 119 |  |  |
| | | from | 119 | | |
Abstract
Widespread interest in smart antennas has continued for several decades due to their use in numerous applications. The term smart antenna is often used in mobile communications to describe an adaptive process designed to improve the capacity of base station by focusing the radiated electromagnetic energy on transmit while improving the gain pattern on receive ITom a mobile system. The enhancement in the capacity of a communication system can be achieved primarily in the implementation of Space Division Multiple Access (SDMA). This is generally carried out using an adaptive process where we havel collection of phased array antennas. To perform adaptive processing in the phased arrays, we need to have some a priori information about the signals that we are trying to detect, we generally know or assume the direction of arrival (DOA) of the signal of interest (SOl) What is unknown is the interference and clutter scenario. Unlike. radar, in mobile communications it is difficult to know a priori the DOA of the desired signal. In this case, DOA estimation is eXploited before performing adaptive processing to yield the desired information. The roadmap of the thesis is divided into two parts:
First the previous work (chapters 1,2 and 3):
In Chapter 1, we present some introductory material and terminology associated with antenna arrays. Adaptive processing is presented with brief description on the several algorithms that have been proposed for adaptive array and various DOA estimation methods are described.
In Chapter 2, we describe a Direct Data Domain Least Squares (D3LS) approach which operates on a single snapshot of data. A D3LS approach is well suited to applications in .¬highly dynamic environment where processing data on a snapshot-by-snapshot basis is appropriate. A snapshot is defined as the set of voltages measured at the antenna elements at certain time. The Conjugate Gradient (CG) method is presented. Electrom~gnetic (EM) analysis utilized to compute the mutual coupling effects between the finite size antenna elements is illustrated. Finally a pre-processing method implemented to compensate the undesired EM effects such as the mutual coupling and the nonuniformity in the spacing between the antenna elements is demonstrated.
In Chapter 3, we present the Matrix Pencil (MP) method and the matrix enhancement and MP method for estimation of the DOA of signals impinging on a one dimensional (1 D) and a two dimensional (2D) array respectively. Again, these are direct data domain method~ computational procedure in the MP method to compute the ID-DOA of the signals impinging on the linear array is described. This procedure reduces the complexity of the computation significantly by using a unitary matrix transformation.
Second the new investigations (Chapters 4, 5 and 6):
In Chapter 4, we apply the methodology described on Chapter and not require formation of a covariance matrix. This implies that we can deal with coherent signals using a single snapshot. Finally, a very efficient 2 on a new configuralio.l (exponentially array) for narrowband signals. The adaptive processing using the previous methodology for wideband signals can be carried out using the proposed configuration. The investigation to carry out adaptive processing of wideband signals using semicircular array (SCA) is also presented. Both designs are covered in details in this chapter. we consider the antenna elements each to be, first, isotropic point source and, second, half wavelength long thin wire dipole (mutual coupling effect).
In Chapter 5, we present three proposed techniques for DOA estimation of signa], impinging on different array configurations, the first technique is based on the rotation of the plane (i.e. main lobe) of the uniform linear array by small angle electronically to estimate ID-DOA. Second, a modified procedure for the MP method described in section 3.4 for 2D-DOA estimation using a uniform planar array is presented. We proved that the method described in section 3.4 gives wrong DOA estimation and some proposed modifications steps in this technique will lead to an accurate determination of the DOA. Finally, a technique using MP method to estimate the three dimE’nsional (3D) exponent-; and hence used to obtain the associated 2D-DOA of signals impinging on the 3D array is proposed. For all the three techniques, we consider the antenna elements each to be isotropic point sources. Also, the effect of the thermal noise on the performance is investigated.
In Chapter 6, we demonstrate how the MP method described in Chapter 5 can be applied to estimate the 2D-DOA of signals impinging on a non-uniform in spacing planar array. Here a pre-processing method described in section 2.5 and implcme:::ed to compensatE’ th: undesired EM effects such as the mutual coupling and the non-uniformity in the spacing between the antenna elements is applied. Also a utilization of the unitary transform on the MP method to estimate the 2D-DOA of signals impinging on uniform or non-nnifnrm in spacing planar array is proposed for efficient computation.