The Research of Householder Transformation Systolic Beamformer
| Autor: | Chin-Tien Huang, 黃進添 |
|---|---|
| Rok vydání: | 1994 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 82 Householder transformation is a data-adaptive transformation which enjoys widespread popularity in modern numerical analysis. In signal processing field, it is desirable to find a fast and stable algorithm that satisfies the demand of matrix computations and the requirement of real-time processing. Householder trans- formation is numerically very stable, especially, for the application of adaptive algorithms. This thesis introduces a stable and highly efficient systolic algorithm. It is regular, recursive, highly pipelined, and possesses communication locality. It can also acheive the goal of real-time signal processing with parallel implementation. These properties can be exploited in developing VLSI parallel/pipelined architectures. Adaptive beamformer has been widely used in military and industrial applications such as radar, sonar, underwater communi- cation, geophysical explorations, etc. This thesis adopts a Householder transformation QR-RLS systolic array as the compu- tational engine and uses the GSC adaptive beamformer as the underlying architecture. The system proposed is a Householder transformation systolic beamformer. This modified architecture is effective in cancelling the interference and noise ; it also converges fast and needs only a very small amount of sampling data.The results gathered from thecomputer simulations have shown that the performance of Householder transformation systolic beamformer is very satisfactory. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|