| Summary: | 碩士 === 輔仁大學 === 電子工程學系 === 90 === In wireless communication systems, the principal factors in system performance including the co-channel interference (CCI) from other users, the intersymbol interference (ISI), and the fading caused by multipath transmission. Blind signal separation (BSS) utilizes the antenna arrays to obtain the spatial signatures of transmitting signals. By using the signal properties such as the difference of the spatial signatures and direction of arrivals (DOA’s) of signals, the BSS techniques are able to recover each desired source correctly. So we can use these techniques to reduce the CCI and mitigate the fading phenomenon caused by multipath transmission. Blind signal separation can be applied to space division multiple access (SDMA). We can use the BSS techniques with the space diversity at antenna array to separate users which share the identical parameter like carrier frequency and spreading code. And further, we can increase the system capacity and improve the quality of communication.
The eigenspace-based DWILSP is presented in this thesis, which utilizes the eigenstructure of the correlation matrix to enhance the performance of the decoupled weighted iterative least-square with projection (DWILSP) algorithm. In the DWILSP the signal estimate is interpreted as the direct-matrix-inversion (DMI) beamforming problem. However, the DMI beamformer is sensitive to the steering vector errors, which cause the signal cancellation effects. We then use the constraint projection beamformer instead of the DMI beamformer to alleviate the signal cancellation, where the output signal-to-interference-plus-noise ratio (SINR) is increased during the estimate of signal of interest. Further, to reduce the computational complexity of the developed algorithm, an efficient implementation approach is proposed. Computer simulations are given to demonstrate that the eigenspace-based DWILSP outperforms the DWILSP.
|
|---|
