TY  - THES
T1  - Performance of adaptive minimum bit-error-rate beamforming algorithms for diversity combining and interference suppression under frequency-flat directional fading channels
A1  - Garcia, Ian Dexter S.
LA  - English
YR  - 2004
UL  - https://ds.mainlib.upd.edu.ph/Record/UP-99796217607612503
AB  - Recent simulation studies have shown that Adaptive Minimum Bit-Error-Rate (AMBER) beamforming algorithms are superior to Wiener-solution-based (i.e. Maximum Signal-to-Noise-Ratio, Minimum Maen-Square, Sample-Matrix Inversion and Least-Mean-Square) beamformers in minimizing BER;s in the presence of so-channel interference under a non-multipath signal environment. In actual wireless environments however, multipath results in signal fading and possibly prevents the AMBER algorithms from achieving optimal BER performance. Simulations performed in this thesis show that the convergence capability of the function minimization algorithm in searching for optimum AMBER beamformer weights is dependent upon the input signal level. Under a fading channel, AMBER beamformers may fail to achieve lower BERs compared to Wiener-solution-based beamformers, if the function minimization method and its associated weight update parameters are not suitably chosen for the signal environment characteristics and the input signal level. The limitation of AMBER for fading environments was addressed by formulating Adaptive Normalized MBER (ANMBER) beamforming algorithms. Through simulations, it was found that under a fading channel, ANMBER beamformers are capable of achieving lower BERs compared to their respective Wiener-solution-based and AMBER beamformer configuration counterparts. In the particular fading channel used in our simulations, ANMBER was able to achieve performance gains of as much as 6 dB over its Wiener-solution-based counter parts.
NO  - Thesis (M.S. Electrical Engineering)--University of the Philippines, Diliman.
CN  - LG 995 2004 E64 G37
KW  - Adaptive signal processing.
KW  - Adaptive antennas.
KW  - Antenna arrays.
KW  - Algorithms.
ER  -