先前使用預先編碼(Pre-Coding)技術在MIMO-OFDM閉合迴路(Closed-Loop)無線通訊系統設計來改善因瞬時通道估測資訊(Channel State Information)在衰落時變通道引起的失真。尚且在開路式的功率傳輸不如閉合式迴路具有更佳的適應性調變，使得較佳的通道信號有較大的輸出功率，因而改善整體信號對雜訊干擾功率比(SINR)。 本論文採用閉合迴路MIMO-OFDM設計，加上貝式決策法則來決定較佳可用之OFDM次載波並加乘其輸出功率，用以加乘之權重由本論文提出的貝式決策法則決定。 論文最後考慮在各式空間交關聯衰落通道下都卜勒(Doppler)和次載波干擾因素，對QPSK和16QAM調變，並比較傳統Alamouti MLD MIMO-OFDM系統模擬之BER和通道特徵值統計分析(EVD)在通道強健性的效益分析。 Unlike using a pre-coding scheme, a pre-weighting cooperated optimum decision rule is employed in our MIMO-OFDM closed-loop wireless system design. For closed-loop MIMO-OFDM system, it has advantages on maximizing received SINR and mitigation of the phase distortion on the feedback Channel State Information, therefor improves the system capacity. Bayes decision algorithm is developed and applied to the receive side, and feedback the optimum weighting vector to the mobile station via fast control feedback channel. The Monte Carlo Simulation in terms of BER performance of QPSK and 16QAM modulation over various spatially correlated fading and inter-subcarrier interference effects are performed and concluded. Meanwhile, the eigenvalue decomposition (EVD) is also investigated and commented for the MIMO-OFDM wireless in which the channel enhancement may not be true as we expected. On this thesis, the perfect CSI is assumed for Bayes decision rule and MIMO-OFDM system model is proposed for the analysis.