Cooperative distributed antenna transmission (CDAT) is a promising transmission technique to achieve a high link capacity by exploiting the spatial distribution of antennas over a macro-cell area. Recently, we proposed a minimum mean square error filtering combined with singular value decomposition (MMSE-SVD) for multi-user spatial multiplexing and showed that it can achieve higher sum capacity with lower computational complexity than BD-SVD which applies SVD after block diagonalization (BD). When using multi-user scheduling, the link capacity of MU-MIMO is affected by the combination of spatially multiplexed user equipments (UEs). In this paper, we evaluate by computer simulation the OFDM downlink capacities of MMSE-SVD and BD-SVD considering three types of multiuser scheduling (Max-SNR, Proportional Fair (PF) and Round Robin (RR)). We show that the MMSE-SVD can achieve higher link capacity with higher fairness than BD-SVD.