Distributed MIMO (multiple-input multiple-output) cooperative transmission can achieve a high link capacity by large-scale antennas which are spatially distributed over a macro-cell area. Authors have been studying a multi-user spatial multiplexing technique called the minimum mean square error filtering combined with singular value decomposition (MMSE-SVD). MU-MIMO processing requires prohibitively high computational complexity for a large number of antennas. The antenna clustering can reduce the computational complexity, however, this produces the inter-cluster interference (ICI) and degrades the transmission performance of a user equipment (UE) near the cluster boundary (namely, cluster-edge UE). In this paper, we propose a dynamic antenna clustering which changes periodically the cluster boundary in order to avoid always falling into the worst condition. The UE link capacities in the cases of static antenna clustering and dynamic antenna clustering is evaluated by computer simulation. It is shown that the proposed dynamic antenna clustering improves the cluster-edge UE link capacity. The median value of the link capacity is increased by about 1.6 times compared to the static antenna clustering.