In addition to the advantage of high capacity in conventional multiple input multiple output (MIMO) systems, distributed large scale (DLS) MIMO systems can further enhance energy efficiency and the system performance at the cell edge. Key challenges in DLS-MIMO systems include interference mitigation and complex signal processing in a large number of antennas. Thus, the concept of antenna clustering is proposed to reduce the complexity of signal processing, but the inter-cluster interference still affect the system performance seriously. In this paper, we present a dynamic antennas clustering algorithm to overcome the inter-cluster interference, while reducing the complexity of signal processing. Simulation results show that our proposed antenna clustering algorithm can improve the spectrum efficiency of 80% for cell edge users and energy efficiency of 20% compared with the existing antenna clustering techniques.