摘要：In a practical network MIMO system, a limited number of transceivers form a super virtual cell, called cluster, to cooperatively transmit data to extremely improve the system performance in terms of sum capacity, user fairness, and coverage area. Clustering algorithm is thus identified as a key enabler to reap such enormous coordinated transmission gain. In this paper, a heuristic algorithm is proposed to dynamically bind together some base stations, which can adjust itself to the ever-changing interference condition. Through utilizing the predefined connection graph, the complexity of proposed scheme is reduced while retaining a similar performance to exhaustive searching algorithm. Compared with other previous clustering methods, the proposed one demonstrates its robustness against the unrealistic feedback channel and mostly matches the performance gain produced by exhaustive searching clustering strategy.
