Partial CSI based affinity propagation dynamic clustering algorithm
一种部分CSI下的近邻传播动态分簇算法

The precondition of whole channel state information (CSI) is impractical in realistic systems. To deal with this problem, this paper proposed a partial CSI based affinity propagation dynamic clustering algorithm. This algorithm used degree of wiliness to cooperate (DWC) to design the input matrix and introduced a cooperation fact to adjust the number of clusters in clustering results. Assuming every BS had the partial CSI information of whole system, during every iteration rounds, it propagated the messages of responsibility and availability to accumulate the coordination incidences among BSs. After that, it constructed several clusters of dynamic size. Simulation results show that the capability of this scheme is almost as same as the whole CSI based benefit-tree clustering algorithm. And the convergence iteration rounds increase slowly as the system size enlarging. This exhibits high convergence speed and is well suited for implementation in realistic system.