Clustering based interference management for QoS guarantees in OFDMA femtocell

The deployment of femtocells is uncoordinated and opportunistic, especially in urban environment where a large number of femtocells are overlapped. Therefore, it is likely that serious femto-femto interference (FFI) may cause when they operate on the same channel. In this paper, the FFI minimization problem is mapped to Max K-Cut problem in graph theory. Based on this model, a heuristic cluster-based FFI minimized subchannels allocation algorithm (HCFM) is proposed, in which cluster is the sub-channel assignment unit. First, by using stochastic geometry to model the distribution of femtocells, the maximum femtocell cluster size is deduced to guarantee the target downlink Outage Probability (OP) for Femtocell User Equipment (FUE). Then, disjoint femtocell clusters with dynamic size and number are simultaneously generated to minimize FFI. Finally, capacity maximized sub-channels allocation algorithm is executed. Simulation results have shown that HCFM outperforms its counterparts and provide statistical Quality of Service (QoS) guarantee (in term of OP).