Clustering-based resource allocation in ultra-dense networks

Abstract

A dynamic cell clustering-based resource allocation algorithm is proposed in this paper aiming at the interference at the same layer between femtocells and the cross-layer interference between femtocell and macrocell. The algorithm includes two steps. In the first step, a subchannel allocation algorithm based on rate fairness is used to allocate subchannels for macro users, followed by power allocation with the water-filling algorithm. In the second step, it firstly conducts clustering for FAPs by utilizing the genetic simulated annealing algorithm according to the size of the weight of the interference among femtocells, and then allocates sub-channels for FUEs by utilizing the heuristic channel allocation algorithm, and then allocates power for users by adopting KKT conditions. According to simulation results, this algorithm can restrain both of the two interferences effectively, satisfying the demands of users on speed and improving the network spectrum effectiveness.