A clustering scheme for resource allocation of OFDMA based cellular systems

This paper presents a cluster based resource allocation scheme for allocating physical resource blocks (PRBs) in downlink of OFDMA based cellular systems. The proposed scheme includes two phases of clustering and sum rate maximization. In the first phase users are clustered by using an interference graph in a way that users in each cluster are allocated the same PRB. This interference graph helps users to avoid severe interference caused by neighbor cell base stations (BSs), since users that receive high interference from each other´s serving BSs are not in the same cluster. In the second phase, the sum weighted rate of users in each cluster is calculated for all PRBs. These values shape the rate matrix. Then, by applying Hungarian algorithm to rate matrix, the optimal resource allocation that maximizes the sum rate of clusters is obtained. Furthermore, users are grouped into cell center users and cell edge ones according to their distance from the BSs. The transmission power levels of users are fixed in two levels, in which the higher level is for cell edge users. Simulation results show that our proposed scheme improves both the throughput of cell edge users and cell center users, compared with the other schemes.