A two-timescale graph-based resource allocation scheme combing dynamic eICIC in Heterogeneous Networks

Heterogeneous Network (HetNet) has been viewed as an effective way to meet the tremendous growth of mobile traffic, in which the cross-tier and co-tier interference are the key limitations. Most of the traditional interference management strategies only focus on addressing either cross-tier interference or co-tier interference. We in this paper propose a resource allocation scheme to systematically cope with both of them in Orthogonal Frequency Division Multiple Access (OFDMA)-based HetNets. Specifically, the proposed scheme combines the dynamic enhanced inter-cell interference coordination (eICIC) and orthogonal resource allocation (ORA), to eliminate the cross-tier and co-tier interference respectively. Furthermore, with tracking the real-time variations of the user information (e.g., locations or service requirements), the proposed scheme is separated into two timescales. At the large timescale, a centralized strategy is performed to roughly allocate resource blocks (RBs) to cells; at the small timescale, a distributed strategy is performed to further schedule RBs among users. Also, we formulate two optimization problems to maximize the minimum of users' throughput satisfaction rate (TSR) for our proposed scheme, followed by two graph-based algorithms to solve them with low complexity. Simulation results manifest the superiority of our proposed scheme in guaranteeing the fairness of users' TSR.