Inter-cell interference coordination based on power control for self-organized 4G systems

Orthogonal Frequency Division Multiplexing (OFDMA) accepted as the multiple access scheme for the 4G Systems provides resistance to inter-symbol and intra-cell interference. However inter-cell interference, when dense frequency reutilization is used, can deteriorate the performance of users with bad channel quality, in particular at cell-edge. Inter-Cell Interference Coordination (ICIC) [1] is a promising mechanism to enhance system performance of 4G. This paper addresses the problem of ICIC in the LTE downlink where the power level selection of resource blocks (RBs) is portrayed as a sub-modular game in the context of self-organizing networks. The existence of Nash equilibriums (NEs) for that type of games shows that stable power allocations can be reached by selfish eNodeBs. To attain these NEs, we propose a semi-distributed algorithm based on a best response algorithm. Based on local knowledge exchanged through the X2 interface in 4G networks [2], each eNodeB will first select a pool of low interference RBs. Then, each eNodeB - to save energy - will make its best to fix the power level on these RBs achieving comparable performances in comparison with a policy serving active users with full power (MAX Power Policy). In order to evaluate our proposal, we compare the obtained results to an optimal global CoMP (Coordinated Multi-Point) solution where a central controller is the decision maker [3].