An improved multi-cell resource allocation scheme based on graph theory

A low-complexity multi-cell orthogonal frequency-division multiple access (OFDMA) downlink channel-assignment method that uses a graphic framework is proposed in this paper. Our aim is to solve the problem of low signal-to-interference-plus-noise ratio (SINR) for the users in the edge cell. A frequency reuse factor of 1 is adopted in cell deployment to ensure maximum spectrum efficiency. Our solution consists of two phases: 1) Inter-cell interference (ICI) is mapped to the minimum graph colouring problem and is solved by greedy algorithm. 2) Channel assignment is accomplished by adopting cost scaling algorithm based on weighted bipartite matching method. Simulation results show that our proposed scheme provides better performance to cell-edge users than the MAX-SNR algorithm. We emphasize the fairness in resource allocation, maximizing the system throughput at the same time.