Dynamic Frequency Selection Based on Carrier Pricing between Cells

We discuss a distributed utility-based algorithm for downlink carrier allocation in a multi-carrier system. The proposed dynamic frequency selection (DFS) algorithm aims to maximize the sum utility of the whole system. To achieve this goal, each base station (BS) selects the carrier allocation strategy to maximize a surplus function comprising both own cell utility and utility prices received from the neighbors. Universal reuse is used as a baseline for comparison. Four different utility functions are considered: mean-rate, weighted fair-rate, proportional fair-rate and max-min. A dense femto deployment scenario is used to evaluate the performance of DFS. Simulation results show that the proposed algorithm is effective in predicting the capacity gain/loss in the system. DFS enhances not only the sum data rate of a system but also the degree of fairness in resource sharing among users.