Multicell Multiuser OFDMA Dynamic Resource Allocation Using Ant Colony Optimization

Evolutionary algorithms like genetic algorithms and ACO are potential candidates for solving any NP-hard problem, because of their ability to obtain acceptable suboptimum (or sometimes could be the optimum) solutions. This paper proposes an Ant Colony Optimization (ACO) based algorithm to solve the centralized resource allocation problem of a multicell multiuser OFDMA network. The proposed ACO is assisted by a water filling algorithm for power allocation. Two metrics used in ACO algorithms: the visibility and the trail intensity are defined so that they are suitable to evaluate the solution. Visibility is used to select subcarriers and power which increase the total transmitted bit of the cell while trail intensity gives solutions which decrease the inter-cell interference. Simulation results show that with these definitions, the proposed algorithm is working successfully and increases the total network transmitted bits without increasing the maximum transmit power level.