Optimizing energy efficiency in multi-user OFDMA systems with genetic algorithm

Resource allocation in an OFDMA system has been an extremely challenging optimization task. Due to the nature of the problem, this optimization problem can be NP-hard. The majority of existing methods are based on Lagrangian duality, in combination with gradient-based methods. In this paper, we apply genetic algorithm to tune the power allocated to each sub-channel in an OFDMA system. This is done by applying GA in tuning the power allocated for the m^th user in the n^th channel (p_m,n) after solving the resource allocation problem. The first part is a binary optimization problem, whereas the latter part is a continuous optimization problem. In solving the energy-efficient resource allocation problem, the objective is to minimize the transmitted power while fulfilling the transmission rate demand per user. Our simulation results showed that the GA method can obtain global solutions with speedy convergence.