Abstract
In this paper we propose to use a recently-developed meta-heuristic algorithm called Firefly Algorithm for joint maximum-likelihood (ML) estimation of carrier frequency offsets (CFO) in an OFDMA uplink. We make use of a pilot preamble for prior estimation of the CFOs in a quasi-static environment. We also introduce a new distance metric that guide the evolution of the fireflies to the optimum solution. This metric is specifically formulated for this problem based on the cost function used. The algorithm gives very good performance with reasonable computational complexity and has the added flexibility of a tradeoff between the population and generation size and performance. Performance is found to be better than many prevailing methods. It is also suitable for all subcarrier assignment schemes employed in OFDMA systems. The advantages of the proposed method are validated through computer simulation.