Adaptive equalisation using particle swarm optimisation for uplink SC-FDMA

An adaptive frequency-domain equaliser for the single carrier frequency division multiple access (SC-FDMA) system using the particle swarm optimisation (PSO) technique is proposed. Unlike the stochastic gradient and recursive least squares algorithms, the PSO is known to have fast convergence which does not depend on the underlying structure. The cost function used in a PSO is formulated based on the respective structure of the equaliser, whether it is a linear equaliser or a decision feedback equaliser. The robustness of the proposed PSO algorithm is demonstrated on a high Doppler scenario. Furthermore, it is shown that the performance improves more when using re-randomisation. Finally, it is shown that the PSO-based frequency-domain equaliser is more computationally efficient than its time-domain counterpart.