Bio inspired Cuckoo Search Algorithm based neural network and its application to noise cancellation

In this paper a Cuckoo Search Algorithm (CSA) based neural network is proposed for noise removal from a signal. A new training function is proposed, which uses Cuckoo Search Algorithm for the training of the network. The trained network is then used to remove noise from a sinusoidal signal contaminated with white Gaussian noise. Various types of random walks and Levy flights are used in the algorithm and their effect on the final output has been analyzed. A comparison is made between the present technique and already existing Levenberg-Marquardtback-propagation(LMBP). The compare-ison shows the superiority of the present approach over the Levenberg-Marquardt back-propagation.