A MLP equalizer trained by variable step size firefly algorithm for channel equalization

The paper intends to present a recent methodology for equalization of nonlinear channels employing Multilayer Perceptron Neural Networks. A hardback methodology regarding instructing the neural network using computational techniques is reported. The presented method used a modified firefly algorithm i.e. variable step size firefly for better equalization. The simulated results validate the superiority of variable step size firefly as compared to the traditional firefly and Particle swarm optimization. The outcomes give an idea that the proposed modified algorithm enriches performance of the proposed equalization process as compared to other two mentioned algorithms and it converges faster with less error to produce optimum solution.