Channel Equalisation of BPSK and QPSK Signal Using Higher Order Neural Networks

A high order feed forward neural network architecture like Multiplicative Neural Network (MNN), Sigma-Pi Network (SPN) and Improved Sigma-Pi Network (ISPN) are used for adaptive channel equalization in this paper. The replacement of summation by multiplication at the neuron results in more powerful mapping because of its capability of processing higher-order information from training data. The equalizer is tested on Rayleigh fading channel with BPSK and QPSK signals. Performance comparison with back propagation (BPN) neural network shows that the proposed equalizer provides compact architecture and satisfactory results in terms of bit error rate performance at various levels of energy per power spectral density ratios for a Rayleigh fading channel, and also in terms of learning rate and training time.