A PAPR reduction of OFDM signal using neural networks with tone injection scheme

This paper proposes a novel peak-to-average power ratio (PAPR) reduction method for OFDM signals. The novelty of the method is the use of Hopfield type neural network (NN) with tone injection (TI) scheme to reduce PAPR. The proposed NN is suitable for global search and PAPR is sufficiently reduced, and side information of parameters for PAPR reduction transmitted to the receiver is not required. By pruning several EFFTs for neuron state updating, the proposed NN has less computational complexity than that of the conventional NNs.