Taspinar, NecmiSeyman, M. Nuri2020-06-252020-06-252010closedAccess978-1-4244-5851-6https://doi.org/10.1109/WCINS.2010.5541934https://hdl.handle.net/20.500.12587/4810IEEE International Conference on Wireless Communications, Networking and Information Security (WCNIS) -- JUN 25-27, 2010 -- N China Elect Power Univ, Beijing, PEOPLES R CHINAIn high data rate communication systems which use orthogonal frequency division multiplexing as a modulation scheme, at receiver channel impulse responses must be estimated for coherent demodulation. In this paper, multilayered perceptrons (MLP) neural network with backpropagation (BP) learning algorithm is proposed as a channel estimator for OFDM systems. Our proposed MLP neural channel estimator is compared to least square (LS) algorithm, minimum mean square error (MMSE) algorithm and radial basis function neural network (RBF) in respect to bit error rate (BER) and mean square error (MSE) criteria in order to evaluate the performances. MLP neural network has better performance than LS algorithm and RBF neural network and its performance is close to MMSE algorithm and the perfect channel impulse responses. Moreover, there is unnecessary of channel statistics, matrix computation and noise information when our proposed neural network is used for channel estimation.eninfo:eu-repo/semantics/closedAccessOFDMchannel estimationneural networkmultilayered perceptron (MLP)back propagationConference Object265+10.1109/WCINS.2010.55419342-s2.0-77957665605N/AWOS:000287768500059N/A