A neural network based channel estimation scheme for OFDM system

• Published in: 2016 International Conference on Communication and Signal Processing (ICCSP) pp. 0438 - 0441
• Main Authors: Hiray, Kalpesh, Babu, K. Vinoth
• Format: Conference Proceeding
• Language: English; Japanese
• Published: IEEE 01.04.2016
• Subjects: Algorithm design and analysis; Artificial neural networks; Back propagation (BP); Channel estimation; least mean square (LMS); MIMO; multi layered perceptron (MLP); neural network (NN); OFDM; orthogonal frequency division multiplexing OFDM; Signal processing algorithms
• DOI: 10.1109/ICCSP.2016.7754174

Recent wireless standards prefer orthogonal frequency division multiplexing (OFDM) along with multiple input multiple output (MIMO) to offer high spectral efficiency services for any time anywhere environment. The full advantages of MIMO-OFDM is accessible only when there exist perfect channel information. Improper channel estimation leads to poor quality. In this work, we have developed a multi layered perceptron (MLP) based neural network (NN) which is trained with back propagation (BP) algorithm to estimate the channel characteristics of OFDM system. Monte-Carlo simulations are used to evaluate the performance of the proposed scheme with the conventional Least Mean Square (LMS) algorithm. The simulation results demonstrate that proposed schemes offers superior performance over the conventional LMS scheme under noisy environment.