Channel estimation using neural network in Orthogonal Frequency Division Multiplexing-Interleave Division Multiple Access (OFDM-IDMA) system

Abstract

In this paper, channel estimation based on neural network trained by Levenberg-Marquardt Algorithm is proposed to estimate the channel coefficients in Orthogonal Frequency Division Multiplexing-Interleave Division Multiple Access (OFDM-IDMA) systems. Conventional pilot based channel estimation algorithms like Minimum Mean Square Error (MMSE) and Least Squares (LS) are also utilized to make comparison with our proposed method with the help of bit error rate (BER) and mean square error (MSE) graphs. In this study, it is demonstrated with the computer simulations that, channel estimation based on neural network ensures better performance than LS algorithm without the requirement of channel statistics and noise information which MMSE algorithm needs to estimate the channel coefficients. Even though MMSE algorithm still shows the best performance in channel estimation, our proposed method has the advantage of being less complex and easy to apply. Because of being multiuser system, the performance of OFDM-IDMA is also evaluated for different user numbers under the channel estimation employing LS, MMSE and neural network methods, respectively. It is shown that the system performance decreases as long as the number of user is increased.