Efficient Machine Learning-Enhanced Channel Estimation for OFDM Systems

Abstract

Recently much research work has focused on employing deep learning (DL) algorithms to perform channel estimation in the upcoming 6G communication systems. However, these DL algorithms are usually computationally demanding and require a large number of training samples. Hence, this work investigates the feasibility of designing efficient machine learning (ML) algorithms that can effectively estimate and track time-varying, frequency-selective channels. The proposed algorithm is integrated with orthogonal frequency-division multiplexing (OFDM) to eliminate intersymbol interference (ISI) induced by the frequency-selective multipath channel and compared with the well-known least square (LS) and linear minimum mean square error (LMMSE) channel estimation algorithms. The obtained results have demonstrated that even when a small number of pilot samples, N P , is inserted before the N subcarriers OFDM symbol, the introduced ML-based channel estimation is superior to the LS and LMMSE algorithms. This dominance is reflected in the bit-error-rate (BER) performance of the proposed algorithm, which attains a gain of 2.5 dB and 5.5 dB over the LMMSE and LS algorithms, respectively, when N P = N 8 . Furthermore, the BER performance of the proposed algorithm is shown to degrade by only 0.2 dB when the maximum Doppler frequency is randomly varied. Finally, the number of iterations required by the proposed algorithm to converge to the smallest achievable mean-squared error (MSE) are thoroughly examined for various signalto-noise ratio (SNR) levels. INDEX TERMS Deep learning (DL), machine learning (ML), orthogonal frequency-division multiplexing (OFDM), intersymbol interference (ISI).The associate editor coordinating the review of this manuscript and approving it for publication was F. K. Wang . computational requirements, which, in turn, increases the complexity of the communications systems. Moreover, since DL networks usually contain a large number of neurons distributed among many layers, these networks require a large number of training samples to converge to the desired output. Unlike DL algorithms, shallow ML algorithms exhibit low computational requirements that make them suitable for implementing fast signal processing schemes, which can fulfil the increasing data traffic requirements imposed by the new generation of wireless communication systems. Consequently, immense attention has been shifted towards exploiting ML algorithms to revolutionize modern wireless communication systems [6]. In essence, conventional communication systems are generally based on well-established mathematical models. Hence, channel estimation algorithms such as the minimum mean square error (MMSE) algorithm

A Comprehensive Review of UAV-Assisted FSO Relay Systems

“…However, in [35], the theoretical Power Density Functions (PDFs) of misalignments and composite channel models were derived from approximation approaches, which resulted in some discrepancies with the experimental data, especially at the tails of the distributions. Such results support the assertion by the authors of [66] that utilizing Machine Learning (ML) algorithms offers a data-driven perspective that does not rely on channel statistics and approximations. This makes ML algorithms suitable for emerging and unique communication configurations where channel statistical modeling is challenging.…”

“…Essentially, it is important to understand which ML method is suited to the nature of the problem at hand, and for this reason, the authors of [70] exploited an extensive dataset for FSO links over a maritime environment to compare the prediction accuracy of different methods, determining that a shallow Artificial Neural Network (ANN) performed best. In [66], the authors also demonstrated the effectiveness of shallow ANNs over deep learning methods when constrained by a smaller training sample size. The effectiveness of ANNs was further shown by the authors of [71], who used a shallow Nonlinear AutoRegressive eXogenous (NARX) ANN to accurately estimate channel state using experimental data from an outdoor FSO link transmitting coherent signals.…”

Artificial intelligence for channel estimation in multicarrier systems for B5G/6G communications: a survey

“…This approach might be seen as an ML interpolation strategy with results similar to the MMSE, and some other discussed NN. A proposed similar channel estimation method using a multiple variable regression approach to design an ML algorithm that does not require any initial information or statistics about the channel is found in [330]. It uses the SGD algorithm for parameter optimization purposes.…”

Future OFDM-based Communication Systems Towards 6G and Beyond: Machine Learning Approaches

“…Training datasets are created offline under various environmental conditions where the received signals are recorded and used as training data for the system to learn. Furthermore, ML can be used to estimate channel impulse response (CIR) coefficients for equalization purposes [14]. In the future, communication using the OFDM signal will be transmitted over the air, space, or underwater channels to create a fully connected world.…”