A deep learning-aided temporal spectral ChannelNet for IEEE 802.11p-based channel estimation in vehicular communications

Abstract

In vehicular communications using IEEE 802.11p, estimating channel frequency response (CFR) is a remarkably challenging task. The challenge for channel estimation (CE) lies in tracking variations of CFR due to the extremely fast time-varying characteristic of channel and low density pilot. To tackle such problem, inspired by image super-resolution (ISR) techniques, a deep learning-based temporal spectral channel network (TS-ChannelNet) is proposed. Following the process of ISR, an average decision-directed estimation with time truncation (ADD-TT) is first presented to extend pilot values into tentative CFR, thus tracking coarsely variations. Then, to make tentative CFR values accurate, a super resolution convolutional long short-term memory (SR-ConvLSTM) is utilized to track channel extreme variations by extracting sufficiently temporal spectral correlation of data symbols. Three representative vehicular environments are investigated to demonstrate the performance of our proposed TS-ChannelNet in terms of normalized mean square error (NMSE) and bit error rate (BER). The proposed method has an evident performance gain over existing methods, reaching about 84.5% improvements at some high signal-noise-ratio (SNR) regions.

A Survey on Deep Learning Based Channel Estimation in Doubly Dispersive Environments

“…Temporal spectral ChannelNet (TS-ChannelNet) [31] is based on applying average decision-directed with time truncation (ADD-TT) interpolation to the received OFDM frame. Thereafter, accurate estimation is achieved by implementing SR-ConvLSTM network to track doubly-dispersive channel variations by learning the vehicular channel's time and frequency correlations.…”

“…Here, historical data can be utilized to forecast future observations[56]. Motiviated by this possibility, the authors in[31] apply SR-ConvLSTM network in addition to the ADD-TT interpolation, where convolutional layers get added to the LSTM network to capture more doubly-dispersive channel features. Consequently,VOLUME 4, 2016…”

CNN aided Weighted Interpolation for Channel Estimation in Vehicular Communications

“…TS-ChannelNet [15] has been proposed to tackle the problem of channel estimation in vehicular communications. It is based on applying ADD-TT interpolation to the received OFDM frame.…”

“…Motivated by the fact that the vehicular time-variant channel can be modeled as a time-series forecasting problem, where historical data can be used to predict future observations [18]. The authors in [15] apply SR-ConvLSTM network on top of the ADD-TT interpolation, where convolutional layers are added to the LSTM network in order to capture more vehicular channel features, and thus improving the estimation performance. Therefore, the ADD-TT estimated channel for the whole received frame is modeled as a low resolution image, and then SR-ConvLSTM network is employed after the ADD-TT interpolation.…”

“…ChannelNet estimator suffers mainly from high computational complexity and considerable performance degradation in high mobility vehicular scenarios. Temporal spectral Chan-nelNet (TS-ChannelNet) has been proposed in [15], where average decision-directed with time truncation (ADD-TT) interpolation that is based on the demodulation and averaging of each received symbol. After that, super resolution convolutional long short-term memory (SR-ConvLSTM) is employed to improve ADD-TT interpolation accuracy.…”