Abstract:End-to-end wireless communication is new concept expected to be widely used in the physical layer of future wireless communication systems (6G). It involves the substitution of transmitter and receiver block components with a deep neural network (DNN), aiming to enhance the efficiency of data transmission. This will ensure the transition of autonomous vehicles (AVs) from self-autonomy to full collaborative autonomy, that requires vehicular connectivity with high data throughput and minimal latency. In this article, we propose a novel neural network receiver based on transformer architecture, named TransRx, designed for vehicle-to-network (V2N) communications. The TransRx system replaces conventional receiver block components in traditional communication setups. We evaluated our proposed system across various scenarios using different parameter sets and velocities ranging from 0 to 120 km/h over Urban Macro-cell (UMa) channels as defined by 3GPP. The results demonstrate that TransRx outperforms the state-of-the-art systems, achieving a 3.5dB improvement in convergence to low Bit Error Rate (BER) compared to convolutional neural network (CNN)-based neural receivers, and an 8dB improvement compared to traditional baseline receiver configurations. Furthermore, our proposed system exhibits robust generalization capabilities, making it suitable for deployment in large-scale environments.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to improve the performance of receivers in the sixth - generation cellular vehicle - to - everything (6G - V2X) communication system, especially in high - mobility and complex channel environments. Specifically, the authors propose a deep neural network receiver (TransRx) based on the Transformer encoder, aiming to replace traditional receiver components to improve data transmission efficiency, reduce the bit error rate (BER), and ensure low latency and high reliability. ### Main Problems and Solutions 1. **Limitations of Existing Technologies** - Traditional receiver designs rely on multiple independent functional modules (such as channel estimation, equalization, demodulation, etc.). These modules are optimized individually but lack overall collaborative optimization. - Existing neural receivers based on convolutional neural networks (CNN) perform well in handling low - to - medium signal - to - noise ratio (SNR) conditions, but are not as good as the minimum mean square error (MMSE) receiver in high SNR conditions. 2. **Proposed Solutions** - **TransRx Architecture**: A deep neural network receiver based on the Transformer encoder that can handle the uplink scenarios in 6G - V2X communication, especially processing frequency - domain signals and predicting the log - likelihood ratio (LLR). This architecture utilizes the attention mechanism and can capture long - distance data dependencies, thereby improving performance. - **Multi - task Joint Optimization**: TransRx integrates tasks such as channel estimation, equalization, and demodulation in a unified neural network framework through end - to - end learning, achieving better collaborative optimization. 3. **Experimental Verification** - The paper verifies the performance of TransRx by conducting tests under different wireless channel characteristics (such as UMa, CDL channel models) and vehicle speeds (0 to 120 km/h). The results show that TransRx outperforms existing state - of - the - art methods in all cases, including CNN - based DeepRx and LS estimators. - Especially in the image transmission application scenario, TransRx exhibits a higher peak signal - to - noise ratio (PSNR), demonstrating its superiority in practical applications. ### Formula Summary - **Channel Estimation**: \[ \hat{H} = y_{i,j} x^*_{i,j} \] where \( y_{i,j} \) is the received symbol, \( x_{i,j} \) is the transmitted symbol, \( H_{i,j} \) is the channel matrix, and \( n_{i,j} \) is the noise. - **LMMSE Equalization**: \[ \hat{y}_{i,j} = (\hat{H}^H_{i,j} \hat{H}_{i,j} + \hat{\sigma}^2 I)^{-1} \hat{H}^H_{i,j} y_{i,j} \] where \( \hat{H}^H \) represents the conjugate transpose, \( I \) is the identity matrix, and \( \hat{\sigma}^2 \) is the estimated noise power. - **LLR Calculation**: \[ LLR_{i,j,k} = \log \frac{P(x_k = 0|\hat{y}_{i,j})}{P(x_k = 1|\hat{y}_{i,j})} \] where \( P(x_k = 0|\hat{y}_{i,j}) \) and \( P(x_k = 1|\hat{y}_{i,j}) \) are the probabilities of the transmitted bit being 0 and 1 respectively given the received symbol \( \hat{y}_{i,j} \). - **Loss Function**: \[ BCE = -\frac{1}{N} \sum_{i = 1}^{n} \left[ y_i \cdot \log(P(y_i))+(1 - y_i) \cdot \log(1 - P(y_i)) \right] \] \[ Loss = 1-\frac{BCE}{\ln(2)} \] where \( N \) is the batch size and \( n \) is each.

TransRx-6G-V2X : Transformer Encoder-Based Deep Neural Receiver For Next Generation of Cellular Vehicular Communications

DeepRx: Fully Convolutional Deep Learning Receiver

ViT LoS V2X: Vision Transformers for Environment-aware LoS Blockage Prediction for 6G Vehicular Networks

A Two-Stage Deep Learning Receiver for High Throughput Power Efficient LEO Satellite System With Varied Operation Status

Modular Neural Network-Based Models of High-Speed Link Transceivers

Joint Vehicle Connection and Beamforming Optimziation in Digital Twin Assisted Integrated Sensing and Communication Vehicular Networks

Deep Neural Network Based Resource Allocation for V2X Communications

Design of a Standard-Compliant Real-Time Neural Receiver for 5G NR

Deep Learning for Joint Adaptations of Transmission Rate and Payload Length in Vehicular Networks

Joint Transceiver Optimization for Wireless Communication PHY with Convolutional Neural Network

Transformer Masked Autoencoders for Next-Generation Wireless Communications: Architecture and Opportunities

Deep Reinforcement Learning Assisted Hybrid Precoding for V2I Communications with Doppler Shift and Delay Spread

CRNN-ResNet: Combined CRNN and ResNet Networks for OFDM Receivers

RVE-CV2X: A Scalable Emulation Framework for Real-Time Evaluation of CV2X-Based Connected Vehicle Applications

End-to-End Optimized Transmission over Dispersive Intensity-Modulated Channels Using Bidirectional Recurrent Neural Networks

Semantically Optimized End-to-End Learning for Positional Telemetry in Vehicular Scenarios

Joint Transceiver Optimization for Wireless Communication PHY Using Neural Network.

AI enlightens wireless communication: A transformer backbone for CSI feedback

When Visible Light (backscatter) Communication Meets Neuromorphic Cameras in V2X

Neural Network-based OFDM Receiver for Resource Constrained IoT Devices

End-to-End Learning for Uplink MU-SIMO Joint Transmitter and Non-Coherent Receiver Design in Fading Channels.