Abstract:This article introduces an approach to harness digital twin (DT) technology in the realm of integrated sensing and communications (ISACs) in sixth-generation (6G) Internet of Everything (IoE) applications. We consider moving targets in a vehicular network and use DT to track and predict the motion of the vehicles. After predicting the location of the vehicle at the next time slot, the DT designs the assignment and beamforming for each vehicle. The real-time sensing information is then utilized to update and refine the DT, enabling further processing and decision making. In the DT, an extended Kalman filter (EKF) is used for the precise motion prediction. This model incorporates a dynamic Kalman gain, which is updated at each time slot based on the received echo signals. The state representation encompasses both the vehicle motion information and the error matrix, with the posterior Cram & eacute;r-Rao bound (PCRB) employed to assess sensing accuracy. We consider a network with two roadside units (RSUs), and the vehicles need to be allocated to one of them. To optimize the overall transmission rate while maintaining acceptable sensing accuracy, an optimization problem is formulated. Since, it is generally hard to solve the original problem, the Lagrange multipliers and fractional programming are employed to simplify this optimization problem. To solve the simplified problem, this article introduces both the greedy and heuristic algorithms by optimizing both the vehicle assignments and predictive beamforming. The optimized results are then transferred back to the real space for ISAC applications. Recognizing the computational complexity of the greedy and heuristic algorithms, a bidirectional long short-term memory (LSTM)-based recurrent neural network (RNN) is proposed for efficient beamforming design within the DT. Simulation results demonstrate the effectiveness of the DT-based ISAC network. Notably, the LSTM-based RNN method achieves similar transmission rates as the heuristic algorithm but with significantly reduced computational complexity.

TIME-SEQUENCE CHANNEL INFERENCE FOR BEAM ALIGNMENT IN VEHICULAR NETWORKS

Remote Channel Inference for Beamforming in Ultra-Dense Hyper-Cellular Network.

Enhancing Mmwave Beam Prediction Through Deep Learning with Sub-6 GHz Channel Estimate

Deep learning-based beam alignment in mmWave vehicular networks

Predictive Beamforming for Integrated Sensing and Communication in Vehicular Networks: A Deep Learning Approach

Cooperative Localisation for Multi-Rsu Vehicular Networks Based on Predictive Beamforming

Mixed-Timescale Deep-Unfolding for Joint Channel Estimation and Hybrid Beamforming

Learning-Aided Beam Prediction in mmWave MU-MIMO Systems for High-Speed Railway

Joint Channel Estimation and Hybrid Beamforming Via Deep-Unfolding.

Radar Aided mmWave Vehicle-to-InfrastructureLink Configuration Using Deep Learning

Channel Estimation Based on Deep Learning in Vehicle-to-Everything Environments

Bayesian Predictive Beamforming for Vehicular Networks: A Low-Overhead Joint Radar-Communication Approach

Time-Varying Channel Prediction for RIS-Assisted MU-MISO Networks Via Deep Learning

Beam Alignment for Millimetre Wave Links with Motion Prediction of Autonomous Vehicles

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

Deep Learning Based Predictive Beamforming Design

Learning-Based Remote Channel Inference: Feasibility Analysis and Case Study

Integrated Sensing and Communications for V2I Networks: Dynamic Predictive Beamforming for Extended Vehicle Targets

Low Overhead Beam Alignment for Mobile Millimeter Channel Based on Continuous-Time Prediction

Deep CLSTM for Predictive Beamforming in Integrated Sensing and Communication-enabled Vehicular Networks

Radar-assisted Predictive Beamforming for Vehicle-to-Infrastructure Links