Deep Learning-Based Data Fusion of 6G Sensing and Inertial Information for Target Positioning: Experimental Validation

The sixth-generation (6G) cellular technology will be deployed with a key feature of Integrated Sensing and Communication (ISAC), allowing the cellular network to map the environment through radar sensing on top of providing communication services. In this regard, the entire network can be considered as a sensor with a broader Field of View (FoV) of the environment, assisting in both the positioning of active and detection of passive targets. On the other hand, the non-3GPP sensors available on the target can provide additional information specific to the target that can be beneficially combined with ISAC sensing information to enhance the overall achievable positioning accuracy. In this paper, we first study the performance of the ISAC system in terms of its achievable accuracy in positioning the mobile target in an indoor scenario. Second, we study the performance gain achieved in the ISAC positioning accuracy after fusing the information from the target's non-3GPP sensors. To this end, we propose a novel data fusion solution based on the deep learning framework to fuse the information from ISAC and non-3GPP sensors. We validate our proposed data fusion and positioning solution with a real-world ISAC Proof-of-Concept (PoC) as the wireless infrastructure, an Automated Guided Vehicle (AGV) as the target, and the Inertial Measurement Unit (IMU) sensor on the target as the non-3GPP sensor. The experimental results show that our proposed solution achieves an average positioning error of $3~\textrm{cm}$, outperforming the considered baselines.

Digital Twin of Industrial Networked Control System based on Value of Information

The paper examines a scenario wherein sensors are deployed within an Industrial Networked Control System, aiming to construct a digital twin (DT) model for a remotely operated Autonomous Guided Vehicle (AGV). The DT model, situated on a cloud platform, estimates and predicts the system's state, subsequently formulating the optimal scheduling strategy for execution in the physical world. However, acquiring data crucial for efficient state estimation and control computation poses a significant challenge, primarily due to constraints such as limited network resources, partial observation, and the necessity to maintain a certain confidence level for DT estimation. We propose an algorithm based on Value of Information (VoI), seamlessly integrated with the Extended Kalman Filter to deliver a polynomial-time solution, selecting the most informative subset of sensing agents for data. Additionally, we put forth an alternative solution leveraging a Graph Neural Network to precisely ascertain the AGV's position with a remarkable accuracy of up to 5 cm. Our experimental validation in an industrial robotic laboratory environment yields promising results, underscoring the potential of high-accuracy DT models in practice.