Lidar aided dynamic digital twins

Abstract

The continuous advancement of wireless communication technologies, especially the transition towards sixth generation (6G) networks, demands new methodologies for network planning, optimization, and intelligent resource allocation. The increasing complexity of wireless environments, characterized by dynamic obstacles, real-time user mobility, and varying signal propagation conditions, necessitates the integration of artificial intelligence (AI), machine learning (ML), and digital twin (DT) technologies. Digital twins provide an advanced approach by creating high-fidelity virtual replicas of physical networks, enabling extensive performance evaluation, real-time monitoring, and predictive network optimization before real-world deployment.

This research presents a digital twin based simulation framework that models dynamic physical environments using NVIDIA Isaac Sim which is a powerful simulation platform designed for AI-based robotics and environmental modeling. By leveraging LiDAR sensor data, the framework captures real time environmental dynamics, such as structural changes, object movements, and urban variations, and integrates these into NVIDIA Sionna, which is a ray tracing engine for wireless communication simulation. The machine learning based processing pipeline extracts critical environmental features and embeds them into the ray tracing environment to enhance the accuracy of wireless signal propagation modeling.

By incorporating ML driven techniques, the proposed digital twin framework provides a virtual testbed for AI powered radio access networks (AI-RAN), allowing for real time network adaptation, predictive analysis, and intelligent decision making. The developed system enables wireless communication researchers and engineers to simulate, evaluate, and optimize 6G networks efficiently while reducing the reliance on costly and time-consuming real-world experimentation.

The results of this study demonstrate that ML enhanced, digital twin based simulations significantly improve network adaptability, signal prediction accuracy, and energy efficiency. The proposed framework serves as a scalable, AI powered solution for optimizing next-generation wireless networks, enabling future 6G systems to handle the increasing demands of ultra reliable, high speed and low latency communication in complex, real world environments.