Digital Twins and Testbeds for Supporting AI Research with Autonomous Vehicle Networks

TL;DR

The paper investigates how digital twins can accelerate AI-enabled development for autonomous vehicle networks by comparing simulation, SITL DTs, HITL sandboxes, and physical testbeds, and by presenting an integrated AERPAW DT platform that couples virtual development with RW validation. It shows that SITL DTs, when calibrated with RW measurements and simulations, offer an efficient, scalable environment for AI experimentation in AVN, reducing cost and iteration time compared to pure RW testing. A key case study on AI-aided signal source localization demonstrates how DT-generated data, combined with RW calibration, improves sim-to-RW transfer for localization algorithms like particle filters and CNN-based fingerprinting. The work highlights a practical, reproducible workflow where DT development precedes sandbox testing and final RW validation, enabling rapid, safe, and scalable AI innovation for AVN.

Abstract

Digital twins (DTs), which are virtual environments that simulate, predict, and optimize the performance of their physical counterparts, hold great promise in revolutionizing next-generation wireless networks. While DTs have been extensively studied for wireless networks, their use in conjunction with autonomous vehicles featuring programmable mobility remains relatively under-explored. In this paper, we study DTs used as a development environment to design, deploy, and test artificial intelligence (AI) techniques that utilize real-world (RW) observations, e.g. radio key performance indicators, for vehicle trajectory and network optimization decisions in autonomous vehicle networks (AVN). We first compare and contrast the use of simulation, digital twin (software in the loop (SITL)), sandbox (hardware-in-the-loop (HITL)), and physical testbed (PT) environments for their suitability in developing and testing AI algorithms for AVNs. We then review various representative use cases of DTs for AVN scenarios. Finally, we provide an example from the NSF AERPAW platform where a DT is used to develop and test AI-aided solutions for autonomous unmanned aerial vehicles for localizing a signal source based solely on link quality measurements. Our results in the physical testbed show that SITL DTs, when supplemented with data from RW measurements and simulations, can serve as an ideal environment for developing and testing innovative AI solutions for AVNs.

Figures (6)

• Figure 1: DT assisted AVN technology development cycle from idea to rw deployment, including continuous refinement of DT through validation and tuning.
• Figure 2: Interactions between pt, dt, sandbox, and simulation for experiment development and testing.
• Figure 3: Interaction of various vehicle and fixed-node software through the environment emulator in the AVN-DT. AI inference engines at each vehicle and fixed node, as well as a centralized entity, are also illustrated.
• Figure 4: AVN application examples where AI techniques can be developed in a DT environment for jointly optimizing AV trajectory control, swarm formation, and wireless network parameters before they get deployed in the RW.
• Figure 5: rssi variation in the dt and the pt across the UAV's trajectory. The deep fades in the rw environment adversely affect the localization performance if the training is performed exclusively based on the dt environment.