Bridge to Real Environment with Hardware-in-the-loop for Wireless Artificial Intelligence Paradigms

TL;DR

This paper introduces a hardware-in-the-loop system for testing AI and vehicular communication solutions in both simulated and real environments, reducing costs and risks associated with real-world testing.

Contribution

The paper presents a novel hardware-in-the-loop platform for integrated testing of AI, services, and HD map data in vehicular networks, bridging simulation and real-world deployment.

Findings

Successful implementation of hardware-in-the-loop for AI testing in vehicular networks

Enhanced reliability of testing by combining simulation and real-world data

Potential to reduce costs and risks in deploying vehicular AI solutions

Abstract

Nowadays, many machine learning (ML) solutions to improve the wireless standard IEEE802.11p for Vehicular Adhoc Network (VANET) are commonly evaluated in the simulated world. At the same time, this approach could be cost-effective compared to real-world testing due to the high cost of vehicles. There is a risk of unexpected outcomes when these solutions are implemented in the real world, potentially leading to wasted resources. To mitigate this challenge, the hardware-in-the-loop is the way to move forward as it enables the opportunity to test in the real world and simulated worlds together. Therefore, we have developed what we believe is the pioneering hardware-in-the-loop for testing artificial intelligence, multiple services, and HD map data (LiDAR), in both simulated and real-world settings.