Autonomous Systems in the Internet of Vehicles
by Balamurugan Balusamy, Sandeep Kumar Mathivanan, Prabhu Jayagopal, S. K. B. Sangeetha, Ali Kashif Bashir
3Design of ML-CASF: Multilayer Context-Aware Sensor Fusion for Autonomous Vehicles in the Internet of Vehicles
Sukumar R.1* and Sathishkumar V.E.2
1Department of Computational Intelligence, SRM Institute of Science and Technology, Kattankulathur, India
2School of Engineering and Technology, Sunway University, Petaling Jaya, Selangor Darul Ehsan, Malaysia
Abstract
With perception optimization and decision-making accuracy, sensor fusion techniques have significantly enhanced autonomous Internet of Vehicles (IoV) systems. Real-time processing, heterogeneous environmental conditions, and faulty sensors plague the existing techniques. In this study, a novel multilayer context-aware sensor fusion (ML-CASF) model based on graph attention network (GAT) adaptive weighting is proposed to fuse LiDAR, radar, camera, and vehicle-toeverything communication data. Besides that, a Bayesian uncertainty estimation module provides robustness against malicious attacks and sensor noise. ML-CASF achieves 96.4% localization accuracy, 94.7% object detection accuracy, and 92.1% collision avoidance success rate, according to IoV dataset experiments. These measures are the extent to which ML-CASF works to provide secure and reliable autonomous navigation in realistic urban environments. These outcomes show the capability of ML-CASF in providing safe and accurate autonomous navigation in challenging city environments.
Keywords: Sensor fusion, graph attention network, IoV systems, LiDAR, radar, camera, V2X ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access