Autonomous Systems in the Internet of Vehicles
by Balamurugan Balusamy, Sandeep Kumar Mathivanan, Prabhu Jayagopal, S. K. B. Sangeetha, Ali Kashif Bashir
7Reinforcement Learning–Driven Multisensor Fusion for Real-Time Navigation in Intelligent and Opportunistic Vehicular Networks
Mahalakshmi1*, Suma T.2, Soya Mathew3 and Nitya S.4
1Department of Mathematics, Nitte Meenakshi Institute of Technology, Bengaluru, Karnataka, India
2Department of Mathematics, New Horizon College of Engineering, Bengaluru, Karnataka, India
3Department of Mathematics, Kristu Jayanti College (Deemed to be University), Bengaluru, Karnataka, India
4School of Electronics and Communication, REVA University, Bengaluru, Karnataka, India
Abstract
Conventional models of navigation foresee traffic congestion, calculate routes optimally, and avoid collisions inefficiently because they have static routing processes and heuristics decision-making. Such a system often leads to suboptimal trip times and huge safety risks due to high latency, poor use of real-time multisensor information, and lack of flexibility toward immediate traffic fluctuation. The present study proposes a multisensor fusion architecture from GPS and vehicle-to-everything communication data for enhanced navigation to bridge these deficiencies. The proposed architecture optimizes route choice, predicts congestion, and improves crash avoidance through the integration of graph neural networks and deep Q-networks in a reinforcement learning paradigm. To prevent data futility during its transmission, a self-attention mechanism is adopted to minimize network latency and signal loss. The SUMO (simulation ...
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