Related Experiment Videos
Zero knowledge verifiable, semi asynchronous federated learning for trajectory prediction on permissioned blockchain
K Raveendra Reddy1, A Muralidhar2
1School of Computer Science and Engineering, Vellore Institute of Technology, Chennai, 600127, Tamil Nadu, India.
Scientific Reports
|April 20, 2026
Summary
ChainDrive-FL-VRA enhances vehicle trajectory prediction using federated learning on a blockchain. It ensures privacy and robustness against unreliable participants and connectivity issues.
Area of Science:
- Distributed Systems and Networking
- Artificial Intelligence and Machine Learning
- Cybersecurity and Privacy
Background:
- Vehicle trajectory prediction in Internet-of-Vehicles (IoV) faces challenges with sensitive data, intermittent connectivity, and untrusted participants.
- Existing federated learning approaches struggle to ensure data privacy and model integrity in decentralized, partially trusted environments.
Purpose of the Study:
- To develop a secure and robust federated learning framework for vehicle trajectory prediction in IoV.
- To address challenges of intermittent connectivity, participant trust, and data privacy using blockchain and zero-knowledge proofs.
Main Methods:
- Implemented ChainDrive-FL-VRA, a system coordinating semi-asynchronous federated learning on a permissioned consortium ledger with Practical Byzantine Fault Tolerance (PBFT).
- Kept raw trajectory data and model updates off-chain, utilizing on-chain headers with commitments, hashes, and zero-knowledge proofs for data integrity and privacy.
- Employed staleness- and reputation-aware robust weighting for validator updates and a contextual-bandit trigger for adaptive aggregation timing under client churn.
Main Results:
- Achieved improved prediction accuracy, demonstrated by lower Average Displacement Error (ADE), Final Displacement Error (FDE), and Root Mean Square Error (RMSE) on NGSIM US-101 and I-80 datasets.
- Showcased enhanced robustness against data staleness and anomalous updates compared to traditional methods.
- Maintained minimal on-chain storage footprint, with kilobyte-scale artifacts per update and aggregation event.
Conclusions:
- ChainDrive-FL-VRA effectively enhances vehicle trajectory prediction by integrating federated learning with blockchain technology for improved security, privacy, and robustness.
- The system's design successfully mitigates challenges posed by intermittent connectivity and untrusted participants in IoV environments.
- The framework offers a scalable and efficient solution for collaborative learning on sensitive trajectory data.
Related Concept Videos
01:26Orthogonal Trajectories
282
Orthogonal trajectories describe the geometric relationship between two families of curves that intersect each other at right angles. One illustrative case involves a family of parabolas that open sideways along the x-axis. These curves share a common shape but differ by a scaling parameter, resulting in a set of curves that all pass through the origin and widen at different rates.Determining Orthogonal TrajectoriesTo identify the orthogonal trajectories for these parabolas, the first step...
282
01:12Observational Learning
1.5K
Albert Bandura's observational learning, also known as imitation or modeling, occurs when a person observes and imitates another's behavior. It is a quicker process than operant conditioning. A well-known example is the Bobo doll study, where children who saw an adult acting aggressively towards the doll were more likely to act aggressively when left alone, compared to those who observed a nonaggressive adult. Many psychologists view observational learning as a form of latent learning...
1.5K