Blockchain and connected vehicles: recording trust during cyberattacks
A new paper combines distributed learning with evidence recording. Tests were carried out on research datasets.
Blockchain can be used to record trust and evidence in connected vehicle networks. That is its role in a new cyberattack detection paper published on 27 September in Cluster Computing.
Learning across the network
The proposed system combines federated learning, vehicle movement information and trust weighting for participants. Federated learning enables collaborative model training without necessarily pooling all raw data in one place.
The architecture includes on-chain and off-chain records, adversarial testing and consensus processes for managing the model. Evaluation used the UNSW NB15, CIC IDS2018 and VeReMi datasets.
Research evaluation, not a commercial service
The reported performance does not establish operation in a city’s vehicle fleet. Response latency, communication costs and practical data protection remain important. The study concerns transport security, not energy trading. It provides a concrete smart-city example of managing trust between connected nodes.
English translation of the Greek edition · Read in Greek