Threat Intelligence Sharing Platforms for Autonomous Vehicle Security

Abstract

As connected, autonomous vehicles advance to market, cyber-physical security assumes profound importance. A system capable of protecting vehicle sensor data from cyber-security risks and physical-layer attacks for Intelligent Transportation Systems and Smart-City applications has been developed. The system is endowed with the capacity to render deep learning sensor data, immunize it in the time domain, recover if necessary, and run safe machine learning decision-making processes using a combination of robust deep neural networks immune to cyber-physical attacks. This cyber-physical reinforcement learning framework is designed to work on multiple connected autonomous vehicles simultaneously.  Besides cybersecurity, the challenge of protecting sensor data and providing privacy through cutting-edge privacy-preserving solutions for both hardware and software are considerable factors [1, 10] of complexity in security coordination between autonomous vehicles, smart city infrastructures, IT providers, cyber-insurance companies, and central and local governance levels. These data-protection security layers include IoT data routers, permissioned blockchains, and data transformer agents [1].