Started: 2021. 06. 01.

Coordinator: Balázs Sonkoly

Members: 5 seniors, 5 PhD student, 2 MSc students

Abstract: Fog computing is an emerging concept extending traditional cloud computing by deploying compute resources closer to end devices. This approach, closely integrated with 5G carrier-networks, enables several future services, such as coordination of autonomous vehicles by adaptive data collection methods through on-vehicle sensors, safe platooning or cooperative adaptive cruise control, and fog-assisted detection of misbehaving drivers, crossing pedestrians or even vehicular botnets. Mission critical applications requiring the resources of fog computing must complete tasks within a predefined deadline. Furthermore, secure and privacy-preserving mechanisms are advocated that prevent the transferred data from being monitored, filtered, or tampered with by an adversary (e.g., an eavesdropper, man-in-the-middle). Here, we focus on the collision avoidance function of a self-driving application that forecasts imminent emergencies within milliseconds by collecting and processing data from multiple cars, and making decisions and transmitting data back to vehicles to allow them to react in time. In this project, as a basis for such diverse applications, we propose and implement a new fog computing based enabler architecture. In particular, the envisioned architecture encompasses novel methods to provide the applications with sufficient resources to meet strict latency, bandwidth, and computing requirements. The platform logically consists of two layers: (i) resource orchestration; (ii) privacy-preserving and secure data management and AI-powered analytics. The first layer (i) includes new operating mechanisms that guarantee optimal resource utilization while meeting application-specific requirements. On the other hand,  (ii) includes features enabling real-time processing of large amounts of data in a geographically distributed environment  in a secure and privacy-preserving manner in order to protect the users, i.e., drivers, from malicious attacks and information leaks.

Acknowledgement: This work is supported by the Ministry of Innovation and Technology of Hungary from the National Research, Development and Innovation Fund through project 2019-2.1.11-TÉT-2020-00183 under the 2019-2.1.11-TÉT funding scheme.

THE AMOUNT OF FUNDING:
4.7 MILLION HUF