Real-time management of railway CPS secure administration of IoT and CPS infrastructure

摘要

Railways constitute a main mean of mass transportation. Public, private, and military settings are traversing long distances everyday. Railway controlling software must collect spatial information and effectively manage these systems. The need to manage the underlying embedded systems, brought forward by the wider adoption of pervasive computing, is vital in the context of safety- and security-critical applications. This work presents RailwaySPD, a management framework of embedded systems for railway settings. It implements an artificially intelligent reasoning process for controlling and configuring the system at runtime. Metrics for security, privacy, and dependability (SPD) are embodied along with semantic knowledge regarding the safety aspects of the setting. RailwaySPD is applied on a railway application and safeguards two distinct WSNs. The first outdoor network monitors the railway's routes. The second network is located inside a train's carriage that transports dangerous cargo. Two scenarios are modelled where the system is configured at runtime to counter cyber-attacks and manage safety-related events respectively.