Simulating Cloud Environments of Connected Vehicles for Anomaly Detection

• DOI: 10.4271/2024-01-2996
• Publication Date: 2024-07-02T02:15:32Z
• Authors: Matthias Weiß; Johannes Stümpfle; Falk Dettinger; Nasser Jazdi; Michael Weyrich
• Keywords: FOS: Computer and information sciences; B.8.2; Emerging Technologies (cs.ET); Computer Science - Distributed, Parallel, and Cluster Computing; C.2.4; Computer Science - Emerging Technologies; Distributed, Parallel, and Cluster Computing (cs.DC); B.8.2; C.2.4

<div class="section abstract"><div class="htmlview paragraph">The emergence of connected vehicles is driven by increasing customer and regulatory demands. To meet these, more complex software applications, some which require service-based cloud edge backends, are developed. Due to the short lifespan software, it becomes necessary keep these environments their applications up date with security updates new features. However, as behavior introduced system, high complexity interdependencies between components can lead unforeseen side effects in other system parts. As such, challenging recognize whether deviations intended occurring, ultimately resulting higher monitoring efforts slower responses errors.</div><div paragraph">To overcome this problem, a simulation environment running parallel proposed. This approach enables live comparison simulated real behavior. Therefore, concept developed mirroring existing into simulation. collect data, an observability platform presented, capturing telemetry architecture information. Subsequently, designed that converts model simulates its dynamic workload utilizing captured communication data.</div><div proposed evaluated real-world application scenario for electric vehicle charging: Vehicles apply unoccupied charging station at service backend, latter manages all incoming requests performs assignment. Benchmarks conducted comparing collected data results under different loads injected faults. The show regular mirrored well misbehavior due fault injection visible, indicating simulations promising source anomaly detection during operation.</div></div>