Kolloquiumsvortrag 16. April 2024, Tobias Konheiser (Funda)

Bild Besprechungsraum 04.137
Bild der Präsentationsfläche

Implementation and Performance Evaluation of Service Blocking Queues in Hardware-in-the-Loop Systems

ZF is developing an autonomous driving system, which requires extensive testing of the developed devices and software on hardware-in-the-loop (HIL) systems. Therefore, a robust and high-performing   HIL system is essential. The purpose of a HIL system is to replay recorded data to the device-under-test. Recordings are loaded, processed and streamed to the device-under-test with real-time requirements. This streaming chain includes processing nodes and queues. This requires careful management of queue configurations. An overflow in the queue will result in packet loss, while an underflow may violate the real-time constraint. This study aims to develop and evaluate concepts for service blocking queues. These concepts block or pause the incoming service to a queue when necessary to avoid queue overflows and associated data loss. However, an out-of-the-box solution is not available and different approaches affect the behaviour and performance of
the system. Therefore, the developed concepts are evaluated against each other and against the existing system based on selected performance parameters in specific scenarios. The scenarios cover a wide range of situations, reflecting standard input data with varying numbers of parallel streams and bottleneck scenarios forcing queue overflows
or blockages. The developed service blocking queue concepts eliminate data loss in all scenarios, but introduce overhead, resulting in reduced system performance. However, the service blocking queue concept using a modified token-bucket approach proved to be the best solution, as the elimination of data loss justifies the additional overhead. This concept is proposed for implementation and deployment on the HIL system.

 

Zeit: 10:15 Uhr

Ort: Raum 04.137, Martensstr. 3, Erlangen

oder

Zoom-Meeting beitreten:
https://fau.zoom-x.de/j/68350702053?pwd=UkF3aXY0QUdjeSsyR0tyRWtLQ0hYUT09

Meeting-ID: 683 5070 2053
Kenncode: 647333