Estimation of Deadline Miss Rate for DAG Mixed Timer-Driven and Event-Driven Nodes
Daichi Yamazaki, Takuya Azumi · 2023
Autonomous-driving systems are becoming increasingly complex, making it more difficult to predict timing. Autonomous-driving applications, such as localization and path planning, can be represented by a Directed Acyclic Graph (DAG) consisting of timer-driven and event-driven nodes. Various estimation methods have been proposed for determining the response time of a DAG. However, most existing studies rely on worst-case execution time estimation, leading to pessimistic results. Additionally, there is a lack of research on DAGs that incorporate a combination of timer-driven and event-driven nodes. To address this issue, a proposed method estimates the response time and deadline miss rates considering the variation in execution time for DAGs with mixed node types. By considering execution time variation, the proposed method allows for multiple values of response time and deadline miss rate based on probability. When a sufficient number of cores are available, the proposed method accurately estimates the deadline miss rate.