Jitter Propagation in Task Chains

Chains of tasks are ubiquitous and used in a broad spectrum of applications. In these chains, tasks execute according to their timing. Then, they communicate by writing to and reading from shared memory. The schedule of tasks and the read/write instants are naturally subject to uncertainties (variability in the execution time, interference due to shared resources of higher priority tasks, etc.). Despite the impact of uncertainties, we believe that current analysis of task chains cannot handle them properly. In this paper, we borrow the notion of jitter to model uncertainties and we propose a novel event model that explicitly captures jitter in read and write operations, decoupled from task scheduling. We develop a (linear-time complexity) compositional analysis framework that tracks how this jitter propagates across chains and impacts metrics such as reaction time, data age, and end-to-end latency. Our model supports arbitrary communication paradigms (e.g., implicit, LET, mid-execution) and is applicable to the analysis of real-world frameworks such as ROS2 without requiring intrusive changes.