Ensuring Data Freshness in Multi-Rate Task Chains Scheduling

TL;DR

This paper presents a novel task scheduling framework that enforces data freshness constraints in multi-rate task chains, improving efficiency and responsiveness in safety-critical autonomous systems.

Contribution

It introduces a data-driven scheduling approach with task offsets and a formal methodology for decomposing data dependency graphs, avoiding artificial latency and maintaining schedulability.

Findings

Enforces end-to-end data freshness without added latency.

Maintains 100% schedulability capacity of Global EDF.

Eliminates redundant sampling overhead.

Abstract

In safety-critical autonomous systems, data freshness presents a fundamental design challenge. While the Logical Execution Time (LET) paradigm ensures compositional determinism, it often does so at the cost of injected latency, degrading the phase margin of high-frequency control loops. Furthermore, mapping heterogeneous, multi-rate sensor fusion requirements onto rigid task-centric schedules typically implies in resource-inefficient oversampling. This paper proposes a Task-based scheduling framework extended with data freshness constraints. Unlike traditional models, scheduling decisions are driven by the lifespan of data. We introduce task offset based on the data freshness constraint to order data production in a Just-in-Time (JIT) fashion: the completion of the production of data with strictest data freshness constraint is delayed to the instant its consumers will be ready to use it. This allows for flexible task release offsets. We introduce a formal methodology to decompose Data Dependency Graphs into Dominant Paths by tracing the strictest data freshness constraints backward from the actuators. Based on this decomposition, we propose a Consensus Offset Search algorithm that synchronizes shared producers and private predecessors. This approach enforces end-to-end data freshness without the artificial latency of LET buffering. We formally prove that this offset-based alignment preserves the 100\% schedulability capacity of Global EDF, ensuring data freshness while eliminating the computational overhead of redundant sampling.