Model-based optimization of ARINC-653 partition scheduling

Abstract

The architecture of ARINC-653 partitioned scheduling has been widely applied to avionics systems owing to its robust temporal isolation among applications. However, this partitioning mechanism causes the problem of how to optimize the partition scheduling of a complex system while guaranteeing its schedulability. In this paper, a model-based optimization approach is proposed. We formulate the problem as a parameter sweep application, which searches for the optimal partition scheduling parameters with respect to minimum processor occupancy via an evolutionary algorithm. An ARINC-653 partitioned scheduling system is modeled as a set of timed automata in the model checker UPPAAL. The optimizer tentatively assigns parameter settings to the models and subsequently invokes UPPAAL to verify schedulability as well as evaluate promising solutions. The parameter space is explored with an evolutionary algorithm that combines refined genetic operators and the self-adaptation of evolution strategies. The experimental results show the applicability of our optimization method.

Randomized Reachability Analysis in Uppaal: Fast Error Detection in Timed Systems

“…We consider more schedulability systems modelled as stopwatch automata. Table 6 shows experiments for two different sets of schedulability problems: ARINC-653 partition scheduling of integrated modular avionics systems [11] (denoted as IMAOptim) and schedulability of Java bytecode systems, originating from TetaSARTS project [21], that are encoded as networks of automata and represent the original layered structure of Java bytecode systems. Our randomized methods discover the target state within 20 seconds even for a huge system with almost 12 thousands of locations, where other techniques are either not applicable or run out of memory.…”

On methods and tools for rigorous system design

“…Papers [3,18] contribute to the design of correct-by construction high-level models by defining a highlevel modelling formalism [18] and by providing an approach for debugging CPS models [3]. Papers [9,14,22,24,28] contribute to the design and proof of domain-specific abstractions. They provide techniques for ensuring the correctness of randomised consensus protocols [9], program block parallelisation [14], usage control policies [22], and for ensuring optimality of partition schedules [24] and energy consumption [28].…”

“…In addition, the paper describes a prototype implementation of the language in Java and illustrates the application of the approach to model 3 different examples from different domains. -The paper "Model-based optimization of ARINC-653 partition scheduling" by Han, Zhai, Nielsen, and Nyman [24] is an extension of the MeTRiD 2018 paper [23], by the same authors. This work introduces a framework for generating optimal ARINC-653 partitioned schedules, i.e.system partitions scheduled in mutual exclusion in order ensure temporal isolation of applications in safety-critical systems.…”