Schedulability test for IMA systems based on mixed integer linear programming formulation

Abstract

Integrated modular avionics (IMA) architecture is widely adopted for the design of modern aircraft. It simplifies the system development process and improves the system security and reliability. In IMA systems, avionics applications are packed into various partitions, and integrated into a standard computing platform. How to determine the schedulability of systems is one of the key problems. In this paper, using the characters of avionics systems, a partition model with a strict period is built, and constraints in space, time, and communication are analyzed. Based on the mixed integer linear programming formulation, a solution to determine the schedulability of IMA systems is presented. Experience reveals that this solution not only determines the system schedulability, but also achieves the required minimum number of modules and guides the design of IMA systems.