A Software-Based Monitoring Framework for Time-Space Partitioned Avionics Systems

Abstract

Recently, avionics systems have evolved into a time and space partitioning (TSP)-based integrated modular avionics (IMA) structure for integration into a single system from a variety of existing independently configured federated systems. The TSP-based IMA architecture is suitable for solving size, weight, and power problems in avionics systems. Partitioning real-time operating systems (RTOSs) to support TSP-based IMA have been researched, and the international aviation industry has established the ARINC 653 standard for a partitioning RTOS. The ARINC 653 standard has defined the health monitoring (HM) function for debugging. However, the HM of the ARINC 653 standard does not support monitoring and debugging functions, such as snapshot, cycle, and, redundancy monitor, which makes the system development hard. To this end, the purpose of this paper is to introduce a monitoring framework that supports high reliability and stability for RTOS and application software based on TSP structure used in avionics systems. The proposed monitoring framework is designed for Qplus-AIR, an RTOS based on the TSP structure that conforms to the ARINC 653 for aircraft systems. It is also applicable to other RTOSs based on TSP structure that does not conform to ARINC 653. It supports monitoring functions, such as snapshot, trigger, and cycle as well as various debugging functions. It also supports debugging and monitoring operations under the redundancy of avionics systems, and minimizes the intrusive effect, which is a disadvantage of the software-based debugging approach. These functionalities enable avionics system developers to monitor and measure the performance of TSP structure-based RTOS and application software in flight control system for unmanned aerial vehicles. Our evaluation results show that the proposed monitoring framework is suitable for monitoring and debugging of RTOS and application software based on TSP structure.