Partition Scheduling on an IMA Platform with Strict Periodicity and Communication Delays

Integrated Modular Avionics (IMA) architectures employ a high-integrity, partitioned environment that hosts multiple avionics functions of different criticalities on a shared computing platform. IMA provides benefits of flexibility and scalability and allow for weight savings. However the complexity of the integration process is in-creased due to the sharing of resources. One of the main problem encountered during the integration process is to find a proper scheme for scheduling avionics functions onto offered processing units while ensuring at the same time execution, resource and safety demands. In this paper, we present a MILP formulation for the scheduling of avionics functions on the resources of an IMA architecture. This mixed integer linear program-ming formulation takes into account resource constraints as well as temporal constraints related to ensuring a fea-sible schedule for the strictly periodic avionics functions. The objective function is to maximize the evolution margin for the functions, so that spare resources can be allocated to meet the resource demand growth of the hosted func-tions. 1