Synthesis of safe, QoS extendible, application specific schedulers for heterogeneous real-time systems

Abstract

We present a new scheduler architecture, which permits adding QoS policies to the scheduling decisions. We also present a new scheduling synthesis method which allows a designer to obtain a safe scheduler for a particular application and at the same time helps him in analysing the task interactions and the overall system behaviour. Our scheduler architecture and scheduler synthesis method have not been developed for a particular application model and, therefore, can be used for heterogeneous applications, where there are periodic tasks, event-driven ones and tasks which are always enabled and where the tasks communicate through various synchronisation primitives. Finally, we present a prototype implementation of this scheduler architecture and related mechanisms on top of an open-source OS for embedded systems.

Modular Schedulability Analysis of Concurrent Objects in Creol

“…The works of [2,10] is similar to ours as they extract automata from code for schedulability analysis. However, first of all they deal with programming languages and timings are usually obtained by profiling the real system.…”

A model-based approach for multiple QoS in scheduling: from models to implementation

“…Here, we show how one can use it to perform a fine-grain analysis of systems and implement these in a way guaranteeing safety properties (i.e., deadlocks, deadlines) and that can furthermore be easily extended to support other quality aspects of the system (e.g., jitter, memory, energy). We show how one can achieve these goals through a new methodology, which increases the applicability and benefits of scheduler synthesis [29]. Our system analysis and scheduler synthesis methods do not make any assumptions on the tasks comprising the system, their periodicity (or lack thereof), their synchronisation patterns, etc.…”

Thunderstriking constraints with JUPITER

“…The rest of the paper is structured as follows: we first present the design of the controller synthesiser of JUPITER and the improvements we introduced from earlier versions of it [7,6]. Then we present the refinement methodology we are using and, following that, we show how JUPITER helps in the analysis of the synthesised controllers.…”

“…• Refinement Methodology [7,6] : The refinement methodology we propose is based on the usual wish to consider easier properties first, gradually enriching the system model so as to prove more difficult ones. In this way, we increase our chances of being able to attack realistically sized problems, at the expense of slightly over-constraining the system, a small price to pay usually.…”

“…Second, how to successively refine the models used for the controller synthesis so as to obtain a series of more easily understandable and more robust controllers. Last but not least, how to structure & explain the synthesised controllers and provide hints to designers for further optimisations through the use of Machine Learning techniques.Thus, JUPITER can help in the design and analysis of multi-constraint systems through the automatic synthesis of control logic for certain of the constraints and the aid it provides to designers for discovering further optimisations.The controllers it synthesises can be easily implemented on top of a standard real-time OS [7,6]. …”