7th International Real-Time Scheduling Open Problems Seminar (RTSOPS 2016)

The 7^th International Real-Time Scheduling Open Problems Seminar (RTSOPS 2016) provides a venue for the exchange of ideas and the discussion of interesting unsolved problems in real-time scheduling. The format of the seminar encourages interaction between participants and provides ample time for relaxed discussions. The goal of the seminar is to promote a spirit of co-operation and collaboration within the real-time scheduling community.

RTSOPS 2016 is organized around presentation and collaboration sessions. Each presentation session provides the opportunity to hear about a number of unsolved problems in real-time scheduling, highlighted via brief presentations. The following collaboration session gives participants the opportunity to interact in small groups, exchange ideas with the presenters, explore how the problems might be solved, and take the first steps towards a solution. In addition, authors of open problem submissions from previous years' RTSOPS are invited to submit short status reports on progress made towards a solution to their open problem. This year's seminar will include a session for presentation of such advances.

The call for papers is available as both PDF and plain text.

Scope

Real-time scheduling theory has provided a foundation for understanding and solving resource allocation and scheduling problems in systems that have real-time constraints. New fundamental results are needed to address recent advances and trends in real-time systems design. RTSOPS 2016 encompasses all aspects that are relevant in the study of real-time scheduling.

RTSOPS 2016 invites extended abstracts of open problems in areas such as, but not limited to:

• Single-, Multi- and Many-core scheduling;
• New models for real-time systems;
• Scheduling in cyber-physical systems;
• Mixed-criticality scheduling;
• Interactions between WCET (worst-case execution time) analysis and scheduling.

Instructions for authors

Extended abstracts (for either new open problems or status reports on previously-presented problems) for RTSOPS 2016 must be written in English, and not exceed two A4 pages in length in single column, 10pt format, including relevant references.

• Download the MS Word template here
• Download the L^aT_eX template here

The main purpose of the abstract for new open problem submissions should be to provide a clear description of the problem. Abstracts may be submitted describing well-known but as yet unsolved problems. However, all abstracts should contain some element of original work that has not been published before; for example, a new problem, a new way of looking at an existing problem, new intuition or ideas on how a problem might be solved, possible frameworks for solutions, or overviews of special cases that may be useful in solving a problem. Abstracts describing progress on an open problem presented at previous RTSOPS meetings should include a brief statement of the open problem and a description of the technical advances made towards a solution.

Submissions

The submission page is now closed

Authors of accepted abstracts are expected to give a brief presentation of their open problem, and be prepared to work on the problem with other participants during the collaboration sessions. Instructions for preparing final copy and presentations will be given when the abstract is accepted. RTSOPS 2016 will publish proceedings as a seminar booklet.

By submitting an extended abstract, the authors agree and confirm that: neither this extended abstract nor a version close to it is under submission or will be submitted elsewhere before notification by RTSOPS 2016, and if accepted, at least one author will register for the RTSOPS 2016 seminar by the special registration deadline set in the notification of acceptance, and present the open problem at the seminar in person. Please note that extended abstracts that do not describe real-time scheduling problems will not be accepted. Submissions will be refereed for quality and relevance. Submissions exceeding the page limit may be rejected without review.

As in the past few years, the authors of the paper that will present the most interesting open problem will be rewarded with an IPad!

Important dates

• Submission deadline: 26th May, 2016
• Acceptance notifications: 7th June, 2016
• Final versions: 17th June, 2016
• Seminar date: 5th July, 2016

Program

8:30	Registrations open
Session 1 (9:30 – 11:00)
9:30	Keynote: Luca Santinelli, ONERA, France Title of the talk: "Probabilistic real-time: Achievements and Open Problems" Probabilistic real-time means doing real-time modeling and real-time schedulability analysis with the use of probabilities at representing tasks behavior. With respect to the deterministic real-time i.e. the real-time without probabilities, the probabilistic real-time is less mature but full of potential benefits. Multiple are the open problems with respect to the actual achievements e.g., how to guarantee timing constraints with probabilities (how to make probabilities trustable), how to make probabilistic models reliable and robust, how to develop effective and less pessimistic probabilistic real-time schedulability analyses. The key note intends to explore both the achievements and the open problems related to probabilistic real-time, clarifying some notions and misconceptions. Download the slides
10:15	Johnson’s Procedure: Mechanization and Parallelization. Mamoun Filali-Amine and Nabil Zaidi Download the slides
10:30	Collaboration Session 1
11:00	Coffee break
Session 2 (11:30 - 13:00)
11:30	Efficient Execution of Dependent Tasks on Many-Core Processors. Hamza Rihani, Claire Maiza and Matthieu Moy Download the slides
11:45	Tardiness Bounds for Global EDF Scheduling on a Uniform Multiprocessor. Kecheng Yang and James H. Anderson Download the slides
12:00	The Feasibility Analysis of Mixed-Criticality Systems. Saravanan Ramanathan, Xiaozhe Gu and Arvind Easwaran Download the slides
12:15	How Does Task Scheduling Affect Engine Control Performance? Paolo Pazzaglia, Alessandro Biondi, Marco Di Natale and Giorgio Buttazzo Best paper award Download the slides
12:30	Collaboration Session 2
13.00	Lunch
Session 3 (14:30 - 16:00)
14:30	Keynote: Victor Jegu, Digital Cores Basic Software Expert, Airbus, France Title of the talk: "Airbus predictable scheduling strategy for Timing Critical applications" This key note addresses the problem of scheduling tasks in systems with hard timing requirements and strong industrial and avionic certification constraints. Different problems have to be addressed. The WCET problem: providing a predictable scheduling so that the WCET method can soundly assess the processor and caches state. The functional verification: insuring no race conditions or data inconsistencies can arise and lead to unwanted behaviors. The timing verification: insuring latency and jitters requirements are met. And finally facilitating the incremental development: minimizing the verification cost of functional evolutions. About the speaker: Victor Jégu joined Airbus in 2004. He worked before on the development and certification of multiple Avionic projects, with special interest for processors, languages and low level programming. From 2004 to 2008, he worked on the A380 and A350 programs, static WCET methodology for complex processors, and USB models for bus accesses interferences. Since 2007, he participates to research projects, Airbus only and in cooperation with academic and industrial partners, relating to multicore processors and SCADE code generation. He has one patent with ONERA on "deterministic multicore processor execution”.
15:15	Collaboration Session 3
15:45	Vote for the "Best Problem Award"
16:00	Coffee Break
Session 4 (4:30 - 5:15)
16:30	RTSOPS impact: A presentation and discussion on past RTSOPS papers and their impact on real-time scheduling research Download the report
17:15	End of the Seminar

Instructions for participants

Every speaker is requested to prepare a 15 minutes presentation to highlight the open problems that he would like the audience to think about. Each session at the workshop will be followed by a 30 minutes collaboration time during which the presenters will have ample opportunities to interact with the workshop audience and to address some of the open problems that he raised.

Proceedings

Download the proceedings