Static Approximation Algorithms for Regularity-based Resource Partitioning

As a hierarchical real-time system framework, the Regularity-based Resource Partition Model allocates physical resources in time intervals determined by integral numbers of a time unit to tasks in different applications. A Regularity-based Resource Partition is characterized by its regularity and availability factor. An important problem is how to schedule a group of resource partitions with the specification of their regularities and availability factors. Mok and Feng have provided the AAF-Single algorithm which works only on a single-resource platform. Li and Cheng have recently extended AAF-Single to AAF-Multi for multiresource platforms without violating the schedulability bound given by Feng. However, the resource utilization of these two algorithms could be significantly improved. This paper is dedicated to developing optimized resource partitioning algorithms for the Regularity-base Resource Partition Model. We first decompose the resource partitioning problem into two sub problems, and invoke the Pfair algorithm to solve the first sub problem. Then we introduce a category of approximation algorithms called Static Approximation Algorithms (SAA) to solve the second sub problem. An SAA adjusts the availability factors of the resource partitions with a specific boundary sequence. We prove that the schedulability bound of any feasible SAA is at most 0.5. Furthermore, we develop an optimal SAA called Magic7. Simulation shows that Magic7-enhanced algorithms improve the resource utilization by 10% or more.

[1]  Aloysius K. Mok,et al.  Design of real-time virtual resource architecture for large-scale embedded systems , 2004 .

[2]  Sanjoy K. Baruah,et al.  Fast scheduling of periodic tasks on multiple resources , 1995, Proceedings of 9th International Parallel Processing Symposium.

[3]  Aloysius K. Mok,et al.  Resource partition for real-time systems , 2001, Proceedings Seventh IEEE Real-Time Technology and Applications Symposium.

[4]  Insup Lee,et al.  Periodic resource model for compositional real-time guarantees , 2003, RTSS 2003. 24th IEEE Real-Time Systems Symposium, 2003.

[5]  Insup Lee,et al.  Compositional Analysis Framework Using EDP Resource Models , 2007, RTSS 2007.

[6]  Aloysius K. Mok,et al.  Towards compositionality in real-time resource partitioning based on regularity bounds , 2001, Proceedings 22nd IEEE Real-Time Systems Symposium (RTSS 2001) (Cat. No.01PR1420).

[7]  Insup Lee,et al.  Hierarchical Scheduling Framework for Virtual Clustering of Multiprocessors , 2008, 2008 Euromicro Conference on Real-Time Systems.

[8]  Björn Andersson,et al.  Sporadic Multiprocessor Scheduling with Few Preemptions , 2008, 2008 Euromicro Conference on Real-Time Systems.

[9]  Chung Laung Liu,et al.  Scheduling Algorithms for Multiprogramming in a Hard-Real-Time Environment , 1989, JACM.

[10]  Sanjoy K. Baruah,et al.  Proportionate progress: A notion of fairness in resource allocation , 1993, Algorithmica.

[11]  Sanjoy K. Baruah,et al.  Proportionate progress: a notion of fairness in resource allocation , 1993, STOC '93.

[12]  Shinpei Kato,et al.  Semi-partitioned Fixed-Priority Scheduling on Multiprocessors , 2009, 2009 15th IEEE Real-Time and Embedded Technology and Applications Symposium.

[13]  S. Shirero,et al.  On the schedulability conditions on partial time slots , 1999, Proceedings Sixth International Conference on Real-Time Computing Systems and Applications. RTCSA'99 (Cat. No.PR00306).

[14]  Albert Mo Kim Cheng,et al.  Regularity-Based Partitioning of Uniform Resources in Real-Time Systems , 2012, 2012 IEEE International Conference on Embedded and Real-Time Computing Systems and Applications.