J. Stankovic, M. Spuri, K. Ramamritham, and G. Buttazzo. DEADLINE SCHEDULING FOR REAL-TIME SYSTEMS -- EDF and Related Algorithms. Kluwer Academic Publishers, Boston, Massachusetts, USA, 1998.  Home/Search   Document Not in Database   Summary   Related Articles   Check

.... fresh inputs. It is not at all obvious that the software architecture meets its deadline requirements. In this section, we perform a formal analysis to check whether this is the case. We use results from real time scheduling theory, in particular, xed priority scheduling theory (e.g. see [8, 7, 5, 14, 1, 2, 15]) and the so called HKL model and analysis [3, 4] First, we cast the platoon application into the formal model. Then, we estimate the execution times and other latencies involved in the system, and compute the total CPU utilization. This is found to be about 74 , that is, less than 1, which is a ....

J. Stankovic, M. Spuri, K. Ramamritham, , and G. Buttazzo. Deadline Scheduling For Real-Time Systems: EDF and Related Algorithms. Kluwer Academic Publishers, 1998.

....more general purpose than the synchronous languages mentioned above. Implementation of asynchronous languages typically relies on an operating system. The latter is responsible for scheduling, which is usually based on static priorities. Real time scheduling theory (e.g. 11] 12] 18] and [32]) provides techniques, such as rate monotonic analysis [26] that guarantee satisfaction of simple time constraints, such as deadlines. Unfortunately, these results are often applicable only to simple models, and are difficult to generalize. Neither of the two paradigms faces the implementation ....

J. Stankovic, M. Spuri, K. Ramamritham, and G. Buttazzo, Deadline Scheduling for Real-Time Systems: EDF and Related Algorithms. Norwell, MA: Kluwer, 1998.

....deadlines (MEDF) It will be shown that applying MEDF to a network of communicating processes will lead called critical instant 1.3. THIS WORK S BACKGROUND: RAPID PROTOTYPING 5 to an EDF processing sequence of these processes. Thus, feasibility analysis algorithms like [Gre93b, Gre93a] SSRB98] or [Jef92] may be applied to prove the required timeliness. Unfortunately, there exists no known MEDF implementation in state of the art real time operating systems. Therefore, necessary run time system support has to be provided. Schedulability analysis requires all best (BCET) and ....

....too. Again, there exist innumerable research papers addressing the relaxation of the task model strictness assumed by Liu and Layland. EDF analysis algorithms to schedule task with precedences, aperiodic task activations and protocols to synchronize tasks with shared resources are summarized in [SSRB98] Feasibility conditions for sporadic task sets with common resources are presented for example in [Jef92] and [CLB99] Event driven systems that are triggered by stimuli with sporadic or even burst behaviour are investigated in [Gre93a, Gre93b] Gresser s EDF scheduling theorems and his Event ....

[Article contains additional citation context not shown here]

John A. Stankovic, Marco Spuri, Krithi Ramamritham, and Giorgio C. Buttazzo. Deadline Scheduling for Real--Time Systems --- EDF and Related Algorithms. Kluwer Academic Publishers, Dordrecht, The Netherlands, 1998. 5, 13, 55, 56

....modifying the plans when requests cannot be satisfied. Our approach is specially suited to unpredictable environments, where resources have to be allocated in a dynamic way that cannot be predicted in advance. We rely on the wide body of algorithms that exists in the area of real time scheduling [9] and load balancing [2] Our architecture provides support for distribution of resources across robots, use of shared resources, and seamless integration of autonomous and human supervised control [10] 3 Distributed Robotic Control The Scout robot is a cylindrical, two wheeled robot that is 40 ....

J. Stankovic, M. Spuri, K. Ramamritham, and G. Buttazzo. Deadline Scheduling For Real-Time Systems: EDF and Related Algorithms. Kluwer Academic Publishers, Boston, 1998.

....the release of the job, which re ects the importance of the job. To ease our discussion, let us de ne the value density of a job to be its value divided by the processing time. The importance ratio of a system is the ratio of the largest possible value density to the smallest possible one. See [18] for more discussion on deadline scheduling. We analyze the performance of online algorithms with respect to their competitiveness (see, e.g. 5] and [17] for general background) In this paper, we say that an online algorithm A is c competitive if for any job sequence, A guarantees to obtain at ....

John Stankovic, Marco Spuri, Krithi Ramamritham, and Giorgio Buttazzo. Deadline scheduling for real-time systems: EDF and related algorithms. Kluwer Academic Publishers, Boston, Mass., 1998. 13

....Tak Wah Lam yz Tsuen Wan Johnny Ngan xz Kar Keung To December 2, 2002 1 Introduction This paper is concerned with online algorithms for scheduling jobs with deadlines. A typical example is the earliest deadline rst (edf) algorithm, which has been widely used in real time systems (see [15] for a survey) It is well known that edf is optimal for a single processor system that is underloaded, i.e. whenever there exists an o line schedule meeting the deadlines of all jobs released, edf can always do so [7] However, when the system is overloaded or involves more than one processor, ....

John A. Stankovic, Marco Spuri, Krithi Ramamritham, and Giorgio C. Buttazzo. Deadline scheduling for real-time systems: EDF and related algorithms. Kluwer Academic Publishers, Boston, Mass., 1998. 15

....multiple unit speed processors cannot. 1 Introduction Online algorithms for scheduling jobs with deadlines on a processor have been studied extensively in the literature. A typical example is the earliest deadline rst (EDF) algorithm, which has been widely used in many real time systems (see [19] for a survey) For scheduling underloaded systems, EDF is optimal, i.e. whenever there exists a schedule meeting the deadlines of all jobs released, EDF can always do so [7] However, when the system is possibly overloaded, no algorithm has a worst case performance guarantee in the sense that ....

....any processor. We consider both settings where job migration is allowed at no cost and disallowed, respectively. In general, a system may be overloaded in the sense that there is no schedule meeting the deadlines of all jobs released. For more details of rm deadline scheduling, one can refer to [19]. The value of a job re ects its importance and is not necessarily related to the processing time. The value density of a job is de ned to be its value divided by its processing time, and the importance ratio k of a system is the ratio of the largest possible value density to the smallest ....

J. Stankovic, M. Spuri, K. Ramamritham, and G. Buttazzo. Deadline scheduling for real-time systems: EDF and related algorithms. Kluwer Academic Publishers, Boston, Massachusetts, 1998.

....and deadline of a job are known only when the job is released. Preemption is allowed at no cost. In general, a system may be overloaded in the sense that there may be no schedule meeting the deadline of every job released. A widely used algorithm is the earliest deadline first (EDF) algorithm (see [4, 13] for more background on on line scheduling) In this paper, we focus on schedulers that maximize processor utilization on jobs that meet the deadlines. In other words, jobs that complete after their deadlines give no credits. From a theoretical viewpoint, EDF is optimal for underloaded systems, ....

John Stankovic, Marco Spuri, Krithi Ramamritham, and Giorgio Buttazzo. Deadline scheduling for real-time systems: EDF and related algorithms. Kluwer Academic Publishers, Boston, Mass., 1998.

....priority among all active threads. If a thread with a higher priority becomes active, the currently executed thread is preempted. The Earliest Deadline First (EDF) scheme executes the thread closest to its deadline. Therefore, the only necessary parameter for this scheme is the deadline. 11] and [15] show, that for periodic threads on a single processor system this is an optimal scheme. It guarantees all deadlines up to 100 processor utilization. The Least Laxity First (LLF) scheme can be considered as an extension to the Earliest Deadline First scheme. Additionally to the deadline, the ....

J. A. Stankovic, M. Spuri, K. Ramamritham, and G. Buttazzo. Deadline Scheduling for Real-Time Systems: EDF and Related Algorithms. Kluwer Academic Publishers, Dordrecht Norwell, 1998.

....caches and branch prediction are more suitable than highly speculative complex microprocessors. Second, the operating system used should be a real time one. It should support at least one of the common real time scheduling policies like xed priority preemptive or earliest deadline rst (EDF) 1] [2]. The process service described in section 2.1 adapts to di erent process models and scheduling policies. Real time memory management is a useful feature which can be used by OSA as will be seen later in this section. In case of no real time memory management is available the operating system ....

J. A. Stankovic, M. Spuri, K. Ramamritham, G. C. Buttazzo. Deadline Scheduling for Real-Time Systems: EDF and Related Algorithms. Kluwer Academic Publishers, 1998.

....transportation systems and medical electronics. To obtain high performance for real time heterogeneous systems, scheduling algorithms play an important role. While a scheduling algorithm maps real time tasks to processors in the system such that deadlines and response time requirements are met [29], the system must also guarantee its functional and timing correctness even in the presence of faults. The proposed algorithm, referred to as eFRCD (efficient Fault tolerant Reliability Cost Driven Algorithm) endeavors to comprehensively address the issues of fault tolerance, reliability, ....

....as the reliability cost, defined to be the product of processor failure rate and task execution time. In addition, the time for detecting and handling of a permanent fault is incorporated into the scheduling scheme, thus making the algorithm more practical. While the various algorithms studied in [1 11,13 29] share one or two features with eFRCD, in terms of the assumed operational conditions, as explained in Section 2, the latter is arguably the most comprehensive, in terms of the number of different scheduling issues addressed, and outperforms several quantitatively comparable algorithms in the ....

J. Stankovic, M. Spuri, K. Ramamritham and G.C. Buttazzo, "Deadline Scheduling for Real-time systems: EDF and Related Algorithms," Kluwer Academic Publishers, 1998

....control. Limited communication bandwidth is a serious problem when robots have to transmit large amounts of data, such as live video, and when many robots need to share the bandwidth, as in the examples we presented. A wide body of literature exists in the area of real time scheduling algorithms [13]. We plan on experimenting with various scheduling and negotiation algorithms for resource allocation in the context of our architecture. For instance, a promising method for negotiation for real time systems has been proposed [1] in the context of automated ight control. The method is based on ....

J. Stankovic, M. Spuri, K. Ramamritham, and G. Buttazzo. Deadline Scheduling For Real-Time Systems: EDF and Related Algorithms. Kluwer Academic Publishers, Boston, 1998.

.... the worst case response time remains very interesting in the dynamic case; it is useful not only for the analysis of single processor systems, but also for the analysis of distributed real time systems, e.g. in the holistic schedulability analysis by Tindell and Clark [19] Spuri and colleagues [17, 18] have shown how to compute for arbitrary deadline systems the worst case response times with the deadline driven scheduler and outlined the fact that this worst case does not always occur in the initial period of the synchronous schedule. In this paper we shall also exhibit two interesting ....

....(i.e. 0; L 0 ) is often called the first busy period in the literature; we shall call the interval [0; L) the first elementary busy period in order to distinguish them. It may be noticed that an idle point is equivalent to an idle slot of zero length some authors used (see for instance [17, 18]) to extend (without a proof) Liu and Layland s results. 4 Worst case response time Recently, Spuri [17, 18] has extended the computation of the worst case response time, already considered for static priority rules, to the dynamic case (more exactly to the deadline driven scheduler) It may be ....

[Article contains additional citation context not shown here]

STANKOVIC, J. A., SPURI, M., RAMAMRITHAM, K., AND BUTTAZZO, G. C. Deadline Scheduling for Real-Time Systems EDF and Related Algorithms. Kluwer Academic Publishers, 1998.

....priority versions of the deferrable server, priority exchange server and sporadic server. At the same time, Spuri et al. in [22] also introduced five different servers under dynamic priorities. The issue of using dynamic servers and its variants has received continuous attention, for example see [23, 6]. The work of Ghazalie et al. in [9] also considers exact schedulability tests for fixed priority servers in order to compare them with their dynamic versions. Also in that paper there is the first reference to the problem of double blocking which is reviewed and discussed in detail in section ....

J. A. Stankovic, M. Spuri, K. Ramamritham, and G. Buttazzo. Deadline Scheduling for real-time systems: EDF and Related Algorithms. Kluwer academic publishers, 1998.

....that the specification has to satisfy for this technique to be useful. Key words: timed automata, trajectories, schedulers, synthesis 1 Introduction There is no doubt that scheduling plays a central role in the development of real time systems. There is a large body of knowledge in this area [13,6]. The principal focus of this work has been to study the behaviour of general algorithms (such as earliest deadline first, rate monotonic scheduling etc. This then identifies a class of real time applications for which a particular algorithm is ideal (e.g. results in optimal behaviour) A ....

J. Stankovic, M. Spuri, K. Ramamritham, and G. Buttazzo. Deadline Scheduling for Real-Time Systems: EDF and Related Algorithms.Kluwer Academic Publishers, 1998.

....this paper is an improved analysis of the competitiveness for weighted deadline scheduling. 1 Introduction This paper is concerned with online algorithms for deadline scheduling. A typical example is the earliest deadline rst (EDF) algorithm, which is widely used in many real time systems (see [13] for a survey) Yet, from a theoretical viewpoint, EDF except in some simple settings has no performance guarantee, i.e. its performance cannot match or even be competitive against the o line adversary. It is indeed known that in many settings of deadline scheduling, no online algorithm has this ....

J. A. Stankovic, M. Spuri, K. Ramamritham, and G. C. Buttazzo. Deadline scheduling for real-time systems: EDF and related algorithms. Kluwer Academic Publishers, 1998. 10

....threads is possible. ISRs can only be interrupted by other ISRs, but not by threads. Threads allow flexible priority schemes. For example, earliest deadline first or guaranteed percentage scheduling can be used instead of fixed priority preemptive. Both, the earliest deadline first (EDF) scheme [2] and the guaranteed percentage scheme allow processor charges of 100 . The guaranteed percentage scheme offers the possibility to assign a rate of the full processing power to each thread of a system of threads. Herewith response times and data rates can be guaranteed even for several ....

J. A. Stankovic, M. Spuri, K. Ramamritham, G. C. Buttazzo. Deadline Scheduling for Real-Time Systems: EDF and Related Algorithms. Kluwer Academic Publishers, 1998.

....The generation delay includes the application processing time needed to generate the contents of the message and the time taken to queue the message. This issue has been extensively addressed in the literature related with tasks worst case response time analysis in single processor systems ([6,7] are just two examples) The queuing delay is a consequence not only from the contention between message requests from the same master but also with message requests from other masters. The impact of the first factor in the overall queuing delay depends on the policy used to queue message ....

Stankovic, J., Spuri, M., Ramamritham, K. and Buttazzo, G.: "Deadline Scheduling for Real-Time Systems -- EDF and Related Algorithms", Kluwer Academic Publishers, Boston, 1998. - 15 -

....threads and other threads is possible. ISRs can only be interrupted by other ISRs, but not by threads. ffl Threads allow flexible priority schemes. For example, earliest deadline first (EDF) or guaranteed percentage scheduling can be used instead of fixed priority preemptive. Both, the EDF scheme [2] and the guaranteed percentage scheme allow processor charges of 100 . ffl The guaranteed percentage scheme offers the possibility to assign a rate of the full processing power to each thread. Herewith response times and data rates can be guaranteed even for several concurrent events independent ....

J. A. Stankovic, M. Spuri, K. Ramamritham, G. C. Buttazzo. Deadline Scheduling for Real-Time Systems: EDF and Related Algorithms. Kluwer Academic Publishers, 1998.

....sets consisting of skippable periodic and soft tasks. They present a scheduling algorithm based on EDF which exploits skips to minimise the average response time of the soft tasks. It is known that under overload conditions, deadline based scheduling systems perform badly due to the domino e ect [14]. This has lead to the de nition of scheduling strategies which rely on the value of the task rather than the deadline under overload conditions. For example, Buttazzo et al. in [15] studies the value based scheduling under overloads. Other works are just based on mechanisms that integrate hard ....

John A. Stankovic, Marco Spuri, Krithi Ramamritham, and Giorgio Buttazzo, Deadline Scheduling for real-time systems: EDF and Related Algorithms, Kluwer academic publishers, 1998.

No context found.

J. A. Stankovic, M. Spuri, K. Ramamritham, and G. C. Buttazzo, Deadline Scheduling for Real-Time Systems -- EDF and Related Algorithms, Kluwer Academic Publishers, 1998.

No context found.

J. A. Stankovic, M. Spuri, K. Ramamritham, and G. C. Buttazzo, Deadline Scheduling for Real-Time Systems - EDF and Related Algorithms, Kluwer Academic Publishers, 1998.

No context found.

J. Stankovic, M. Spuri, K. Ramamritham, and G. Buttazzo. DEADLINE SCHEDULING FOR REAL-TIME SYSTEMS -- EDF and Related Algorithms. Kluwer Academic Publishers, Boston, Massachusetts, USA, 1998.

No context found.

J. Stankovic, M. Spuri, K. Ramamritham, and G. Buttazzo, Deadline Scheduling For Real-Time Systems: EDF and Related Algorithms. Norwell, MA: Kluwer, 1998.

No context found.

J. Stankovic, M. Spuri, K. Ramamritham, and G. Buttazzo, Deadline Scheduling For Real-Time Systems: EDF and Related Algorithms. Norwell, MA: Kluwer, 1998.

No context found.

John A. Stankovic, Marco Spuri, Krithi Ramamritham, and Giorgio C. Buttazzo. Deadline Scheduling for Real-Time Systems: EDF and Related Algorithms, chapter 5, Planning-based scheduling. Kluwer Academic Publishers, October 1998.

No context found.

J. A. Stankovic et al., Deadline Scheduling for RealTime Systems -- EDF and Related Algorithms, Kulwer Academic Publications, New York, 1998.

No context found.

John A. Stankovic, Marco Spuri, Krithi Ramamritham, and Giorgio C. Buttazzo. Deadline scheduling for real-time systems: EDF and related algorithms. Kluwer Academic Publishers, Boston, Mass., 1998. 15

No context found.

J.A. Stankovic, M. Spuri, K. Ramamritham, and G.C. Buttazzo. Deadline Scheduling for Real-Time Systems: EDF and Related Algorithms, volume 460 of Kluwer International Series in Engineering and Computer Science. Kluwer Academic Publishers, 1998.

No context found.

J.A. Stankovic, M. Spuri, K. Ramamritham, and G.C. Buttazo. Deadline Scheduling for Real-Time Systems: EDF and Related Algorithms. Kluwer Academic Publisher, 1998.

No context found.

J. A. Stankovic, M. Spuri, K. Ramamritham, and G. C. Buttazzo. Deadline scheduling for real-time systems: EDF and related algorithms. Kluwer Academic Publishers, 1998. 9

No context found.

J.A. Stankovic, M. Spuri, K. Ramamritham, and G.C. Buttazo. Deadline Scheduling for Real-Time Systems: EDF and Related Algorithms. Kluwer Academic Publisher, 1998.

No context found.

J. Stankovic, M. Spuri, K. Ramamritham, and G. Buttazzo. DEADLINE SCHEDULING FOR REAL-TIME SYSTEMS -- EDF and Related Algorithms. Kluwer Academic Publishers, Boston, Massachusetts, USA, 1998.

No context found.

J. A. Stankovic, M. Spuri, K. Ramamritham, and G. C. Buttazzo. Deadline scheduling for real-time systems: EDF and related algorithms. Kluwer Academic Publishers, 1998. 9

No context found.

J.A. Stankovic, M. Spuri, K. Ramamritham, and G.C. Buttazo. Deadline Scheduling for Real-Time Systems: EDF and Related Algorithms. Kluwer Academic Publisher, 1998.

No context found.

J.A. Stankovic, M. Spuri, K. Ramamritham, and G.C. Buttazo. Deadline Scheduling for Real-Time Systems: EDF and Related Algorithms. Kluwer Academic Publisher, 1998.

No context found.

J. Stankovic, M. Spuri, K. Ramamritham, and G. Buttazzo. Deadline Scheduling for Real-Time Systems: EDF and Related Algorithms. Kluwer Academic Publishers, 1998.

No context found.

J. Stankovic, M. Spuri, K. Ramamritham, and G. Buttazzo. Deadline scheduling for real-time systems: EDF and related algorithms. Kluwer Academic Publishers, Boston, Mass., 1998. 12

No context found.

J. A. Stankovic, M. Spuri, K. Ramamritham, G. Buttazzo, Deadline Scheduling for Real-Time Systems: EDF and Related Algorithms, Kluwer Academic Publishers, Dordrecht Norwell, 1998.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC