J. A. Stankovic, M. Spuri, M. D. Natale, and G. C. Buttazzo, "Implications of classical scheduling results for real-time systems," Computer, vol. 28, no. 6, pp. 16--25, June 1995.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....under the mentioned assumptions are given by Chetto et al. CSB90] while Spuri and Stankovic [SS94] consider also shared resources. Most problems related to hard real time scheduling on multi processor systems under non trivial assumptions have been proven to be NPcomplete [GJ75, Ull75, GJ79, SSDB94] The work of Sun and Liu [SL95] addresses multi processor real time systems where the application exhibits a particular form of precedence relationships among tasks, namely the periodic job shop model. Sun and Liu have provided analytic bounds for the task response time under such assumptions. ....

J. A. Stankovic, M. Spuri, M. Di Natale, and G. C. Butazzo. Implications of classical scheduling results for real-time systems. Technical Report UM-CS-94-089, Computer Science Department, University of Massachusetts, 1994.

....in the previous discussion. The literature review section presents a brief overview of existing scheduling algorithms that can be applied to this problem. Section 3.2 describes how my background and previous research relates to this research. 3. 1 Literature Review Priority scheduling algorithms [9, 16] are all designed based on a simple premise. At any given time, the highest priority runnable tasks are being executed. This assumes that tasks can be preempted if a higher priority task is ready to be executed. Priority scheduling algorithms differ in how priorities are allocated to tasks. ....

J. A. Stankovic, M. Sprui, M. DiNatale, and G. C. Buttazzo. Implications of classical scheduling results for real-time systems. IEEE Computer, 28:16--25, June 1995.

....systems have evolved spectacularly in the last years. These systems are available from personal computers to large scale computational machines. In particular, the use of these powerful computing resources in real time systems has opened several problems concerning scheduling strategies [1,2]. The problem of determining when and where a given task must execute without missing its deadline or compromising other task deadlines in multiprocessor systems often becomes intractable. Besides, when the scheduling is possible, the algorithms that are optimal for uniprocessor systems are not ....

Stankovic, J.A., Spuri, M., Di Natale, M., Butazzo, G.C., "Implications of Classical Scheduling Results for RealTime Systems", IEEE Computer, v. 28, n.6, pp.15-25, 1995

....our approach gives better results than local Slack Stealing schedulers. 1. Introduction Multiprocessor systems have evolved rapidly in the last years. At the same time, the use of these powerful computing resources in real time systems has opened several problems concerning scheduling strategies [1,2]. The problem of determining when and where a given task must execute without missing its deadline or compromising other task deadlines in multiprocessor systems often becomes intractable. Besides, when the scheduling is possible, algorithms that are optimal for uniprocessor systems are not ....

....added uncertainty factor: the results depend on the decisions previously done about which task to execute where and these decisions make some tasks to interfere with others or not. For example, it is well known the multiprocessor anomaly of reducing some execution times can increase the run length [2]. As a consequence, an algorithm to compute the worst case response time for periodic tasks in a multiprocessor would consider the worst situations, giving very pessimistic values. We have done some attempt to find the upper bounds to periodic task response times and the results obtained were far ....

[Article contains additional citation context not shown here]

Stankovic, J.A., Spuri, M., Di Natale, M., Butazzo, G.C., "Implications of Classical Scheduling Results for RealTime Systems", IEEE Computer, v. 28, n.6, pp.15-25, 1995

..... We also assume that the voltage be continuously within its operating voltage range. 3 Static Slowdown Factors We compute static slowdown factor for a system with an underlying Earliest Deadline First scheduler. The problem of scheduling tasks in the presence of resource sharing is NP hard [8]. Resource access protocols have been designed to bound the blocking times and sufficient schedulability tests have been given in the presence of maximum blocking times. The Stack Resource Protocol [2] has been designed to handle tasks scheduled under EDF policy. Our work is based on the use of ....

J. A. Stankovic, M. Spuri, M. D. Natale, and G. Buttazzo. Implications of classical scheduling results for real-time systems. In IEEE Transactions on Computers, 1994.

....in real life applications, tasks share the resources in the system. This could lead to tasks being blocked for a particular resource. Blocking of tasks can cause priority inversion [21] and result in deadline misses. The problem of scheduling tasks in the presenc of resource sharing is NP hard [15, 5, 24]. Resource access protocols have been designed to bound the blocking times and sufficient schedulability tests have been given in the presence of maximum blocking times. Resource access protocols such as priority inheritance protocol, priority ceiling protocol and priority limit protocol [21] deal ....

J. A. Stankovic, M. Spuri, M. D. Natale, and G. Buttazzo. Implications of classical scheduling results for real-time systems. In IEEE Transactions on Computers, 1994.

.... example, some policy may attempt to minimize number of tasks which miss their deadlines, another could try to minimize lateness (time by which a deadline is missed, more precisely, completion time deadline) Among the policies that we have tried out is the earliest deadline rst policy (EDF) [5, 6], which has the desirable property of being optimal. A policy is said to be optimal if it is able to nd a schedule meeting all deadlines if one exists. Another policy called least laxity(deadline completion time) rst (LLF) 5, 6] provides a di erent type of guarantee it minimizes maximum ....

....we have tried out is the earliest deadline rst policy (EDF) 5, 6] which has the desirable property of being optimal. A policy is said to be optimal if it is able to nd a schedule meeting all deadlines if one exists. Another policy called least laxity(deadline completion time) rst (LLF) [5, 6], provides a di erent type of guarantee it minimizes maximum lateness. However it su ers from the drawback of requiring an estimate of completion time of all threads, and poor estimates may lead to skewed scheduling. Compared to LLF, EDF is preferred because it does not need completion time ....

J.A. Stankovic et al. Implications of Classical Scheduling Results for Real Time Systems.

....of system overload. Best effort schedulers will generally allocate a proportional share of the CPU to each process, Rate Monotonic will meet the deadlines of processes with highest priority (generally those with the lowest period) 9] and EDF will miss all deadlines by roughly the same amount [33]. While the overload of either RT scheduler might be considered optimal in some strictly SRT environments, they suffer from the fact that they will starve best effort processes entirely. Figure 3.6 shows the performance of the Linux, BEST, and RM schedulers with three processes, one best effort, ....

John A. Stankovic, Marco Spuri, Marco Di Natale, and Giorgio Buttazo. Implications of classical scheduling results for real-time systems. Computer, 28(6):16--25, June 1995.

....of a set of tasks in multiprocessor system is an NP complete problem. Clearly then, a heuristic scheduling algorithm has to be designed in order to reduce the scheduling costs. Much of the work reported in the area of scheduling problem in distributed systems is on task allocation problems[1,2,3,5]. In a traditional distributed system, the objectives of a task allocation algorithm is to balance the load among processors and to minimize the communication cost for tasks. Task allocation is a dicult problem even without timing constraints. For example, nding optimal assignment of tasks with ....

....communication graph to four or more processors with di erent speeds is known to be NP hard. Considerable e orts have been spent on more restrictive allocation problems or on developing heuristic algorithms to nd suboptimal solutions using network ow and dynamic programming. For example, Bokhari[1] developed a dynamic programming algorithm to allocate tasks which form a tree to a system with an arbitrary number of processors and has complexity of O(n m ) Ni[9] has identi ed two objectives of any distributed scheduling algorithm: maximize the utilization of resource minimize ....

Stankovic et. al."Implications of Classical Scheduling Results for Real-Time System" , IEEE Computer 1995

....implies that if tasks have deadlines, then the accompanying increase in the schedule length due to the anomaly can result in missing of task deadlines for which guarantees were given before. It is a counter intuitive result. This is because of the nature of timing constraints and multiprocessing [108]. Suppose a task finishes before its finish time, a simple solution to avoid the run time anomaly is to keep the processor idle till the worst case finish time of the task. In such case, the processors are said to be non work conserving. A processor is said to be work conserving if it never be ....

J.A. Stankovic, M. Spuri, M.D. Natale, and G.C. Buttazzo, "Implications of classical scheduling results for real-time systems," IEEE Computer, pp.16-25, June 1995.

.... delay computation, logic optimization, and combinational equivalence checking can be formulated as SAT problems [Mar00] An example of a heuristics for solving SAT problem is presented in [Bay96] Summery of important classical scheduling theory results for real time computing can be found in [Sta95] and [Pau89] An example of an ILP based formulation for Scheduling is presented in [Geb91] 3. PRELIMINARIES 3.1. SAT Let U= u 1 , u 2 , u m be a set of Boolean variables. A clause over U is a set of literals that represents their disjunction. A clause is satisfied by a truth assignment ....

J. Stankovic, et al. "Implications Of Classical Scheduling Results For Real-Time Systems." IEEE Computer, Vol.28, pp. 16-25, 1995.

....that is efficient, accurate, modular and retargetable. We present preliminary results for sample embedded programs to demonstrate the applicability of the proposed approach. 1. INTRODUCTION Static timing analysis is essential in real time systems development, where the schedulability analysis [1] of programs with hard real time constraints depends on the estimated extreme case performance. It is also useful in verification of timing critical systems, hardware software co design of embedded systems, and early design space exploration. Because of its importance, many researchers have ....

J. Stankovic, M. Spuri, M. Di Natale, and G. Buttazzo. "Implications of classical scheduling results for real-time scheduling". IEEE Computer, pages 16--25, June 1995.

....data and another video data. The results that consider multiprocessor systems fit a more flexible design in which the arriving tasks are stored in a global buffer of the shared memory (See e.g. 4, 1] Previous work: So far the main research efforts considered preemptive scheduling (See e.g. [10]) Dertouzos [2] showed that the earliest deadline first algorithm is optimal for arbitrary task sets. Only few works considered the case of non preemptive scheduling. Leung [8] considered the non preemptive scheduling prob lem for multiprocessor computing system, where the tasks are of uniform ....

J. Stankovic, M. Spuri, M. Di Natale, and G. Buttazzo, "Implications of Classical Scheduling Results For RealTime Systems," IEEE Computer, Vol. 28, No. 6, pp.1625, June 1995.

....section 6 we evaluate our approach experimentally. The last section draws some conclusions. 2 Related work Much work has been done in the field of task scheduling with fixed parameters (periods, execution times, etc. Results are summarized in several surveys such as those by Stankovic et al. [10], Fidge [5] and Audsley et al. 3] However, only recently researchers have focused on scheduling policies, schedulability analysis and performance analysis of tasks with stochastic parameters. Atlas and Bestavros [2] extend the classical rate monotonic scheduling policy with an admittance ....

J. A. Stankovic, M. Spuri, M. Di Natale, G. Butazzo, "Implications of Classical Scheduling Results for RealTime Systems", IEEE Computer, June 1995, pp. 16--25

.... transactions [Yu94] If the two phase locking protocol (2PL) Bern87] is used for RTDBS, priority inversion may occurs [Rajk91] which is highly undesirable to real time scheduling as it can greatly degrade the schedulability of the adopted CPU scheduling algorithm and affect the system performance [Liu73, Stan95]. In recent years, different real time concurrency control protocols have been proposed [Agra95, Best95, Hari92, Huan92, Lam95, Lee94, Ulus94] Most of them are based on 2PL owing to its popularity and simplicity in implementation [Eswa76] The correctness notion adopted in these protocols is ....

J. Stankovic, Spuri, M. and Natale, M.D., "Implications of Classical Scheduling Results for Real-time Systems", IEEE Computer, volume 28, number 6, pages 16-25, 1995.

....timing constraints. Processes are independent if their correct execution does not imply need for synchronization. Note that processes can be independent even when they interchange data between them or have a set of imposed control and timing constraints [Sha and Goodenough 1990] Stankovic et al. [Stankovic et al. 1995] provided an excellent survey of a number of real time A Methodology and Algorithms for the Design of Hard Real Time Multi Tasking ASICs 9 of 30 Preliminaries scheduling algorithms, their computational complexity, assumed timing and hardware models. We review a few results of direct importance to ....

....to this problem, if preemption is allowed, is to schedule the processes in non decreasing due date order. Unfortunately, neither Jackson s rule nor any other known algorithm can optimally solve the same problem if preemption is not allowed, since the associated optimization problem is NP complete [Stankovic et al. 1995]] The rate monotonic scheduling address the problem of scheduling a set of periodic processes. It belongs to the class of static priority driven preemptive approaches where no explicit scheduling is provided, but instead always the process with the current highest priority among the available ....

Stankovic, J.A., Spuri, M., Natale, M.D., and Buttazzo, G.C. 1995. "Implications of Classical Scheduling Results for Real-Time Systems, IEEE Computer, 28, 6, 16-25.

....the tasks admitted to the system and for which it can be proved that each task will complete before its deadline. Scheduling has been studied for long time and many proposals exist for static scheduling and (much fewer) dynamic approaches. Without entering in details on the individual algorithms [Stankovic et al. 1995] it is remarkable to note that guarantees (absolute or probabilistic) are given by making many, often unrealistic, assumptions on the number of tasks, their maximum execution time, precedence constraints and resource conflicts, etc. Moreover very restrictive assumptions are often made on the ....

J.A. Stankovic, M. Spuri, M. Di Natale and G.C. Buttazzo, "Implications of Classical Scheduling Results for Real-Time Systems," IEEE Computer, vol. 28, no. 6, pp.16-25, 1995. page 26

....MURST. # active replication, where each task is always executed on two or more processors (active copies) if any task fails its active copies will continue to execute. Many scheduling problems have been found to be NPcomplete: most likely, there are no optimal polynomialtime algorithms for them [15]. In particular, scheduling periodic tasks with arbitrary deadlines is NP complete, even if only a single processor is available [8, 3] Several heuristic algorithms for scheduling periodic tasks on uniprocessor and multiprocessor systems have been proposed. Liu and Layland [10] proposed the ....

J. A. Stankovic, M. Spuri, M. D. Natale, and G. Buttazzo. Implications of classical scheduling results for real-time systems. IEEE Computer, 28(6):16--25, June 1995.

....transactions, and concurrency control protocols, such as two phase locking [7] and optimistic concurrency control (OCC) 7] were used to guarantee database consistency. In particular, the well known priority ceiling protocol (PCP) which is for the scheduling of tasks in hard realtime systems [6,8], has been extended for the scheduling of hard real time transactions [9] Sha, Rajkumar, and Lehoczky [10] proposed the PCP in which processes can inherit the higher priority of a process they block. The priority ceiling of a 2 resource is the priority of the highest priority process which may ....

Stankovic, J., Spuri, M. and Natale, M.D. (1995.) Implications of Classical Scheduling Results for Real-time Systems. IEEE Computer, 28(6),16-25.

.... For example, some policy may attempt to minimize number of tasks which miss their deadlines, another could try to minimize lateness (time by which a deadline is missed, more precisely, completion time deadline) Among the policies that we have tried out is the earliest deadline rst policy(EDF) [5, 6], which has the desirable property of being optimal. A policy is said to be optimal if it is able to nd a schedule meeting all deadlines if one exists. Another policy called least laxity(deadline completion time) rst (LLF) 5, 6] provides a di erent type of guarantee it minimizes maximum ....

....that we have tried out is the earliest deadline rst policy(EDF) 5, 6] which has the desirable property of being optimal. A policy is said to be optimal if it is able to nd a schedule meeting all deadlines if one exists. Another policy called least laxity(deadline completion time) rst (LLF) [5, 6], provides a di erent type of guarantee it minimizes maximum lateness. However it su ers from the drawback of requiring an estimate of completion time of all threads, and poor estimates may lead to skewed scheduling. Compared to LLF, EDF is preferred because it does not need completion time ....

J.A. Stankovic et al. Implications of Classical Scheduling Results for Real Time Systems. IEEE Computer, Vol. 28, No. 6, pp. 16-25, 1995.

....considers some of the practical issues a system must deal with in order to dispatch a set of tasks using priority based scheduling. Not only is priority based scheduling simple to implement, but it is exible enough to support a variety of static and dynamic scheduling algorithms (for details see [11, 24, 36]) With static scheduling, decisions are based on the entire task set, while with dynamic scheduling, decisions are based only on the current task set. In most cases, the di erences between static and dynamic algorithms lie in the ability to handle aperiodic tasks. Many real time tasks tend to ....

....schedule a priori for static scheduling to calculating appropriate task scheduling parameters for dynamic scheduling. The on line component ranges from dispatching the tasks according to a xed schedule with static scheduling to determining if a new task can be scheduled with dynamic scheduling [36]. Scheduling decisions are based on a variety of constraints and criteria, e.g. 1) period: inter arrival time between successive occurrences of the same task, 2) computation: worstcase execution time for an instance of the task, 3) deadline: time by which an instance of the task must be ....

[Article contains additional citation context not shown here]

J. A. Stankovic, M. Spuri, M. Di Natale, and G. C. Buttazzo. Implications of classical scheduling results for real-time systems. Computer, 28(6):16-25, June 1995.

.... begun investigating the impact of overload with respect to CC, and an initial set 2 0 0 1 4 10 3 8 TASK T1 TASK T2 time Figure 1: Effective Processor Utilization (EPU) EPU metric has been widely used in the analysis of real time scheduling algorithms under conditions of overload (e.g. [15, 4, 3, 13, 17]) A detailed discussion of the applicability of this metric to real time systems is provided in [15] In particular, our goal is to compare the EPU performance of on line scheduling algorithms against that of an optimal off line (or clairvoyant) algorithm. On line schedulers make scheduling ....

J. Stankovic, M. Spuri, M Di Natale, and G. Buttazzo. Implications of classical scheduling results for real-time systems. In IEEE Computer, June 1995. 15

....For statically evaluated least slack, slack is measured once when a transaction arrives and is defined by: slack = deadline (arrival time service time) where service time is the total service time required at all devices. During overload conditions, EDF su#ers from the Domino E#ect [64, 32], whereby a response which is already late is given the highest priority, thus delaying responses which could otherwise more easily meet their deadlines. To stabilize overload performance, Haritsa, Livny and Carey [32] propose algorithms called Adaptive Earliest Deadline (AED) and Hierarchical ....

....works well at lower loads, allowing most of the responses to complete within their deadlines. At heavy loads some other scheduling policies perform better but we are most interested in cases where most responses are completed on time. During periods of overload, EDF su#ers from the domino e#ect [64], whereby a response which is already late is given the highest priority, thus delaying responses which could otherwise more easily meet their deadlines. To stabilize overload performance, Haritsa, Livny and Carey [32] propose an algorithm called Adaptive Earliest Deadline. This research has ....

J. Stankovic, M. Spuri, M. Di Natale, and G. Buttazzo. Implications of classical scheduling results for real-time systems. IEEE Computer, 28(6):16--25, June 1995. BIBLIOGRAPHY 141

....is allowed to execute at some instant of time so that the maximum number of tasks meet their deadlines. Unfortunately, optimal scheduling decisions easily become prohibitively expensive at run time or even computationally intractable, specifically for multiprocessors and distributed systems [15]. In these cases, heuristics may serve as viable alternatives, providing good enough behaviour at acceptable run time overhead. While certain properties of sophisticated heuristics can be derived analytically, it is often desirable to verify these results or even to nd new approaches ....

J. A. Stankovic, M. Spuri, M. Di Natale, and G. Buttazzo. Implications of classical scheduling results for real-time systems. IEEE Computer, 28(6):16-25, June 1995.

....is allowed to execute at some instant of time so that the maximum number of tasks meet their deadlines. Unfortunately, optimal scheduling decisions easily become prohibitively expensive at run time or even computationally intractable, speci cally for multiprocessors and distributed systems [25]. Some systems, however, cannot provide su cient information a priori to allow for e ective clairvoyant scheduling [4, 5] In these cases, heuristics may provide viable alternatives, providing good enough behaviour at acceptable run time overhead. While certain properties of sophisticated ....

J. A. Stankovic, M. Spuri, M. Di Natale, and G. Buttazzo. Implications of classical scheduling results for real-time systems. IEEE Computer, 28(6):1625, June 1995.

....physical or computational resources 6 . It is beyond the scope of this dissertation to completely describe that scheduling work, but here we provide a brief look at some work in the area. A good survey of classical (mostly operations research) scheduling results can be found in Stankovic et al. [Stankovic et al. 1995] . There has been a large amount of research in the area of scheduling real time tasks with hard deadlines. In hard real time scheduling the emphasis is on finding solutions that never miss a deadline, because missed deadlines can lead to complete system failure. For example, nuclear power plant ....

....collisions) We would like to extend our scheduling algorithm to schedule effectively given resource constraints. This would bring our scheduler into the realm of AI based job shop scheduling [Smith et al. 1990, Hildum, 1994] as well as systems oriented scheduling approaches [Graham et al. 1979, Stankovic et al. 1995] Full embedding in an application. As described in Chapter 4 we have began exploring the design to time approach in real systems, but we would like to verify the usefulness of the approach by completely embedding it in an actual application. Possible issues to be addressed would include ....

John A. Stankovic, Marco Spuri, Marco Di Natale, and Giorgio C. Buttazzo. Implications of classical scheduling results for real-time systems. IEEE Computer, 28(6):16--25, June 1995.

No context found.

J. A. Stankovic, M. Spuri, M. D. Natale, and G. C. Buttazzo, "Implications of classical scheduling results for real-time systems," Computer, vol. 28, no. 6, pp. 16--25, June 1995.

No context found.

Stankovic, J., M. Spuri, M. Di Natale and G. Buttazzo (1995). Implications of classical scheduling results for real-time systems. IEEE Computer, 28(6), 16 -- 25.

No context found.

J. Stankovic, M. Spuri, M. D. Natale, and G.Buttazzo. Implications of classical scheduling results for real-time systems. IEEE Computer, pages 16--25, Jun. 1995.

No context found.

J.A. Stankovic et. al. Implications of classical scheduling results for real-time systems. IEEE Computer, Volume 28, Number 6, pp. 16-25, June 1995.

No context found.

J. A. Stankovic, M. Spuri, M. Di Natale, and G. Buttazzo, "Implications of Classical Scheduling Results for Real-time Systems", IEEE Computer, June, 1995

No context found.

J. Stankovic, M. Spuri, M. Di Natale, and G. Buttazzo, "Implications of Classical Scheduling Results for Real-Time Systems," Computer, June 1995.

No context found.

J.A.Stankovic, M.Spuri, M.Di Natale, G.Buttazzo, \Implications of Classical Scheduling Results for Real-Time Systems," IEEE Computer, Vol. 28, No. 6, June 1995.

No context found.

J. A. Stankovic, M. Spuri, M. D. Natale, and G. Buttazzo. Implications of classical scheduling results for real-time systems. IEEE Computer, 28(6):16--25, June 1995.

No context found.

Stankovic, J. A., Spuri, M., Di Natale, M., and Buttazzo, G. C. Implications of classical scheduling results for real-time systems. Computer 28, 6 (June 1995), 16-25.

No context found.

Stankovic, J., "Implications of Classical Scheduling Results for Real-Time Systems", IEEE Computer, 1995.

No context found.

J.A. Stankovic et al., Implications of Classical Scheduling Results for Real-Time Systems, IEEE Computer, 1995

No context found.

J.A. Stankovic, M. Spuri, M. Di Natale and G.C. Buttazzo, "Implications of Classical Scheduling Results for Real-Time Systems," IEEE Computer, Vol. 28, pp. 16-25, 1995.

No context found.

J. A. Stankovic, M. Spuri, M. Di Natale, and G. Buttazzo, "Implications of Classical Scheduling Results for Real-time Systems", IEEE Computer, June, 1995

No context found.

J. A. Stankovic, M. Sprui, M. DiNatale, and G. C. Buttazzo. Implications of classical scheduling results for real-time systems. IEEE Computer, v. 28, n.6, pp15-25, 1995.

No context found.

J.A. Stankovic et al., "Implications of Classical Scheduling Results for Real-Time Systems," Computer, June 1995, pp. 16-25.

No context found.

John A. Stankovic, Marco Spuri, Macro Di Natale, and Giorgio Buttazzo. Implications of classical scheduling results for real-time syatems. IEEE computer, 28(6):16--25, 1995.

No context found.

J. A. Stankovic, M. Spuri, M. Di Natale, G. Butazzo, "Implications of Classical Scheduling Results for Real-Time Systems", IEEE Computer, June 1995

No context found.

J. A. Stankovic, M. Spuri, M. D. Natale, and G. Buttazo, \Implications of classical scheduling results for real-time systems," Computer 28, pp. 16-25, June 1995.

No context found.

Stankovic, J., "Implications of Classical Scheduling Results for Real-Time Systems", IEEE Computer, 1995.

No context found.

J. Stankovic, "Implications of Classical Scheduling Results for Real-time Systems", IEEE Computer, Vol.28, No.6, June 1995.

No context found.

J. Stankovic, M. Spuri, M. Di Natale, and G. Buttazzo, Implications of Classical Scheduling Results For Real-Time Systems, IEEE Computer, Vol. 28, No. 6, pp. 16-25, June 1995.

No context found.

John A. Stankovic, Marco Spuri, Marco Di Natale, and Giorgio C. Buttazzo, Implications of Classical Scheduling Results for Real-Time Systems, Computer, 16-25, 1995. 97

No context found.

J. A. Stankovic, M. Spuri, M. D. Natale and G. C. Buttazzo, "Implications of Classical Scheduling Results for Real-Time Systems," IEEE Computer, June 1995.

First 50 documents  Next 50

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