Lehoczky, J. P., and Sha, L. "Performance of Real-Time Bus Scheduling Algorithms". ACM Performance Evaluation Review, Special Issue 14, 1 (May, 1986).  Home/Search   Document Not in Database   Summary   Related Articles   Check

.... of each other, it has been shown that the rate monotonic algorithm is an optimal static priority scheduling scheme [ 11 ] By optilnal we mean that, if a process set can be scheduled by any fixed priority algorithm, then it can also be scheduled by the rate monotonic scheduling algorithm [ 12]. The rate monotonic algorithm requires a preemptire scheduler; all processes am allocated a priority according to tleir period. The shorter the period, the higher their priority. For this simple scheme, the priorities remain fixed and therefore implementation is straightforward. Overheads are ....

SHA, L., and LEHOCZKY, J.P. :'Performance of real-time bus scheduling algorithms', ACM Perform. Eval. Rev., 1986, 14, (1)

....yet are rarely considered in embedded system scheduling or codesign. Arbitrated busses, in particular, are used in PCI [3] and Small PCI Compact PCI [4] embedded systems. Communication scheduling is sometimes considered as a topic completely independent from other system considerations (e.g. [5], 6] 7] While it may make sense in some circumstances to treat communication separately (e.g. when a good deal of control over communication scheduling is possible or when communication time is insigni cant when compared to computation times) assuming that communication doesn t a ect the ....

John P. Lehoczky and Lui Sha, \Performance of real-time bus scheduling algorithms," ACM Performance Evaluation Review, Special Issue, vol. 14, no. 1, pp. 44-53, May 1986.

....the flexibility can be viewed as increasing an application s likelihood of obtaining good performance with a fixed set of resources. Predictability for applications has been extensively studied in the real time systems literature. Several well known scheduling schemes, such as rate monotonic [19, 25] and 20 CHANG, KARAMCHETI, AND KEDEM 0 0.2 0.4 0.6 0.8 10 30 50 70 throughput laxity arrival interval 0 500 1000 1500 2000 2500 3000 3500 4000 0 0.2 0.4 0.6 0.8 10 30 50 70 z 0 500 1000 1500 2000 2500 3000 3500 throughput laxity arrival interval Benefit ....

L. Sha and J. Lehoczky. Performance of real-time bus scheduling algorithms. ACM Perform. Eval. Rev., 14(1), 1986.

....the flexibility can be viewed as increasing an application s likelihood of obtaining good performance with a fixed set of resources. Predictability for applications has been extensively studied in the real time systems literature. Several well known scheduling schemes, such as rate monotonic [19, 25] and earliest deadline first (EDF) exist for scheduling real time tasks in an uniprocessor environment. Gillies [13] has studied the scheduling of tasks in AND#OR graphs. However, the validity of most of these results is restricted to 1441 APPLICATION TUNABILITY either uniprocessor systems or a ....

L. Sha and J. Lehoczky, Performance of real-time bus scheduling algorithms, ACM Perform. Eval. Rev. 14, 1 (1986).

....qualities: the issue then is of maximizing the achieved job quality. As with most non trivial scheduling problems both the above formulations are NP hard. Furthermore, unlike the uniprocessor case, where there are well established algorithms for real time scheduling (such as rate monotonic (RM) [12, 16], earliest deadline first (EDF) and least slack time first (LST) that are all optimal in the sense that if any algorithm can schedule a given task set with no missed deadlines, then all of them can as well) no such algorithms are known for the parallel case. Consequently, in the next section, we ....

....scenarios; the flexibility can be viewed as increasing an application s likelihood of obtaining good performance with a fixed set of resources. Predictability for applications has been well studied in the real time systems literature. Several wellknown scheduling schemes, such as rate monotonic [12, 16] and earliest deadline first (EDF) exist for scheduling real time tasks in an uniprocessor environment. Gillies [8] has studied the scheduling of tasks in AND OR graphs, a problem very similar to our problem of scheduling task chains in tunable applications. However, the validity of most of these ....

L. Sha and J. Lehoczky. Performance of real-time bus scheduling algorithms. ACM Perform. Eval. Rev., 1986, 14(1).

....closed loop control across the network and high level system diagnostics. Among the more recent suggestions is TTP [1] where a TDMA scheme quarantees a fixed bandwith to each node, a tight clock synchronization between nodes as well as direct information about liveliness and internal state. In [3] and [5] approximate network preemptivness as well as prioritization among nodes is assumed in order to facilitate the translation of results from scheduling theory ( 2] and [ to real time network communication. In this paper an existing network technology is analyzed w.r.t. real time ....

J. P. Lehoczky and L. Sha, Performance of Real-Time Bus Scheduling Algorithms," ACM Performance Evaluation Review, Special Issue 14(1), 1986.

....the deadline monotonic algorithm, which uses fixed priorities and the priority of a task is assigned to be the inverse of its deadline. Liu and Layland s result [Liu and Layland 1973] of the worst case utilization bound for the rate monotonic algorithm has been generalized by Lehoczky and Sha [Lehoczky and Sha 1986]and Lehoczky [Lehoczky 1990] for the case that deadlines of tasks occur earlier or at the end of the periods. It is common in practice that task sets can be feasibly scheduled with the RMS algorithm even if they have utilizations greater than the worst case bound. This has been described by ....

J. P. Lehoczky and L. Sha. Performance of Real-Time Bus Scheduling Algorithms. ACM Performance Evaluation Review, 14, 1986.

....scenarios; the flexibility can be viewed as increasing an application s likelihood of obtaining good performance with a fixed set of resources. Predictability for applications has been well studied in the real time systems literature. Several well known scheduling schemes, such as rate monotonic [12, 17] and earliest deadline first (EDF) exist for scheduling real time tasks in an uniprocessor environment. Gillies [8] has studied the scheduling of tasks in AND OR graphs. However, the validity of most of these results is restricted to either uniprocessor systems or a sequential task model. Our ....

L. Sha and J. Lehoczky. Performance of real-time bus scheduling algorithms. ACM Perform. Eval. Rev., 14(1), 1986.

....new request period begins, the heuristic may reevaluate its notion of the highest priority request, possibly preempting an existing task. This real time scheduling problem and heuristic are widely studied. Results have been derived by Liu Layland [6] Lehoczky [8] Lehoczky, Sha, and Strosnider [10] [9] Shih, Liu, and Liu [15] Tindell [17] and many others. In particular, Liu Layland showed that the rate monotonic priority assignment was optimal among all priority assignments, and that all deadlines would be met if the task system utilization P (C i =T i ) ln 2. In this paper we ....

J. Lehoczky and L. Sha, Performance of real-time bus scheduling algorithms, ACM Performance Evaluation and Sigmetrics, May 1986, pp. 35-43.

....(or advanced) by some fixed factor Delta 0; the deadline of a task is Delta times its period. The results described there for 1 2 Delta 0 are applicable to our model in the special case that all tasks share the same skip factor s ( Delta = 1=s) The utilization bound quoted there (due to [11, 16]) is of 1=s for all s 2. One can see that even in this special case while the task set of example 5.1 fails the test of [11, 16] because its utilization is greater than 1=2) it is still proven schedulable using the bound derived above. 6 Conclusion Like telemarketing solicitations and ....

....1 2 Delta 0 are applicable to our model in the special case that all tasks share the same skip factor s ( Delta = 1=s) The utilization bound quoted there (due to [11, 16] is of 1=s for all s 2. One can see that even in this special case while the task set of example 5. 1 fails the test of [11, 16] (because its utilization is greater than 1=2) it is still proven schedulable using the bound derived above. 6 Conclusion Like telemarketing solicitations and political speeches, repeated signals arriving to a controller contain a great deal of redundancy. It is therefore sometimes acceptable to ....

J. P. Lehoczky and L. Sha. Performance of realtime bus scheduling algorithms. ACM Performance Evaluation Review, 14, 1986.

....qualities: the issue then is of maximizing the achieved job quality. As with most non trivial scheduling problems both the above formulations are NP hard. Furthermore, unlike the uniprocessor case, where there are well established algorithms for real time scheduling (such as rate monotonic (RM) [12, 16], earliest deadline first (EDF) and least slack time first (LST) that are all optimal in the sense that if any algorithm can schedule a given task set with no missed deadlines, then all of them can as well) no such algorithms are known for the parallel case. Consequently, in the next section, we ....

....scenarios; the flexibility can be viewed as increasing an application s likelihood of obtaining good performance with a fixed set of resources. Predictability for applications has been well studied in the real time systems literature. Several wellknown scheduling schemes, such as rate monotonic [12, 16] and earliest deadline first (EDF) exist for scheduling real time tasks in an uniprocessor environment. Gillies [8] has studied the scheduling of tasks in AND OR graphs, a problem very similar to our problem of scheduling task chains in tunable applications. However, the validity of most of these ....

L. Sha and J. Lehoczky. Performance of real-time bus scheduling algorithms. ACM Perform. Eval. Rev., 1986, 14(1).

.... projects, among them we can cite the BSY 1 Submarine Combat System trainer developed by IBM, and the space station data management system of NASA [4] It has also been used to analyze systems developed with the help of some open standards like the IEEE FutureBus or the POSIX real time extensions [9, 13]. 3 Contribution and novelties 3.1 The Project This electricity meter monitors the electric installation of a customer for billing purposes, but also in order to allow for some actions to be performed on the installation itself (e.g. circuit switching, tariff programming) The functions of the ....

J.P. Lehoczky and L. Sha. Performance of real-time bus scheduling algorithms. ACM Performance Evaluation Review, Special Issue, 14(1), May 1986.

.... projects, among them we can cite the BSY 1 Submarine Combat System trainer developed by IBM, and the space station data management system of NASA [12] It has also been used to analyze systems developed with the help of some open standards like the IEEE FutureBus or the POSIX real time extensions [19, 25]. It was also the basis of a previous and much simpler study at Schlumberger Industries. 3.2.1 Rate Monotonic Scheduling The RM scheduling algorithm has been defined by Liu and Layland [21] on periodic and fully preemptable tasks, having no more interaction than the share of a common processor. ....

J.P. Lehoczky and L. Sha. Performance of real-time bus scheduling algorithms. ACM Performance Evaluation Review, Special Issue, 14(1), May 1986.

....[Leh90] He considered task systems where all the deadlines and periods were related by a constant, 8i d i DeltaT i and Delta 0. Using the same fixed priority scheme as RMS, Lehoczky derived worst case utilization bounds for such systems. In particular, Lehoczky, Sha, Peng and Shin [ Leh90] LS86] and [PS89] in [vTK91] generalized the RMS schedulability test so that a periodic task system with 8i d i DeltaT i is schedulable if X 8i U i 8 : n( 2 Delta) 1 n Gamma 1) 1 Gamma Delta) 1 2 Delta 1 Delta; 0 Delta 1 2 2.3.2 RMS Extensions Burns and Welling ....

J. P. Lehoczky and L. Sha. Performance of real-time bus scheduling algorithms. ACM Performance Evaluation Review, 14, 1986.

.... algorithm s schedulability performance advantage is more than offset by the additional overhead incurred, and have pointed to the instability of the earliest deadline algorithm under transient overload conditions [3] A summary of the various extensions to rate monotonic scheduling may be found in [4, 5, 6]. More information on dynamic priority scheduling may be found in [7, 8, 9] Both Liu and Layland s, and Lehoczky, Sha and Ding s work assumed zero overhead and perfect preemption. Katcher, Arakawa and Strosnider [10] proposed operating system scheduling models which bridge the gap between ....

J. Lehoczky and L. Sha, "Performance of real-time bus scheduling algorithms," ACM Performance Evaluation Review, Special Issue, vol. 14, May 1986.

....for the uniprocessor task scheduling. This is due to a few reasons as the agreement problem, the limited number of priority bits that are made available by the network protocols and the limited preemption allowed by message packetization. A clear description of these phenomena can be found in [4]. In this paper, we will present a dynamic technique to schedule real time messages in a LAN. The technique is a simple variation of the earliest deadline scheduling policy. We will first show how to check the schedulability of a set of packetized messages, under the assumption of a perfect ....

Lehoczky J.P. and Sha L., "Performance of realTime Bus Scheduling Algorithms," ACM Performance Evaluation Review, 1986.

....a sequence of processors) is used before and after a certain service is attained. An example is a database that is queried before and updated after a specific operation. The periodic flow shop model is a generalization of both the traditional flow shop model and the traditional periodic job model [11, 13]. As in the traditional periodicjob model, the periodic job system J to be scheduled in a flow shop consists of n independent periodic jobs; each job consists of a periodic sequence of requests for the same computation. In our previous terms, each request is a task. The period p i of a job J i in ....

....P m j=1 ffi j 1. We can postpone the phase of each subjob J ij on P j by ffi j p i units. This generates a feasible schedule where all precedence constraints and all deadlines are met. Given the parameters of J , we can compute the set fu j g and use the existing schedulability bounds given in [11, 12, 13] to determine whether there is a set of fffi j g where ffi i 0 and P m j=1 ffi j 1. The job system J can be feasibly scheduled in the manner described above if such a set of ffi j exists. With a small modification of the required values of ffi j , this method can handle the case where the ....

J. P. Lehoczky and L. Sha. Performance of real-time bus scheduling algorithms. ACM Performance Evaluation Review, 1986.

....priority levels can be denoted as B l , and is addressed below. A smaller number of priority levels than required by the scheduling algorithm causes a potential loss in schedulability. The impact of limited priority levels on the utilization based schedulability test [LL73] has been discussed by [LS86]. The impact of limited priorities on the time based schedulability test of [LSD89] is discussed in [SKS92] This is a technique to quantify the schedulability loss due to limited priorities for a particular connection set on a network with a known number of priority levels. The technique finds ....

J. Lehoczky and L. Sha. Performance of real-time bus scheduling algorithms. ACM Performance Evaluation Review, Special Issue, 14(1), May 1986.

....otherwise the system is said to have failed. Aperiodic tasks have been included within the periodic scheduling framework through the use of aperiodic servers in [5, 6] Other modifications to the original periodic scheduling framework include adding synchronization [7] and scheduling for buses [8]. A good overview of real time systems is found in [9] A real time task set is composed of n tasks. Each task i ; 1 i n, is described by a period, T i , a deadline, D i , and a worst case estimated execution time, C i . We assume that all deadlines are before or at task periods. Each ....

J. Lehoczky and L. Sha, "Performance of realtime bus scheduling algorithms," ACM Performance Evaluation Review, Special Issue, vol. 14, May 1986.

No context found.

Lehoczky, J. P., and Sha, L. "Performance of Real-Time Bus Scheduling Algorithms". ACM Performance Evaluation Review, Special Issue 14, 1 (May, 1986).

....real time system, we must be able to manage both the logical complexity and timing complexity by using a disciplined approach. The logical complexity is addressed by software engineering methodology, while the timing complexity is addressed by research in real time scheduling algorithms [2, 3, 5, 6, 7, 8, 10, 12, 14]. An important class of scheduling algorithms is known as static priority scheduling algorithms. These algorithms have several attractive properties. First, they are simple to implement. Second, they have good performance. The utilization bound of a randomly chosen periodic task set is 88 [4] ....

....algorithms translate complex timing constraints into simple resource utilization constraints. As long as the utilization constraints of the CPU, I O channels and communication media are observed, the deadlines of periodic tasks and the response time requirements of aperiodic tasks will both be met [2]. This means that the realtime software can be modified freely as long as the utilization bounds are observed. Furthermore, should there be a transient overload, the tasks that will miss deadlines will miss them in reverse order of importance, and the number of tasks missing their deadlines will ....

Lehoczky, J. P., and Sha, L. Performance of Real-Time Bus Scheduling Algorithms. ACM Performance Evaluation Review, Special Issue vol. 14, no. 1 , May, 1986.

....The utility of the scheduling models is demonstrated by analyzing several common system buses. Keywords: real time scheduling, bus architecture, bus scheduling, scheduling framework. 1.0 Introduction Previous work on real time bus scheduling focused on fixed priority scheduling. Specifically [LEH86] introduced bus scheduling based on rate monotonic analysis and analyzed the effects of task preemption, insufficient priority levels, and buffer management for an ideal bus. In addition [SHA86] included scheduling analysis between the DMA, bus, and the CPU. SHA91] describes the operation of ....

J. Lehoczky and L. Sha, "Performance of Real-time Bus Scheduling Algorithms," Performance 1986 and ACM Sigmetrics 1986, May 1986.

No context found.

J. P. Lehoczky and L. Sha, "Performance of real-time bus scheduling algorithms," ACM Performance Evaluation Review, vol. 14, 1986.

No context found.

J. P. Lehoczky and L. Sha. Performance of real-time bus scheduling algorithms. ACM Performance Evaluation Review, 14:44--53, 1986.

No context found.

J. P. Lehoczky and L. Sha. "Performance of Real-Time Bus Scheduling Algorithms". ACM Performance Evaluation Review, Special Issue 14(1): 44-53, May, 1986.

First 50 documents

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