Results 1 
9 of
9
Improved conditions for bounded tardiness under EPDF fair multiprocessor scheduling
 In Proceedings of the 12th International Workshop on Parallel and Distributed RealTime Systems
, 2004
"... 1 This author was also supported by an IBM Ph.D. fellowship. ..."
Abstract

Cited by 11 (8 self)
 Add to MetaCart
1 This author was also supported by an IBM Ph.D. fellowship.
The partitioned multiprocessor scheduling of deadlineconstrained sporadic task systems
 IEEE Transactions on Computers
, 2006
"... Abstract—A polynomialtime algorithm is presented for partitioning a collection of sporadic tasks, each constrained to have its relativedeadline parameter be no larger than its period parameter, among the processors of an identical multiprocessor platform. Since the partitioning problem is easily s ..."
Abstract

Cited by 11 (2 self)
 Add to MetaCart
Abstract—A polynomialtime algorithm is presented for partitioning a collection of sporadic tasks, each constrained to have its relativedeadline parameter be no larger than its period parameter, among the processors of an identical multiprocessor platform. Since the partitioning problem is easily seen to be NPhard in the strong sense, this algorithm is unlikely to be optimal. A quantitative characterization of its worstcase performance is provided in terms of resource augmentation: It is shown that any set of sporadic tasks that can be partitioned among the processors of an mprocessor identical multiprocessor platform will be partitioned by this algorithm on an mprocessor platform in which each processor is ð3 1=mÞ times as fast. Index Terms—Sporadic tasks, partitioned scheduling, multiprocessors, resource augmentation. 1
The Partitioned Scheduling of Sporadic RealTime Tasks on Multiprocessor Platforms
"... In the sporadic task model, a task is characterized by three parameters — an execution requirement, a relative deadline, and a period parameter — and has the interpretation that it generates an infinite sequence of jobs, such that (i) the arrivaltimes of any two successive jobs are separated by a t ..."
Abstract

Cited by 6 (3 self)
 Add to MetaCart
In the sporadic task model, a task is characterized by three parameters — an execution requirement, a relative deadline, and a period parameter — and has the interpretation that it generates an infinite sequence of jobs, such that (i) the arrivaltimes of any two successive jobs are separated by a timeinterval at least as long as the period parameter; (ii) each job has a deadline that is separated from its arrivaltime by a timeinterval exactly equal to the relative deadline parameter of the task; and (iii) each job must execute for an amount equal to its execution requirement by its deadline. Most previous research concerning the scheduling of collections of sporadic tasks upon multiprocessor platforms has added the additional constraint that all tasks have their relative deadline parameters equal to their period parameters. In this research, we consider the scheduling of systems of sporadic tasks that do not necessarily satisfy this additional constraint, upon preemptive multiprocessor platforms. We propose, and evaluate, an algorithm for partitioning a given collection of arbitrary sporadic tasks upon a specified number of preemptive processors such that all deadlines are guaranteed to always be met.
Schedulable utilization bounds for EPDF fair multiprocessor scheduling (full paper). Available at http://www.cs.unc.edu/˜anderson/papers.html
"... Abstract. The earliestpseudodeadlinefirst (EPDF) algorithm is less expensive than other known Pfair algorithms, but is not optimal for scheduling recurrent realtime tasks on more than two processors. Prior work established sufficient pertask weight (i.e., utilization) restrictions that ensure t ..."
Abstract

Cited by 3 (3 self)
 Add to MetaCart
Abstract. The earliestpseudodeadlinefirst (EPDF) algorithm is less expensive than other known Pfair algorithms, but is not optimal for scheduling recurrent realtime tasks on more than two processors. Prior work established sufficient pertask weight (i.e., utilization) restrictions that ensure that tasks either do not miss their deadlines or have bounded tardiness when scheduled under EPDF. Implicit in these restrictions is the assumption that total system utilization may equal the total available processing capacity (i.e., the total number of processors). This paper considers an orthogonal issue — that of determining a sufficient restriction on the total utilization of a task set for it to be schedulable under EPDF, assuming that there are no pertask weight restrictions. We prove that a task set with total utilization at most 3M+1 is correctly scheduled under EPDF on M processors, 4 regardless of how large each task’s weight is. At present, we do not know whether this bound is tight. However, we provide a conterexample that shows that it cannot be improved to exceed 86 % of the total processing capacity. Our schedulability test is expressed in terms of the maximum weight of any task, and hence, if this value is known, may be used to schedule task sets with total utilization greater than 3M+1
A schedulable utilization bound for the multiprocessor EPDF Pfair algorithm
 RealTime Systems
, 2008
"... The earliestpseudodeadlinefirst (EPDF) Pfair scheduling algorithm is less expensive than some other known Pfair algorithms, but is not optimal for scheduling recurrent realtime tasks on more than two processors. In prior work, sufficient pertask weight (i.e., utilization) restrictions were esta ..."
Abstract

Cited by 3 (0 self)
 Add to MetaCart
The earliestpseudodeadlinefirst (EPDF) Pfair scheduling algorithm is less expensive than some other known Pfair algorithms, but is not optimal for scheduling recurrent realtime tasks on more than two processors. In prior work, sufficient pertask weight (i.e., utilization) restrictions were established for ensuring that tasks either do not miss their deadlines or have bounded tardiness when scheduled under EPDF. Implicit in these restrictions is the assumption that the total system utilization may equal the total available processing capacity (i.e., the total number of processors). This paper considers an orthogonal issue, namely, determining a sufficient restriction on the total utilization of a task set for it to be schedulable (i.e., a schedulable utilization bound) under EPDF, assuming that there are no pertask weight restrictions. We prove that a task set with total utilization at most 3M+1 4 is correctly scheduled under EPDF on M processors, regardless of how large each task’s weight is. At present, we do not know whether this value represents the worstcase for EPDF, but we do provide a counterexample that shows that it cannot be improved to exceed 86 % of the total processing capacity. The schedulable utilization bound we derive is expressed in terms of the maximum weight of any task, and hence, if this value is known, may be used to schedule task sets with total utilization greater than 3M+1
Improved Conditions for Bounded Tardiness under EPDF Pfair Multiprocessor Scheduling ✩
"... The earliestpseudodeadlinefirst (EPDF) Pfair algorithm is more efficient than other known Pfair scheduling algorithms, but is not optimal for scheduling recurrent realtime task systems on more than two identical processors. Although not optimal, EPDF may be preferable for realtime systems insta ..."
Abstract
 Add to MetaCart
The earliestpseudodeadlinefirst (EPDF) Pfair algorithm is more efficient than other known Pfair scheduling algorithms, but is not optimal for scheduling recurrent realtime task systems on more than two identical processors. Although not optimal, EPDF may be preferable for realtime systems instantiated on lesspowerful platforms, those with soft timing constraints, or those whose task composition can change at runtime. In prior work, Srinivasan and Anderson established a sufficient pertask utilization restriction for ensuring a tardiness of at most q quanta, where q ≥ 1, under EPDF. They also conjectured that under this algorithm, a tardiness bound of one quantum applies to task systems that are not subject to any restriction other than the obvious restrictions, namely, that the total system utilization not exceed the available processing capacity and pertask utilizations not exceed 1.0. In this paper, we present counterexamples that show that their conjecture is false and present sufficient pertask utilization restrictions that are more liberal than theirs. For ensuring a tardiness bound of one quantum, our restriction presents an improvement of 50 % over the previous one. Key words: soft realtime, multiprocessors, Pfairness, scheduling, tardiness bounds
Approved by:
, 2008
"... AARON D. BLOCK: Adaptive Multiprocessor RealTime Systems (Under the direction of James H. Anderson) Over the past few years, as multicore technology has become costeffective, multiprocessor systems have become increasingly prevalent. The growing availability of such systems has spurred the develop ..."
Abstract
 Add to MetaCart
AARON D. BLOCK: Adaptive Multiprocessor RealTime Systems (Under the direction of James H. Anderson) Over the past few years, as multicore technology has become costeffective, multiprocessor systems have become increasingly prevalent. The growing availability of such systems has spurred the development of computationallyintensive applications for which singleprocessor designs are insufficient. Many such applications have timing constraints; such timing constraints are often not static, but may change in response to both external and internal stimuli. Examples of such applications include tracking systems and many multimedia applications. Motivated by these observations, this dissertation proposes several different adaptive scheduling algorithms that are capable of guaranteeing flexible timing constraints on multiprocessor platforms. Under traditional task models (e.g., periodic, sporadic, etc.), the schedulability of a system is based on each task’s worstcase execution time (WCET), which defines the maximum amount of time that each of its jobs can execute. The disadvantage of using WCETs is
Hierarchical AED Scheduling Algorithm for RealTime Networks
, 2004
"... Abstract: Earlier studies have observed that in moderatelyloaded realtime systems, using an earliest deadline policy to schedule tasks results in the fewest missed deadlines. However, when the realtime system is overloaded an earliest deadline schedule performs worse than most other policies. Thi ..."
Abstract
 Add to MetaCart
Abstract: Earlier studies have observed that in moderatelyloaded realtime systems, using an earliest deadline policy to schedule tasks results in the fewest missed deadlines. However, when the realtime system is overloaded an earliest deadline schedule performs worse than most other policies. This is due to the earliest deadline giving the highest priority to the tasks that are close to missing their deadlines, thus delaying other transactions that might still be able to meet their deadline. In this research, an enhanced priority assignment algorithm is presented, called the Adaptive Earliest Deadline (AED), which features a feedback control mechanism that detects overload conditions and modifies packet priority assignments accordingly. Using a detailed simulation model, the performance of AED is compared and analyzed with Earliest Deadline First (EDF). Furthermore, an enhanced AED algorithm called the Hierarchical AED is proposed in a manner in which it obtains a better packetserving performance by using the concept of priority based on Quality of Service (QoS) of network traffic rather than using a random priority assignment when doing the packet group assignment. Finally, the performance of Hierarchical AED scheduling algorithm is compared with both EDF and the AED scheduling algorithms under the same operating environment.
Hybrid Multiprocessor RealTime Scheduling Approach
"... Realtime scheduling is one of the most important aspects of a realtime system design. To achieve a realtime system’s requirement, especially to be fast, multiprocessor systems are used. Generally, multiprocessor realtime scheduling algorithms fall into one the two wellknown approaches: Partitio ..."
Abstract
 Add to MetaCart
Realtime scheduling is one of the most important aspects of a realtime system design. To achieve a realtime system’s requirement, especially to be fast, multiprocessor systems are used. Generally, multiprocessor realtime scheduling algorithms fall into one the two wellknown approaches: Partitioning or Global. The partitioning approach has acceptable overhead for underlying system but can NOT guarantee to provide an optimal schedule. The global approach can provide this guarantee by holding some preconditions and considerable overheads. In this paper, an intermediate hybrid multiprocessor realtime scheduling approach is proposed in which optimality will be reached via the minimum overheads for underlying system. Presenting and analyzing different feasible paradigms for combination of the two existing approaches, the proposed hybrid approach satisfies the two major goals of this combination: optimality and lightweightness. Experimental results show that the hybrid approach outperforms the two existing ones.