An Algorithm to Reduce the Time Complexity of Earliest Deadline First Scheduling Algorithm in Real-Time System
In this paper I have study to Reduce the time Complexity of Earliest Deadline First (EDF), a global scheduling scheme for Earliest Deadline First in Real Time System tasks on a Multiprocessors system. Several admission control algorithms for Earliest…
Authors: Jagbeer Singh
An Algorithm to Reduce the Time Complexity of Earliest Deadline First Scheduling Algorithm in Real-Time System Jagbeer Sin gh Dept. of Computer Science and Engine ering, Gandhi Institute of Engineering and Technology Gunupur, Rayagada (Orissa), India-765022, E-mail:- willybokadia@gmail.com , Mob No. +91943 9286506, +917873 887979 ABSTRACT In this paper I have study to Reduce the time Complexity of Earliest Deadline First ( EDF ), a global scheduling scheme for Earliest Deadline First in Real Time System tasks on a Multiprocessors system. Several admission control algorithms for Earliest Deadline First ( EDF ) are presented, both for hard and soft real-time tasks. The average performance of these admission control algorithms is co mpared with the performance of known partitioning schemes. I have applied some modification to the global Earliest Deadline First ( EDF ) algorithms to decrease the number of task migration and also to add predictability to its behavior. The Aim of this work is to provide a sensitivity analysis for task deadline context of multiprocessor system by using a new approach of EFDF (Earliest Feasible Deadline First) algorithm. In order to decrease the number of migrations we prevent a job from moving one processor to another processor if it is among the m higher priority jobs. Ther efore, a job will continue its execution on the same processor if possible (processor affinity). The result of these comparisons outlines some situations where one scheme is preferable over the other. Partitioning schemes are better suited for hard real-time systems, while a global scheme is preferable for soft real-time systems. Key Words- Real-time system, task migration, earliest deadline first, earliest feasible deadline first. INTRODUCTION In Earliest Deadline First ( EDF ) scheduling, at every scheduling point the task having the shortest deadline is taken up for scheduling. The basic principle of this algorithm is very intuitive and simple to understand. The schedulability test for EDF is also simple. A task is schedule under EDF , if and only if it satisfies the condition that total processor utilization ( u i ) due to the task set is less than 1 . With scheduling periodic processes that have deadlines equal to their periods, EDF has a utilization bound of 100%. Thus, the schedulability test for EDF is: where the are the worst-case computation-times of the n processes and the are their respective inter- arrival periods (assu med to be equal to the relative deadlines). For example let us Consider 3 periodic processes scheduled using EDF , the following acceptance te st shows that all deadlines will be met. Process Execution Time = C Period = T P1 1 8 P2 2 5 P3 4 10 The utilization will be: The theoretical limit for any number of processes is 100% and so the system is schedulable. EDF has been proven to be an optimal uniprocessor scheduling algorithms [8].This means that if a set of tasks is unschedulable under EDF , then no other scheduling algo rithm can feasible schedule this task set. The EDF algorithm chooses for execution at each instant in the time currently activ e job(s) that have the nearest deadlines. The EDF implementation upon uniform parallel machines is according to the following rules [2], No Processor is idled while there are active jobs waiting for execution, when fewer then m jobs are active, they are required to execute on the fastest processor while the slowest are idled, and higher priority jobs are executed on faster processors. A formal verification which guarantees all deadlines in a real-time system would be the best. This verification is called feasibility test. Three different kinds of tests are available:- • Exact tests with long execution times or simple models [11], [12], [13]. • Fast sufficient tests which fail to accept feasible task sets, especially those with high utilizations [14], [15]. • Approximations, which are allowing an adjustment of performance and acceptance rate [1], [8]. For many applications an exact test or an approximation with a high acceptance rate Must be used. For many task sets a fast sufficient test is adequate. EDF is an appropriate algorithm to use for online scheduling on uniform multiprocessors. However, their implementation suffers from a great number of migrations due to vast fluctuations caused by finishing or arrival of jobs with relatively nearer deadlines. Task migrati on cost might be very high. For example, in loosely coupled system such as cluster of workstation a migrat ion is performed so slowly that the over load resulting from excessive migration may prove unacceptable [3]. Another disadvantage of EDF is that its behavior becomes unpredictable in overloaded situations. Therefore, the performance of EDF drops in overloaded condition such that it cannot be considered for use. In this paper I am presenting a new approach, call the Earliest Feas ible Deadline First ( EFDF ) which is used to reduce the time complexity of earliest deadline first algorithm by some assumptions. REVIEW OF RELATED WORK Each processor in a uniform multiprocessor machine is characterized by a speed or Computing capacity, with the interpretation that a job executing on a processor with speed s for t time units completes ( s * t ) units of execution. The earliest-deadline first ( EDF ) scheduling of real-time systems upon uniform multiprocessor machines is considered. It is known that onli ne algorithms tend to perform very poorly in scheduling Such real-time systems on multiprocessors; resource-augmentation techniques are presented here that permit online algorithms in general ( EDF in particular) to perform better than may be expected given these inherent limitations. It is shown that EDF scheduling upon uniform multiprocessors is robust with respect to both job execution requirements and processor computing capacity. PROPOSED APPROACH I have applied some m odification to the global Earliest Dead line First ( EDF ) algorithms to decrease the number of task migration and also to add predictability to its behavior. In order to decrease the number of migrations we prevent a job from moving to another processor if it is among the m higher priority jobs. Ther efore, a job will continue its execution on the same processor if possible ( processor affinity 1 ). In Earliest Deadline First ( EDF ) scheduling, at every scheduling point the task having the shortest deadline is taken up for scheduling. The basic principle of this algorithm is very intuitive and simple to understand. The schedulability test for Earliest Deadline First ( EDF ) is also simple. A task is schedule under EDF , if and only if it satisfies the condition that total processor utilization due to the task set is less than 1. For a set of periodic real- time task { T 1 , T 2 , T n }, EDF schedulibility criterion can be expressed as:- Where e i is the execution time, p i is the priority of task and u i is the average utilization due to the task T i and n is the total number of task in set. EDF has been proven to be an optimal uniprocessor scheduling algorithm [8]. This means that if a set of task is unschedulable under Ea rliest Deadline First ( EDF ), then no other scheduling algorithm can feasible schedule this task set. In the simple schedulability test for EDF we assumed that the period of each task is the same as its deadline. However in practical problem the period of a task may at times be different from its deadline. In such cases, the schedulability test needs to be changed. If p i >d i , then each task needs e i amount of computing time every min( p i , d i ) duration time. Therefore we can write: However, if p i 0 for all j. In this section I am presenting five steps of EFDF algorithm. Obviously, each task which is picked for up execution is not considered for execution by other processors. 1. Perform a feasibility check to specify the task which has a chance to meet their deadline and put them in a set A , Put the remaining tasks in set B . 2. Sort both task sets A and B according to their deadline in a non-descending order. Let k denote the number of tasks in set A , i.e. the number of tasks that have the opportunity to meet their deadline. 3. For all processor j , ( j ≤ min ( k,m )) check whether a task which was last running on the j th processor is among the first min( k,m ) tasks of set A . If so assign it to the j th processor. At this point there might be some processors to which no task has been assigned yet. 4. For all j , ( j ≤ min ( k, m )) if no task is assigned to the j th processor , select the task with earliest deadline from remaining tasks of set A and assign it to the j th processor. If k ≥ m , each processor have a task to process and the algorithm is finished. 5. If k
Original Paper
Loading high-quality paper...
Comments & Academic Discussion
Loading comments...
Leave a Comment