Wyniki wyszukiwania - BazTech

Ograniczanie wyników

Znaleziono wyników: 2

Liczba wyników na stronie

Wyniki wyszukiwania

Wyszukiwano:
w słowach kluczowych: schedulability analysis

Sortuj według:

Ogranicz wyniki do:

A new sufficient schedulability analysis for hybrid scheduling

Zhang F., Zhai Y., Liao J.

International Journal of Applied Mathematics and Computer Science

2016

Vol. 26, no. 3

683--692

Earliest deadline first (EDF) and fixed priority (FP) are the most commonly used and studied scheduling algorithms for real-time systems. This paper focuses on combining the EDF and FP strategies in one system. We provide a new sufficient schedulability analysis for real-time hybrid task systems which are scheduled by EDF and FP. The proposed analysis has a polynomial time complexity and no restrictions on task parameters, where the relative deadline of each task could be less than, equal to, or greater than its period. By extensive experiments, we show that our proposed analysis significantly improves the acceptance ratio compared with the existing results of the sufficient schedulability test for hybrid scheduling systems.

Podejście probabilistyczne do zagadnień szeregowania zadań w systemach czasu rzeczywistego

Gajer M.

Kwartalnik Elektroniki i Telekomunikacji

2003

Vol. 49, z. 1

33-45

Systemy czasu rzeczywistego znajdują coraz powszechniejsze zastosowanie w urządzeniach telekomunikacyjnych, instalacjach militarnych, lotnictwie oraz w nowoczesnych zakładach produkcyjnych. Uogólniona teoria szeregowania według monotonicznego tempa (GRMS) stanowi stosunkowo nowe osiągnięcie informatyki teoretycznej, które wywołało duży wpływ na kierunek rozwoju wielu praktycznie zrealizowanych systemów czasu rzeczywistego. W artykule zaprezentowano podstawy teorii szregowania zadań GRMS. Następnie zaprezentowano nowatorskie statystyczne podejście do zagadnień związanych z procesem szeregowania zadań. Uzyskane rezultaty pozwalają na udzielenie odpowiedzi na pytanie dotyczące wartości prawdopodobieństwa zdarzenia polegającego na tym, że zadany w sposób arbitralny zbiór zadań okresowych będzie szeregowalny w przypadku zastosowania algorytmu GRMS. Wykazano, że wartość tego prawdopodobieństwa zależy w sposób istotny od stopnia ziarnistości zadań oraz od wartości współczynnika łącznego wykorzystania czasu pracy procesora przez wszystkie zadania.

At present the real-time systems constitute the well-defined class of computer systems. The real-time systems are getting moreand more popular in many fields of industry and communication. The real-time systems are used to control telecommunication devices and systems, defence systems, avionics and modern factories. In fact many modern facilities cannot do without them. For example without the real-time systems there would be no nuclear power plants, space ships, modern factories with robots etc. In the case of the real-time systems the computer program is organized into a set of task. Each task performs one of the system functions. In most cases the tasks are realized as periodic tasks. At the beginning of a periodic task some sensors are scanned, then some computations are performed, and in the end of the task the calculated control values are sent to the actuators. These operations are repeated continuously with a given period. Each task posses its deadline point, before which its execution must be finish. The most often the deadline is equal to elapse of the period time of the task. A very important matter is to guarantee that all the task are to be finished before their deadline points. In the case of the hard real-time systems any exceeding task deadline is absolutely intolerable. Such event if happens may lead to uncontrollable behavior of the system, which can cause a disaster, for example an aircraft crash, a loss of human life etc. To guarantee that the task deadlines will be always met, the task scheduling theory was developed. The main goal of the task scheduling theory is to demonstrate at the system development phase that under all the possible to foreseen circumstances the task deadlines will always be met. The so-called Generalized Rate-Monotonic Scheduling (GRMS) theory is a recent development that has had a large impact on the development of the real-time systems. The main ideals of the Generalized Rate-Monotonic Scheduling are the following. Each task is assigned a priority. The more often the task is activated, it is the shorter period it has, it is assigned the higher priority. In a given moment only the task with the highest priority is performed. Any activation of the task with higher priority is ready for the execution. The Generalized Rate-Monotonic Scheduling gives the sufficient condition on the task set schedulability. This condition says that the set of task is always schedulable if no more than 69% some sets of task may be not schedulable and some of them may be still schedulable, but this must be checked in each individual case. In the paper the thorough overview of the Generalized Rate-Monotonic Scheduling is presented. Further this author has proposed a novel statistical approach to periodic task scheduling. The purpose is to answer the question about the value of probability that the task set which uses more than 69% of the processor time is scheduable. In the paper this probability values were measured during the computer simulations of the scheduling process and put into tables. The conducted numerical experiments revealed that the value of probability depends strongly on the tasks granularity level, it is the granularity is finer the probability that the set of tasks is schedulable is higher.