Real time system scheduling is the way used for
managing the order of the job which is performed by a CPU of a
computer. The goal of scheduling is to engage the CPU at its
maximum capacity limit and no process shall wait for longer
time period and to finish the entire task in small possible time.
In this paper, we discuss various types of Scheduling algorithms
and Compare their performance on terms of throughput and
waiting period. The four basic type of Scheduling are: Round
Robin (RR) scheduling, First Come First Served (FCFS),
Priority Based scheduling, Shortest Job First (SJF). These entire
algorithms scheme have some drawback and have not been
optimized yet. In this paper we propose a Median based .Time
quantum based scheduling algorithm which is combination of
SJF & RR where all the jobs in the queue are first aligned as
per their burst time in ascending order and them Round robin is
applied for improving the performance. We are mainly work on
DPZL and FPZL scheduling algorithms.
Priya Yeole : Student, Department of Computer Technology, RGCER, Nagpur-441110, Maharashtra, India
Monika Atey : Student, Department of Computer Technology, RGCER, Nagpur-441110, Maharashtra, India
Ashwini Pathade : Student, Department of Computer Technology, RGCER, Nagpur-441110, Maharashtra, India
Manjusha Badge : Student, Department of Computer Technology, RGCER, Nagpur-441110, Maharashtra, India
Scheduling
Multiprocessor
Computation Time
Multicore
Multithreaded
FPZL & DPZL
The motivation for our work was the desire to improve
upon current state-of-the-art global scheduling methods
for hard real-time systems in terms of practical techniques
that enable the efficient use of processing capacity. The
intuition behind our work was that dynamic priority
scheduling has the potential to schedule many more task
sets than fixed task or fixed job priority algorithms, and
yet this theoretical advantage has to be tempered by the
need to avoid prohibitively large overheads due to a high
number of pre-emption. This led us to consider minimally
dynamic scheduling algorithms which permit each job to
change priority at most once during its execution. One
such algorithm is EDZL. We applied the zero-laxity rule
from EDZL to global FP scheduling, forming the FPZL
scheduling algorithm. The number of context switches
with FPZL is at most two per zero-laxity task, and one per
ordinary task. As there are at most m zero-laxity tasks, the
increase in overheads compared to global FP scheduling is
tightly bounded.
[1] Huang Shujuan and Zhu yian, “Base-Utilization Partitioning
Algorithm for multiprocessor” College of Computer Science
&Engineering Northwestern Polytechnic University 2011.
[2] Euiseong Seo Jinkyu Jeong, Seonyeong Park, and Joonwon
Lee., "Energy Efficient Scheduling of Real-Time Tasks on
Multicore Processors", IEEE transactions on parallel and
distributed systems, vol. 19, no. 11, November 2008, pp 1540-
1552.
[3] Enrico Bini, Thi Huyen Chau Nguyen, Pascal Richard, and
Sanjoy K. Baruah, "A Response-Time Bound in Fixed-Priority
Scheduling with Arbitrary Deadlines", IEEE Transactions On
Computers, Vol. 58, No. 2, February 2009, pp. 279-286.
[4] Robert I. Davis and Alan Burns, “FPZL Schedulability
Analysis” 2011 17th IEEE Real-Time and Embedded
Technology and Applications Symposium