Abstract
This
thesis is an extension of existing fuzzy scheduling algorithms, to schedule
real-time tasks on multiprocessor systems. The existing fuzzy scheduling
algorithm was extended in order to have a better performance at higher system
load. The task scheduling parameters used in this research work are; arrival
time, computation time, and deadline which are inputs to the fuzzy inference
system. The outputs are the runtime priorities which are used to schedule tasks
in a priority (ready) queue for execution on multiprocessor. The performance of
the new fuzzy scheduling algorithm was compared with the existing fuzzy
scheduling algorithm. Results show that the new fuzzy algorithm has a better
performance than the existing algorithm at higher system load; this is as a
result of minimum response time, turnaround time and lower number of tasks that
missed their deadline.
CHAPTER
ONE
GENERAL
INTRODUCTION
1.0
INTRODUCTION
Scheduling
algorithm is more significant in a real-time system to ensure desired and
predictable behaviour of a system (Sagar et al., 2012). Scheduling is
concern with the optimal allocation of resources to tasks in a specified time.
Real-time systems are often embedded in most of domestic devices which are used
for daily activities without our knowledge when we use the devices in which
they reside. Cars, planes and entertainment systems are just some devices in
which real-time systems reside, governing the workings of that device while we
do not consider that such a system exists within the chosen device. A
real-time-system is a computer system in which the key aspect of the system is
to perform tasks on time, not finishing too early nor too late. A classic
example is that of the air-bag in a car; it is of great importance that the bag
inflates neither too soon nor too late in orders to be of help and not be
potentially harmful. In real-time systems, all tasks have specific parameters
such as execution time, deadline, arrival time, priority, etc. Modern embedded
computing systems are becoming increasingly complex (Sagar et al., 2012).
Many
real-time systems are soft in which missing the deadline is tolerable while
some others are hard and missing deadline is catastrophic (Sabeghi et al., 2006).
Nowadays, the use of real-time multiprocessor systems is dramatically
increasing. Unfortunately, little is known about how to schedule
multiprocessor-based real-time systems than that for uni-processors. The
multiprocessor based scheduling have more computational complexity in practical
algorithm which are unknown to most researchers, this open floor for new area
of research in operating systems (Shahzad and Afzal, 2006). It has been
provened that finding a minimal schedule for a set of real-time tasks in
multiprocessor system
is a hard problem (Tanenbaum and Woodfhull, 2005). In this research fuzzy
technique was introduced to schedule tasks on multiprocessors in real-time
system. Many scheduling algorithms have been studied to guarantee the time
constraints of real-time. Scheduling decision of real-time algorithms is
usually based on parameters which are assumed to be crisp. However, in
uncertain and imprecise environment the values of these parameters are vague.
The vagueness of these parameters suggests that we make use of fuzzy logic to
capture uncertainty and imprecision mode. Fuzzy scheduling approach therefore
help to decide in what order the tasks should be scheduled for execution to
better utilize the system and as a result reduce the chance of task’ deadline
being missed.
1.1
PROBLEM STATEMENT AND RESEARCH MOTIVATION
There
are many conventional approaches to scheduling algorithms to guarantee the time
constraints of real-time processes but no one is absolutely ideal. Scheduling
decision of these algorithms is usually based on parameters which are assumed
to be crisp. All practical real-time scheduling algorithms in multiprocessor
systems present a trade-off between their computational complexity and
performance. In real-time systems, tasks have to be performed correctly and
timely. Finding minimal schedule in multiprocessor systems with real-time
constraints is shown to a hard problem (Sheo et al., 2012). Although
some optimal algorithms have been employed in uni-processor systems, they fail
when they are applied in multiprocessor systems. However, the existing
algorithm has a better performance at normal system load but at overloaded
condition the performance decreases. Therefore, this research work will focus
on the development and implementation of a new fuzzy algorithm for fair
scheduling of real-time tasks on multiprocessor in order to improve performance
of an existing system at higher system load.
MSC Project Topics and Complete Thesis in Computer Science
NEW FUZZY SCHEDULING ALGORITHM (NFSA) FOR REALTIME TASKS ON MULTIPROCESSOR SYSTEMS
Department: Computer Science (M.Sc Thesis)
Format: MS Word
Chapters: 1 - 5, Preliminary Pages, Abstract, References, Appendix.
No. of Pages: 92
NB: The Complete Thesis is well written and ready to use.
Price: 20,000 NGN
In Stock
Our Customers are Happy!!!
No comments:
Post a Comment
Add Comment