Site Loader
Rock Street, San Francisco

Abstract

 

OS by a thought
of multi-tasking is applied to raise the ability of the central processor. Multi-scheduling
managed several applications on one processor and plenty of users worked it at
that time. Scheduling provides the simplest way to envision a sequence of a
processor through they that dispatch and keep it at busy. Multiple programming
algorithms are applied for this purpose. Throughout this paper, we tend to use
the hybrid approach to overcome the obstacle of Starvation.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!


order now

Keyword: CPU Scheduling, Multi-scheduling,
Operating System, Resource Scheduling, Time quantum 

1. Introduction

OS
implements the Associate interface between a system and user credentials. Just
in case of the one processor, that process moves in watching for until the
C.P.U stays available. It takes time and that we do not work on multiple
programs through that. To resolve this issue we tend to use multitasking during
which many processors work at the same time and increase the potency of C.P.U.
The main plan of multitasking is to share resources among totally different
processes. Around every one of the resources is processed in a very typical
means before doing it. Scheduling provides a more robust and effective way to
improve the performance of C.P.U. Scheduling provides a more robust route to
watch that process is to run among all the processes.

 

Scheduling
implies once when we need to realize high-speed computer operation and
management multiple programs on a system.CPU scheduling is vital thanks to its
impacts toward the resource allocation, C.P.U. utilization, turnaround
interval, waiting time, output and different performances. Existing C.P.U.
planning algorithms are FCFS, Shortest-Job-First (SJF), Round Robin (RR), and
Priority primarily based planning. Those algorithms are applied to enhance the
efficiency of C.P.U. and minimize the waiting time, work time, waiting time and
amount of context change. There is some planning algorithmic program that
determines that scheduling algorithmic program dead all its parameter and
provides the higher result. These are some planning parameter, on the idea of
those parameters we tend to decide that that one scheduling program is best.                          

These
are some scheduling parameter and that we want to use that rule which can give
higher result in keeping with the state of affairs and setting. These are some
as follow:

Context-Switching:
It’ll happen once one process will interrupt the sequence of execution of a
process. We want to use that kind of scheduling that reduces context switch as
a result of it’s the wastage of time and memory

 

.
CPU Utilization: CPU idles once the CPU works on 100 percent that’s not the
fact. Real-time operating system, CPU work on 40% to 90% which is said to be
lightly loaded too heavily loaded.

 

 Turnaround Time: Time needed for a selected
process to its completion in from ready to its execution.

Waiting
Time:  once the process is in ready queue
and awaiting its flip. A process executed properly when getting into its
execution queue. We want to use that scheduling rule that reduced the waiting
time for a process.

Response Time:
It takes the time to start out the execution of a process and CPU performs in
it’s a way-way when once we minimize the interval time. To overcome that
drawback to achieved the best CPU utilization.

 

 That paper is divided into portions. Section 1
is regarding previous work which others have done with the same topic. Section
2 based on the proposed algorithm which one is better than all others. Section
3 based on results and discussion of our paper. Section 4 based on the
conclusion and section 5 described the future work, how can improve the efficiency
of real-time operating systems.

 

2. Literature Survey

 

There
are many techniques works done on Scheduling algorithm that worked at a time of
arrival and latency. Working on Scheduling improved with the passage of time.
The author (Chhugani & Silvester, 2017) worked on dynamic time quantum
that calculates the parameter of scheduling. The result shows that the way to
increase the time quantum for few processor because of a threshold value. The
author (Rajput & Gupta, 2012) proposed an algorithmic program
that supported priority primarily based algorithmic program and compares with
normal round robin.

 

 The fuzzy technique supported pre priority and
execution time and compare with the various algorithmic program and shows a
higher result in (Kumari, Sharma, & Kumar, 2017. V FJFDRR targeted
on round robin with dynamic time slice and compare with the various
technologies and shows the higher result in {Mohanty, 2012 # 4).(Mohanty & Panigrahi, 2012) planned a replacement technique
that calculates fit factor and dynamic time slice. fit factor supported the
mixture of FCFS, SJF, and priority algorithmic program and show the higher
result as compared to the other scheduling algorithmic program. SJFDRR works on
time quantum and improves the potency of round robin. In this paper, there are
user and system priority. User priority has necessary than system priority and
reduced the context switching in (Gupta, Yadav, & Goyal, 2016). Self Adjustment round Robin
(SARR) solves the matter of dynamic time-quantum that regulate the burst time
in line with the running algorithm. The planned algorithmic program also can be
enforced on a large processor and also the software system itself can determine
the optimum time quantum in (Matarneh, 2009). (Mohanty, Behera, Patwari, Dash, & Prasanna, 2011) represents the algorithmic
program that is called priority-based dynamic round robin that calculates
intelligent time slice for the individual process and changes the time slice
before every execution. FPRRDQ shows the higher result as compared to other
alternative programs that are based on the user priority and quantum time t
when each execution in line with priority and burst time in (Srivastav, Pandey, Gahoi, & Namdev, 2012).

Optimum service
time conception for round-robin algorithmic program works on an associate
optimum priority of every process associated placed in an order of execution in
line with calculated priority in (Saxena & Agarwal, 2012) FCFS work on the idea of the
FIFO. Each process executed according to its number.

 FCFS performs well for smaller values. It
shows poor waiting time, a turnaround for giant computation.SJF worked on the
idea of shortest CPU burst length. Within which short process enter in
execution queue and execute first. SJF perform best for long processes as
compared to FCFS. It’s potential that long Process waits within the prepared
queue for the brief process that complete its task however typically it behaves
like starvation.

 

 RR worked in time quantum. RR worked sensible
for short process and provides the results of minimum average time, minimum
turnaround and minimum throughout. In real time system, the overhead invokes
when every context switch because of context switch exaggerated for short time
quantum. Just in case of long-time quantum, the process executes inside
one-time slice and performs higher result. The priority-based algorithmic
program worked on low and high priority. Typically it becomes suffer a serious
downside known as starvation as a result of low priority failed to execute
because of high priority.  To avoid the
matter of overhead and starvation, a replacement technique should be introduced
to resolve this downside and average waiting time, average turnaround and
average response time should be increased.

 

3. Proposed Algorithm

 

Scheduling
is the technique used to enhance the performance of the CPU. To increase the
CPU utilization and reduced the average waiting time, average turnaround time
and average response time.CPU scheduling algorithm worked on maximize
throughput. I used two CPU scheduling algorithms and combined them in one that
is SJF and round robin. Both can combine and generate new technique that
behaves well effective. In this Technique, the processor is in ready queue in
according to CPU burst length, Shortest burst length is at the top of the
queue. We tend to assume two numbers to represent the burst length of the
largest PCB within the queue and the second one to represent the running time
of all the processes respectively.

 

 A Process control block (PCB) of a process is
often submitted to the system which is connected to the ready queue in
according to the CPU.

The
proposed algorithm that is executed by the CPU linked to the process from the
top of the queue. Executed Process is expired after a given time quantum, which
is defined by the system. After that, new preemption is as follow:

te
= te + quantum time   

Time
quantum applies to boost the efficiency and minimize the average waiting time
average turnaround time and average waiting and context switching between the
processes.

 

In
that case, five states are in the process which is new, ready, running, block
and complete state. The new state admitted the process and dispatch to the
ready state. The ready queue then moves forwards the process to the running
state. If an interrupt occur on ready state then it will back to the ready
state if the processor requires an I/O device then it moves to the block state
and if the process completed then it moves to the complete state. Block State
complete the requirement for the processor such that I/O and then moved to the
ready queue. Comparison of two numbers is as fellow:

 

If
execution time of a process te is less than the largest burst length of the PCB
to then the preempted process PCB is joined to the tail of the ready queue.
After that, the next process is then dispatched from the top of the ready
queue.

If   te ? to

Then
the Process control block (PCB) of the process with the largest CPU burst
length is to start the execution.

 

In
Preemption, SJF is in the ready queue that’s why shortest job entertained
first. The value of te is reset to 0 and the value of the CPU burst length of
the largest PCB is reset that is lying at the tail of the queue. After that,
the next process is then moving towards from the head of the ready queue.

When
a process has accomplished its task it terminates and deleted from the system.
Then te will be:

te
= te  + time to complete process

Process
and actions are same as a preempted process.

 

4. Results and Discussions

 

Proposed
algorithmic rule supported round Robin and also Shortest Job first. It performs
higher result and enhances the efficiency of a central processing unit. This
shows better results on average waiting time, average turnaround and average
response time comparative other algorithms and shows optimum results.

 

5. Conclusions

 

Scheduling
policies play a vital role to enhance the efficiency of the system. Proposed
solution provides an optimum and better solution for the system through which
the real-time system performs in a genuine way. By using this, we want to
overcome the problem of starvation and overhead. Performance of time-sharing
system can be improved through this algorithm.

 

6. Future Recommendations

 

In
Future, we can also enhance the performance of CPU in real-time systems through
embedded different scheduling algorithms. Compare with other techniques and can
generate a better response time to intensify the performance of a real-time
system.

Post Author: admin

x

Hi!
I'm Velma!

Would you like to get a custom essay? How about receiving a customized one?

Check it out