Dynamic Three Stages Task Scheduling Algorithm on Cloud Computing

(IJCSIS) International Journal of Computer Science and Information Security,
Vol. 18, No. 6 June 2020
Dynamic Three Stages Task Scheduling Algorithm
on Cloud Computing
Naglaa Sayed Abdelrehem Fathi Ahmed Amer Imane Aly Saroit
Department of Information Technology
Faculty of Computer and Artificial Intelligence
Cairo University
Cairo, Egypt
Naglaasayed.fci@gmail.com fathi.amer.csis@o6u.edu.eg i.saroit@fci-cu.edu.eg
Abstract— Scheduling process is one of the main challenges in
cloud computing to manage and coordinate between tasks and
their appropriate resources, to get the best and most efficient use
of the available cloud resources. This paper proposes a cloud
scheduling mechanism that works as a three- stage strategy. In
the first stage, a task classification is performed using a job
classifier to pre-create different types of Virtual Machines (VMs),
which saves the time needed through the scheduling process to
create these VMs and decreases failure rate. In the second stage,
tasks are sorted based on their priority, and then check if their
expected execution time is less than or equal to the deadline to
indicate the state of the VM that satisfies the deadline constraint
as successful and reject the tasks that can’t be executed within
their deadline and save them in the database to be detected later.
In the third stage, the tasks are paired dynamically with their
matching VMs with minimum completion time. In order to
evaluate the proposed protocol; a simulation is performed using
the cloud sim plus simulator to simulate the proposed algorithm
and compare it with the Min-Min standard algorithm and the
two-stage scheduling algorithm to show that the proposed
algorithm reduced the average waiting time, the average
makespan, and failure rate and maximized the virtual machine
utilization rate, task guarantee ratio and the VMs load balancing
compared to the two other algorithms.
Keywords- Cloud Computing; Scheduling; Virtual Machines
(VMs), Makespan; Waiting time; Resource
Utilization; Failure rate.
I. INTRODUCTION
Cloud computing is known as an on demand sharing of
resources, services or infrastructure over the internet and
paying only for what is used. Tasks are scheduled depending
on the user different needs [1] [2]. To organize the resource
usage on the cloud, and find the appropriate deployment
method we need to find a suitable scheduling mechanism to
get the efficient use of cloud resources with the minimum
costs. The cloud scheduling process aims to define the most
suitable deployment method to satisfy the user’s requirements
and to help the service providers to get the highest economic
benefits [3]. Many different cloud applications are received by
the data center to get services using the pay-per-use policy.
Attributed to the limited resources with different
functionalities and different capacities on the cloud, Cloud
scheduling has turned into a challenging process [4]. There
are various scheduling algorithms suggested by different
researchers to define the most convenient deployment method
of the resources in the cloud [5] [6]. There are many tasks
scheduled in different cloud environments with different
Quality of Service (Qos) requirements [7]. Some of the
suggested cloud task scheduling mechanisms aim to optimize
the deadline [8]. Others suggested enhancements in the load
balancing [9] [10]. Others suggested to optimize the quality of
service requirements (Qos) and maximize the total revenue
[11], or minimize the costs [12]. Others suggested to get the
best service level agreement and the best energy consumption
level [13]
The cloud scheduling process became an urgent and
sensitive matter to find the best deployment for cloud
resources, which helps enhance the cloud overall performance,
increase the quality of the service, minimize the costs and
failure rate, maximize the utilization and total revenue. The
main problems we face during the scheduling process is to
find the most convenient pair of tasks and VMs to be matched
with, the waiting times sometimes be too high and many tasks
may access the system and waiting to be processed then fail as
they exhausted the deadline constraint. In order to resolve this
problem a new scheduling algorithm is introduced in this
paper to organize the task and virtual machine mapping
process; based on a dynamic three stage strategy where in the
first steps it detects the common task and VM types based on a
historical stored data set; which helps to predict and pre-create
a convenient number of VMs based on the types in the
database; this step saves the time needed to create the tasks
during the scheduling process. The algorithm starts to receive
dynamic task sets in the cloud, classify based on the historical
database, checks first if the coming tasks can be executed
within their deadline to follow the scheduling sequence, or
56 https://sites.google.com/site/ijcsis/
ISSN 1947-5500

Vol. 18, No. 6 June 2020
they can’t and will be rejected to avoid waiting time and be
saved in the historical database to be detected later. In case of
successful tasks that matches the deadline constraint, some
may don’t find a suitable matching VM from the same type, so
that the algorithm will fetch for the most similar available VM
with the minimum completion time. If they can’t find a
convenient one, tasks will be saved in a waiting queue to
create a convenient matching one. This helps to minimize the
waiting time, makespan and the task guarantee ratio, also can
enhance the load balance among several VMs.
The paper is organized as follows; section 2 reviews some of
the related work scenarios. Section 3 explains the proposed
algorithm in details. Section 4 shows the simulation and
evaluation of the proposed scenario and comparison results
with other algorithms. Section 5 concludes the work and show
the suggested future work.
II. RELATED WORK
SaeMi Shin and SuKyoung Lee suggested a scheduling
mechanism to receive all the data center jobs, sort the jobs
in an ascending order to serve the high priority jobs first
then choose the largest backfill job to satisfy the deadline
guarantee constraint. Their algorithm enhanced the
scheduling performance by enhancing the deadline
guarantee ratio and the utilization of resources [2]. Atul
Vikas Lakraa and Dharmendra Kumar Yadavb showed a
simplified survey on different cloud task scheduling
algorithms; each algorithm enhanced one or more of these
factors (quality of service, load balancing, minimized
makespan, consistency, maximum resource utilization,
energy efficiency, effective implementation, and fairness
among tasks, high profits and bandwidth utilization).
Meanwhile, all the algorithms have a problem that they
cannot enhance all of them together so no one of them can
achieve 100% efficiency [4]. IM.Vijayalakshmi, and
IIV.Venkatesa Kumar made a survey on different
scheduling techniques like Round Robin, Minimum
Completion time, Random Resource Selection, and Load
Balancing algorithms, where the RR had minimum cost
compared to the minimum completion time and the load
balancing algorithms. In terms of the total cost, Random
algorithm is the best. Random algorithm and the Round
Rubin algorithm got the same cost when increasing the
number of jobs [6]. Mousa and Abdelouahed Gherbi
suggested a scheduling mechanism that divides tasks into
different groups using Ram and CPU utilization based on
data from log files where various tasks can share the same
VM resources, where their algorithm enhances the QoS
requirements, maximizes the resources usage, increases the
user satisfaction and reduces the number of job rejections
[10]. PeiYun Zhang and MengChu Zhou proposed a cloud
scheduling technique that works as a strategy of two stages.
In the first stage, a task classification is performed
depending on past documented data. In the second stage,
there is matching of the tasks dynamically to their
corresponding VMs. Their algorithm enhanced the load
balancing and the scheduling performance compared to
Min-Min and Max-Min algorithms [14]. Mokhtar A.
Alworafi and Suresha Mallappa suggested a cloud
scheduling mechanism to sort the tasks according to the
length priority in an ascending order, then they referred the
state of the virtual machine that satisfies the deadline
constraint as successful. After that, they pair each task to
its convenient VM. Their algorithm enhanced the task
guarantee ratio and the utilization of resources, and
reduced the average response time and the makespan
compared with other algorithms like (Min-Min, GA, SJF,
and Round Robin) [15]. Dr. Amit Agarwal and Saloni Jain
suggested a Generalized Priority Algorithm where the
priority is defined with respect to the user demands. Their
algorithm has a minimum execution time compared with
(FCFS) and (RR) algorithms [16]. Xiaoping Li and Rub´en
Ruiz suggested a task scheduling mechanism that combines
the priority scheduling algorithm and the RR algorithm to
enhance performance by enhancing the execution time and
throughput values by combining the tasks into two groups
(deadline-based and cost-based). The tasks arranged
ascending based on the deadline and the cost arranged
descending based on the task length [17]. Mehwish
Awan*, and Munam Ali Shah suggested a multi-objective
scheduling algorithm to relate a group of tasks received by
the broker to the received virtual machines list. They
reduced the execution time of workload to the minimum
optimized time. They compared their algorithm to (FCFS)
algorithm and priority scheduling algorithm [18].
III. PROPOSED ALGORITHM
In this section, we explain, the proposed cloud scheduling
strategy named the Dynamic Three Stages scheduling
algorithm. The proposed algorithm aims to find the best
sequence to be followed as a task scheduling mechanism
that maps the tasks dynamically to the most appropriate
virtual machine with the minimum time and the maximum
utilization rate, the algorithm acts as a series of three stages
as we will see in the next sections.
A. The First Stage
In the first stage, a task classification is performed using
a job classifier depending on the last stored historical data
and the cloud environment’s current state. This step saves
the time needed to create virtual machines during the
scheduling process and also decreases the task scheduling
failure rate as it pre-creates a convenient number of the
various predicted virtual machine types.
ISSN 1947-5500

Vol. 18, No. 6 June 2020
Fig. 1 illustrates the sequence followed in the first stage of
the algorithm in brief:
Figure 1: The First Stage Sequence
B. The Second Stage
The second stage works on decreasing the average
makespan and maximizing the resource utilization. The
tasks are sorted depending on the priority of their lengths, in
order to process critical or high priority tasks first and then
the ordinary tasks, then comes the time to check that the
expected execution time for a task is less than or equal to
the task deadline, which means that the task can be
executed within its deadline. After that, the state of the
virtual machine that satisfies the deadline constraint is
referred as successful to complete the process, and discard
the unsuccessful ones after saving them in the database to
be used later as a historical data and create convenient VMs
that can execute them easily.

Fig. 2 illustrates the sequence followed in the second
stage of the algorithm in brief:
Figure 2: The Second Stage Sequence
C. The Third Stage
The third stage is a mapping of the tasks dynamically to
the convenient virtual machines, if they do not have a
convenient matching with a VM from the same type
available, we check if we have extra available VMs from
other similar types to match with, which helps to enhance
the load balancing among all the available VMs and saves
the waiting time the task need to wait in the waiting queue
until we create a matching VM from the same type,
Otherwise, we put the task in the waiting queue to create a
convenient VM type during the dynamic scheduling
process. The algorithm is compared with two old
algorithms which is the Min-Min and the two stages
algorithms.
Figure 3 illustrates the sequence followed in the third stage
of the algorithm in brief:
Figure 3: The Third Stage Sequence
D. The task and VM mapping
Assume we have a set of VMs defined as V = {1, 2… N}
and Vi, where i ∈ {1, 2... N}, represent the VM number i, Vi
defined by four attributes denoted as Vi (a), where a ∈ {1, 2,
3, 4}. Which represent the CPU resources like the CPU clock
speed, memory resources, the network bandwidth and the
hard disk storage, respectively. Where, we have Vi = {Vi (1),
Vi (2), Vi (3), Vi (4)}.
Assume we have a set of tasks provided by the users
defined as T = {1, 2, ... , M} and Tj , where j ∈ {1, 2, ... , M},
represent the task number j. where the task j can be defined
based on some attributes as Tj ={Tj(id) , Tj(r), Tj(d), and
Tj(p),Tj(L) } , where Tj (id) is defined as the unique ID of task
j, Tj (r) is defined as the requirements of task j . and Tj(r) =
{Tj1, Tj2,Tj3, Tj4} specifies the requirements for CPU, memory,
network bandwidth and hard disk storage for task Tj; Tj(d) is
defined as the deadline of task Tj. When the deadline of Tj is
ISSN 1947-5500

Vol. 18, No. 6 June 2020
violated, the task is failed to be scheduled, Tj (p) is defined as
the priority of task j, if Tj is urgent or high payment user’s
job, it is a high priority task otherwise it is a regular job.
Assume that the cloud holds a set of hosts defined as Hk,
where k ∈ {1, 2... K}, represent the host number k that can
create Gk VMs, i.e., Hk = {vnk |n ∈ {1, 2... Gk}}. Where vnk
represents the VM number n of the host number k.
To reduce the complexity of the mapping, we divide tasks
that need to be matched with VMs into task types, assume we
have a set of VM types called VMtype = {1, 2... L}. Where L
represents the number of VM types.
Assume that the data center DC has a set of servers
defined as S= {S1, S2 . . . Ss}, Si = {Vi1, Vi2 . . . V iN} is a set of
virtual machines in the server Si. Each virtual machine has a
specified speed defined by the number of million
instructions per second (MIPS). And the speed of the VM is
defined as (Vs), where the number of instructions per task
(task length) is defined as (TL).
The expected execution time (EET) of the task in each
VM is computed, and compared with the deadline constraint
value to find which of the VM achieve the deadline to be
defined as a successful task to complete the scheduling
sequence. Or which of them don't achieve the deadline to be
defined as an unsuccessful task and saved to the database to
be defined later, the expected execution time is calculated as
in equation 2 [15]:

Where defines the task length, and defines the VM
speed.
The task scheduling process here is defined as a function
that maps tasks to the convenient VM types. We can
compute the matching degree using equation 3 [14].

Where is the maximum Vk (a), k represents the
type-k virtual machine.
Assume having a variable Yi that defines a VM of type
i. For task j, we can compute the probability that task Tj is
related to type Yi, using a Bayes classifier, using equation 4
[14].

E. Model Description
Fig. 4 illustrates a flow chart for the proposed protocol.
Figure 4: A Detailed Task Scheduling Based on a Three Stages Strategy
IV. SIMULATION AND EXPERIMENTS
A. Simulation Environment
We applied our experiments using Lenovo laptop with a
processor intel(R), Core (TM) i3-3110 M CPU, 2.40 GHZ,
memory 4.0 GB, HD graphics (4000 G), windows 10
operating system, NetBeans IDE 8.2, JDK 8.0, Microsoft
SQL server 2012 and cloudSim plus simulator which is a
ISSN 1947-5500

Vol. 18, No. 6 June 2020
well-known and the most used cloud simulator to simulate
the cloud by inheriting and extending some of the cloudSim
classes such as Vm, Data center Broker, Cloudlet and Host
and defined the mapping policies as in the
CloudletSchedulerSpaceShared policy extended from the
CloudletSchedulerAbstract.
We applied the model simulation with a data set
from"http://www.mediafire.com/file/birsbbpo7e8nwom/L
CG._ARCH.swf/file" with a total size 10000. In the first
simulation we used a data set with 1000 record, we divided
the experimental data into five groups each group has a size
200. We started the simulation with a set from 0 to 200, then
added the second set from 200 to 400 to work on a set of
400, then 600,800 to 1000. Each time we add a set of 200 to
the last set until the end we work on f the 1000 set. The
simulation results may be different in values at each time we
simulate based on the dynamic data set each time, although
the result analysis shows a good enhancement in several
measurement factors in all the simulation trials.
B. Experimental Results and Analysis
Based on the above experimental configuration, and
analysing the simulation results for the proposed algorithm,
we compared our algorithm with the last proposed two-
stage scheduling algorithm [14], and the standard Min-Min
algorithm [21] based on some evaluation metrics like:
1. The Time complexity:
The time complexity for the Min–Min algorithm is
calculated as O (n3
) [14] [21] and for the Two- stage
algorithm it is calculated as O (n3
) [14]. After
simulating the algorithm and the theoretical
computation of time complexity to the overall
proposed algorithm, it is also O (n3
) which is the
same as in the two compared algorithms. So, the
three algorithms have equal effect with respect to
the time complexity.
2. The average makespan:
It is defined as the total time needed for tasks to be
scheduled and completed in the cloud, taking into
consideration that the smaller the makespan value,
the better the scheduling and the better the quality of
the service are. It is calculated as in equation 5 [14].
Where is the completion time of task Tj in the
cloud, and M is the total number of tasks
By calculating the average makespan and
according to the simulation results, we found that
the total time taken for tasks to complete the
scheduling process by the proposed three stages
algorithm is 1.88 MS while the total time taken
using the two- stages algorithm is 0.48 MS and the
min-min takes a total time of 2.90. This means that
the total time for tasks to be scheduled on the cloud
using the proposed three stages scheduling
algorithm is less than the total time we can got using
the two other algorithms, which means that the
proposed algorithm minimized the average
makespan value by about 74.5% compared to the
two stages and by about 74.7% compared to the
Min-Min algorithm working on a set of 200 task, by
the same way the algorithm minimized the
makespan value by about 70.7% compared to the
two stages, and by 88% compared with the Min-Min
algorithm using a set of 1000 task.
Fig. 5 shows the results of the simulation for the
three algorithms with respect to the average
makespan working on a set of 200,400,600…..1000
tasks:
Figure 5: The Average Makespan vs Number of Tasks
By analysing the data set that contains 1000
sample, we found it divided into 12 types with a
varying results from type 1 to type 12, the final
results as shown in fig. 6 is that the proposed
algorithm minimized the total makespan value by
71% compared to the two stages algorithm, and
by 88% compared to the Min-Min algorithm
which is a good enhancement ratio.
ISSN 1947-5500

Vol. 18, No. 6 June 2020
Figure 6: The Average Makespan vs Task Types
3. The task Average Waiting Time:
It is defined as the performance of the overall
processing capacity, and the throughput of the cloud,
taking into consideration that the smaller the waiting
time value, the better the scheduling and the better
the quality of the service are. It is calculated as in
equation 6 [14].
Where the waiting time of task Tj, M is the
total number of tasks.
By calculating the average waiting time and
according to the simulation results as shown, we
found that using the proposed algorithm; the tasks
take an average waiting time of 0.07 MS, while using
the two stages, tasks have a waiting time of 0.09 MS
and using the min-min algorithm the tasks have to
wait about 0.24 in average, this refers that the
proposed algorithm minimized the average waiting
time taken by tasks inside the scheduling process,
which means that the proposed algorithm minimized
the average waiting time value by about 22%
compared to the two stages by about 71% compared
to the Min-Min algorithm working on a set of 200
task, by the same way the algorithm enhanced the
waiting time value by about 70% compared to the
two stages, and by 88% compared with the Min-Min
algorithm using a set of 1000 task.
three algorithms with respect to the average waiting
time working on a set of 200,400,600…..1000 tasks:
Figure 7: The Average Waiting Time vs Number of Tasks
algorithm minimized the total average waiting
time by 70% compared to the two stages, and by
88% compared with the Min-Min algorithm, this
results indicates a good enhancement also for the
VMs load balancing; which results from using
the most similar with the minimum completion
time VM, in case there is no available matching
VM from the same type.
Figure 8: Average Waiting Time vs Task Types
4. The VMs Utilization Rate:
It is defined as a ratio that shows how effective the
scheduling is, according to the resource usage and
resource deployment, taking into consideration that
the bigger the utilization ratio value, the better the
scheduling and the better the quality of the service
are. It is calculated as in equation 7.

ISSN 1947-5500

Vol. 18, No. 6 June 2020
Where is the number of successful tasks
By calculating the utilization rate, and according to
the simulation results, we found that the proposed
algorithm has a utilization rate with 86%, while the
two- stages has a utilization rate with 81%, and the
Min-Min has a ratio of 80%, which means that the
proposed algorithm managed the available cloud
resources well, find the best deployment method for
resources and the tasks have got the most appropriate
Vm to be scheduled with, and maximized the
utilization rate value by about 6.2% compared to the
two stages and by about 7.5% compared to the Min-
Min algorithm working on a set of 200 task, by the
same way the algorithm maximized the utilization
ratio by about 10% compared to the two stages, and
by 42.6% compared with the Min-Min algorithm
using a set of 1000 task.
three algorithms with respect to the utilization ratio
working on a set of 200,400,600…..1000 tasks:
Figure 9: The Utilization Rate vs Number of Tasks
algorithm minimized the total average utilization by
70% compared to the two stages, and by 88%
compared with the Min-Min algorithm, The
enhancement in the resource utilization refers to the
good deployment for resources.
Figure 10: The Utilization Rate vs Task types
5. The Task Scheduling Failure Rate:
The Ratio between the number of failed tasks and
the total number of tasks, it measures the cloud's
stability.it is calculated as in equation 8 [15].
Where FT is the number of tasks that have a
scheduling failure, and M is the total number of
tasks.
By calculating the failure rate and according to the
simulation results, we found that the proposed
algorithm has 3.5%, failed tasks among all, while the
two stages has got a minimum failure ratio which is
5.5 %, and the Min-Min has a ratio of 3.5 % failed
tasks, which means that the proposed algorithm
reduced the failure rate by about 36.4% compared to
the two- stages and by 0% compared to the Min-Min
algorithm working on a set of 200 task. By the same
way, the algorithm reduced the failure rate by 0%
compared to the two- stages, and by 17% compared
with the Min-Min algorithm using a set of 1000 task.
three algorithms with respect to the failure rate
ISSN 1947-5500

Vol. 18, No. 6 June 2020
Figure 11: The Failure Rate vs Number of Tasks
algorithm reduced the failure rate by about 0%
compared to the two- stages, and by 17% compared
with the Min-Min. as per the graph results the failure
rate ratio varies between types 1 to 12 from low to
high according to the tasks complexity. Sometimes
the failure rate for the proposed algorithm is a little
greater than or equal to other algorithms. We may
handle this point later.
Figure 12: The Failure Rate vs Number of Tasks
6. The task guarantee ratio:
It defines the ratio of the tasks that are
successfully matched to a convenient VM to the
total number of tasks. We first define a variable
that takes a value 1, only when a task Tj is assigned
to a virtual machine Vi at Host Hk as represented in
equation 9 [14]:

The task guarantee ratio (Gr) value of a given host
Hk at its VMs as in equation 10 [14]:

By calculating the guarantee ratio and according to
the simulation results, we found that the proposed
algorithm has guarantee ratio of 99.965%, while the
two stages has got a value 99.945%, and the Min-Min
has a value of 99.965%, which means that the
proposed algorithm reduced the failure rate by about
0.02% compared to the two- stages and by about 0%
compared to the Min-Min algorithm working on a set
of 200 task. By the same way, the algorithm has the
same guarantee ratio compared to the two- stages,
and reduced the value by 0.005% compared with the
Min-Min algorithm using a set of 1000 task.
three algorithms with respect to the guarantee ratio

Figure 13: The Guarantee Ratio vs Task Types
By analysing the data set that contains 1000 sample,
we found it divided into 12 types with a varying
results from type 1 to type 12, the final results as
shown in figure 14 shows that the guarantee ratio
values vary between types 1 to 12 from higher values
to low according to the task type complexity.
ISSN 1947-5500

Vol. 18, No. 6 June 2020
Figure 14: Guarantee Ratio vs Number of Tasks
CONCLUSION AND FUTURE WORK
In this paper, we suggest a dynamic three stages
scheduling algorithm; which is a three-stage task
scheduling technique that is used to find the best
matched pairs of tasks and VMs, achieve the best
results of task scheduling and execution, enhance the
quality of the service on the cloud based on the latest
stored scheduling data for tasks and their matched
VMs, which helps to pre-create a convenient number
of VMs with different resource attributes, saving
much time and resources and enhancing the quality of
service. The results of the algorithm simulation
compared to other algorithms shows that the
proposed algorithm has got the minimum waiting
time, minimum makespan, and maximum utilization
rate, enhanced the load balance on the virtual
machines and has got a little enhancement in the task
guarantee ratio. In spite of the better enhancement for
the previous factors, the algorithm sometimes has a
failure rate value which is a little higher or equal to
other algorithms, based on the dynamic dataset. So
that, In the future, we need to minimize the failure
rate also we can use the concept of machine learning
to get the system act as a self-learning while working
on any type of data need to be scheduled.
REFERENCES
[1] Panda, S. K., & Jana, P. K. "Efficient task scheduling
algorithms for heterogeneous multi-cloud environment". The
Journal of Supercomputing, Vol.71, No.4, pp.1505–1533,
2015.
[2] SaeMi Shin, Yena Kim, and SuKyoung Lee." Deadline-
guaranteed scheduling algorithm with improved resource
utilization for cloud computing". 2015 12th Annual IEEE
Consumer Communications and Networking Conference
(CCNC), 2015.
[3] Razaque, Abdul, et al. "Task scheduling in cloud
computing." 2016 IEEE Long Island Systems, Applications
and Technology Conference (LISAT). IEEE, pp.1-5, 2016.
[4] Lakra, A. V., & Yadav, D. K "Multi-Objective Tasks
Scheduling Algorithm for Cloud Computing Throughput
Optimization". Procedia Computer Science. Vol.48, pp.107–
113. 2015.
[5] Arunarani, A. R., D. Manjula, and Vijayan Sugumaran.
"Task scheduling techniques in cloud computing: A
literature survey." Future Generation Computer Systems,
Vol.91, pp.407-415, 2019.
[6] IM.Vijayalakshmi, IIV.Venkatesa Kumar" Investigations on
Job Scheduling Algorithms in Cloud Computing",
International Journal of Advanced Research in Computer
Science & Technology (IJARCST) Vol.2, No.1, 2014.
[7] Babur Hayat Malik, Mehwashma Amir, Bilal Mazhar,
Shehzad Ali, Rabiya Jalil, Javaria Khalid. "Comparison of
Task Scheduling Algorithms in Cloud Environment".
(IJACSA) International Journal of Advanced Computer
Science and Applications. Vol.9, No.5, 2018.
[8] Jain, N., Menache, I., Naor, J. (Seffi), & Yaniv, J. "Near-
Optimal Scheduling Mechanisms for Deadline-Sensitive
Jobs in Large Computing Clusters". ACM Transactions on
Parallel Computing, Vol.2, No.1, pp.1–29, 2015.
[9] Raza Abbas Haidri C. P. Katti2, P. C. Saxena3. "A Load
Balancing Strategy for Cloud Computing", IEEE
International Conference on Signal Propagation and
Computer Technology, 2014.
[10] Elrotub, M., & Gherbi, A., "Virtual machine classification-
based approach to enhanced workload balancing for cloud
computing applications", Procedia computer
science, Vol.130, pp.683-688, 2018.
[11] Delimitrou, C., & Kozyrakis, C. "QoS-Aware scheduling in
heterogeneous data centers with paragon". ACM
Transactions on Computer Systems, Vol.31, No.4, pp.1–34,
2013.
[12] Rodriguez, Maria Alejandra, and Rajkumar Buyya.
"Deadline based resource provisioningand scheduling
algorithm for scientific workflows on clouds." IEEE
transactions on cloud computing, Vol.2, No.2, pp.222-235,
2014.
[13] Zhou, Zhou, et al. "Minimizing SLA violation and power
consumption in Cloud data centers using adaptive energy-
aware algorithms." Future Generation Computer
Systems, Vol.86, pp.836-850, 2018.
[14] Zhang, P., & Zhou, M., "Dynamic Cloud Task Scheduling
Based on a Two-Stage Strategy", IEEE Transactions on
Automation Science and Engineering, Vol.15, No.2,
pp.772–783, 2018.
DOI:10.1109/tase.2017.2693688
[15] Alworafi, M. A., & Mallappa, S. "An Enhanced Task
Scheduling in Cloud Computing Based on Deadline-Aware
Model", International Journal of Grid and High-Performance
Computing, Vol.10, No.1, pp.31–53, 2018.
[16] Dr. Amit Agarwal, Saloni Jain."Efficient Optimal Algorithm
of Task Scheduling in Cloud Computing Environment".
International Journal of Computer Trends and Technology
(IJCTT), Vol.9, No.7, pp.344-349, 2014.
[17] Li, X., Qian, L., & Ruiz, R. "Cloud Workflow Scheduling
with Deadlines and Time Slot Availability". IEEE
Transactions on Services Computing, Vol.11, No.2, pp.329–
340, 2018.
[18] Mehwish Awan*, Munam Ali Shah. "A Survey on Task
Scheduling Algorithms in Cloud Computing Environment ".
International Journal of Computer and Information
Technology, Vol.4, No.2, 2015.
[19] Khanghahi, N., & Ravanmehr, R., "Cloud Computing
Performance Evaluation: Issues and Challenges",
International Journal on Cloud Computing: Services and
Architecture, Vol.3, No.5, pp.29–41, 2013.
[20] Zhang, Xinqian, et al. "Energy-aware virtual machine
allocation for cloud with resource reservation." Journal of
Systems and Software, Vol.147, pp.147-161, 2019.
[21] Devipriya, S., and C. Ramesh. "Improved Max-min heuristic
model for task scheduling in cloud." 2013 International
Conference on Green Computing, Communication and
Conservation of Energy (ICGCE). IEEE, 2014.
ISSN 1947-5500

IJCSIS
ISSN (online): 1947-5500
Please consider to contribute to and/or forward to the appropriate groups the following opportunity to submit and publish
original scientific results.
CALL FOR PAPERS
International Journal of Computer Science and Information Security (IJCSIS)
January-December 2020 Issues
The topics suggested by this issue can be discussed in term of concepts, surveys, state of the art, research,
standards, implementations, running experiments, applications, and industrial case studies. Authors are invited
to submit complete unpublished papers, which are not under review in any other conference or journal in the
following, but not limited to, topic areas.
See authors guide for manuscript preparation and submission guidelines.
Indexed by Google Scholar, DBLP, CiteSeerX, Directory for Open Access Journal (DOAJ), Bielefeld
Academic Search Engine (BASE), SCIRUS, Scopus Database, Cornell University Library, ScientificCommons,
ProQuest, EBSCO and more.
Deadline: see web site
Notification: see web site
Revision: see web site
Publication: see web site
For more topics, please see web site https://sites.google.com/site/ijcsis/
For more information, please visit the journal website (https://sites.google.com/site/ijcsis/)

Context-aware systems
Networking technologies
Security in network, systems, and applications
Evolutionary computation
Industrial systems
Evolutionary computation
Autonomic and autonomous systems
Bio-technologies
Knowledge data systems
Mobile and distance education
Intelligent techniques, logics and systems
Knowledge processing
Information technologies
Internet and web technologies, IoT
Digital information processing
Cognitive science and knowledge
Agent-based systems
Mobility and multimedia systems
Systems performance
Networking and telecommunications
Software development and deployment
Knowledge virtualization
Systems and networks on the chip
Knowledge for global defense
Information Systems [IS]
IPv6 Today - Technology and deployment
Modeling
Software Engineering
Optimization
Complexity
Natural Language Processing
Speech Synthesis
Data Mining

Dynamic Three Stages Task Scheduling Algorithm on Cloud Computing

Recommended

More Related Content

What's hot (19)

Similar to Dynamic Three Stages Task Scheduling Algorithm on Cloud Computing (20)

Recently uploaded (20)

Dynamic Three Stages Task Scheduling Algorithm on Cloud Computing