Optimal scheduling in cloud computing environment using the bee algorithm

Đăng ngày 4/2/2019 3:56:41 PM | Thể loại: | Lần tải: 0 | Lần xem: 7 | Page: 6 | FileSize: 0.33 M | File type: PDF
Optimal scheduling in cloud computing environment using the bee algorithm. In this article, we want to use honeybee colony algorithm for resources scheduling. This algorithm is an optimization method based on swarm intelligence and intelligent behavior of honeybee population. Honeybee algorithm involves a group based on search algorithm.
International Journal of Computer Networks and Communications Security
VOL. 3, NO. 6, JUNE 2015, 253–258
Available online at: www.ijcncs.org
E-ISSN 2308-9830 (Online) / ISSN 2410-0595 (Print)
Optimal Scheduling In Cloud Computing Environment Using the
Bee Algorithm
Msc.NASRIN HESABIAN1, 2 and phD.HAMID HAJ SEYYED JAVADI3
1 Department of Computer Engineering, Science and Research Branch, Islamic Azad University, Boroujerd,
Iran
2 Department of Computer Engineering, Boroujerd Branch, Islamic Azad University, Boroujerd, Iran
3 Department of Applied Mathematics, Faculty of Mathematics and Computer Science, Shahed University,
Tehran, Iran
E-mail: 1, 2n.hesabian@gmail.com, 3hamid.h.s.javadi@gmail.com
ABSTRACT
Cloud computing has made a fundamental change in the way of strange information and data and
implementation of application progress. Everything is hosted on a cloud that is a set of several servers and
computer, which can be accessed through internet instead of placing data and application programs on a
personal computer. The challenge of cloud computing system is dedicating the resources to the system
requests. Dedicating resources to the requests is a NP-complete problem due to requests and resource
dynamics. In recent year, one of the most important and promising method to solve such problems is
innovatin methods inspired from the nature. These methods are similar to the social or natural system. In
this article, we want to use honeybee colony algorithm for resources scheduling. This algorithm is an
optimization
method
based on swarm intelligence and intelligent behavior of honeybee population.
Honeybee algorithm involves a group based on search algorithm.
Keywords: Cloud Computing, Scheduling, Resource Dedication, Honeybee Algorithm.
1
INTRODUCTION
computing systems is dedicating the resource to the
system requests. Dedicating the resource to requests
Cloud
computing
is
a
general
and
newfound
technology
delivering
resources
and
application
is a
NP-complete
problem due
to
requests
and
program
to
the
users.
Multi-national
companies
resources
dynamics
[3],[4].
in
cloud
computing
such
as
Microsoft,
IBM,
Google
and
Oracel
systems,
computing
resource
are
presented
as
companies
provide
various
services
of
cloud
virtual
machines
.
In
such
scenario,
scheduling
computing to the clients. Currently, cloud com-
algorithms play a very important role because the
puting is a commercial field [1],[2]. In terms of
aim is scheduling the efficiency of tasks so that
cloud
computing,
companies
organizations
and
time is reduced, and resource exploration can be
individuals don’t pay money for software, hardware
improved.
In
recent
years,
one
of
the
most
or
the
network;
instead,
they
buy
the
required
important and promising methods to solve such
software services and computing power. This idea
problems is innovative methods inspired from the
follows high saving and productivity in information
nature. These methods are similar to social and
technology resources. In fact, cloud computing is a
natural
systems.
These
methods
are
genetic
pattern
of
distributed
computing
,
and
it
is
a
algorithm, ant-colony optimization algorithm and
combination of many resources and requests with
bee-colony optimization problem. In this article, we
the aim of resource subscription in the form of a
want
to
use
bee
colony
algorithm
in
resource
service
in
the
internet.
The
challenge
of
cloud
scheduling.
This
algorithm
is
an optimization
254
N. Hesabian and H. H. S. Javadi / International Journal of Computer Networks and Communications Security, 3 (6), June 2015
method based on swarm intelligence and intelligent
resource, and it does not stop [9],[10].
behavior of bee population. Since bee algorithm is a
* The properties of requests
that
home been
meta-innovative
method
based
on
distributed
entered to the environment are unclear, and they
systems and parallel processing technique, it is used
continuously
change
such
as
entering
time,
for scheduling optimization [15].
executing time and required memory.
*
Cloud
computing
environment
is
a
set
of
2
CLOUD COMPUTING
heterogynous resources.
* Resources home some hardware and software
Distributed
systems
and
parallel
processing
properties such as load volume, the amount of free
techniques are the solution for better and quicker
memory
in
a
system,
it
has
properties
of
utilization of complex information in the present.
communication network such as bandwidth, traffic
Nowadays, there are hundreds of computers and
and etc. Kanlee and his colleagues (2011) proposed
supercomputers with various architecture capacities
scheduling of cloud tasks on the basis
of load
in all over the world, and they are used in scientific,
balancing
ant
colony
optimization
algorithm
military and commercial fields. Mostly, information
(LBACO)[17]. The main policy of this method is
subscription
among
them
is
necessary.
A
load balance of whole system. Also, it has been
distributed system as a set of independent comp-
attempted to minimize the time of Make span. In
uters and users consider it as a coherent system,
this article, the properties of ant algorithm has been
grid computing is a distributed system and it is
considered to schedule the tasks. Sourav Banerjee
infrastructure of cloud computing. This technology
and
his
colleagues
(2002)
presented
scheduling
provides
accessing
remote
resources
by
using
algorithm based on genetic algorithm[11].
computer networks and communication infrastruc-
Kennedy and his colleagues (2001) proposed an
ture as well as by using concepts and facilities of
algorithm that has been inspired from the behavior
distributed
systems.
Computing
resource
of
of birds and fish , and it is
in unline discrete routs
heterogeneous software and hardware systems can
used
in
ants
colony
algorithm.
This
algorithm
be connected to each other without geographical
called
particles
swarm
algorithm
searches
the
limitations, and whole system structure seems as a
solution space by adjusting various factors.
unique and integrated virtual machine. Then, very
Babu and his colleagues (2003) used bee meta-
large and complex application, program, requiring
heuristic
algorithm
to
reach
load
balance
in
high
processing
and
much
input
data
can
be
machines. The proposed algorithm provides tasks
implemented in this virtual machine. In fact, the
priority in machines queue on the basis of the least
purpose is to use computing resources of systems to
waiting
time.
They
found
considerable
perform tasks when they are free [5],[6].
improvement
in
execution
time
and
reducing
Cloud computing has grid properties as well as
waiting time in queue [12][13][14][16].
under properties. In fact, cloud computing as a
pattern
of
distributed
computing,
and
it
is
4
THE PROPOSED ALGORITHM
combination of many resources and request with
the aim of resource subscription in the internet [7,
As we know, various users send various tasks to
8].
the system, so different computing capacities are
required. Dedicating the capacity, similar to cloud
3
LITERATURE REVIEW
resources,
to
various
users
is
undesirable.
In
addition, it is assumed that cloud resources are
Since
tasks
scheduling
is
dynamic
in
cloud
heterogeneous, independent and dynamic. Hence,
computing environment, it is difficult to obtain a
resources properties are temporal and they change.
proper solution that can minimize tasks impleme-
Cloud performance depends on scheduling method
ntation, the time of executing input tasks and can
because cloud environment is dynamic, independe-
consider load balance between resources.
nt and heterogeneous, and the space of tasks is
The number of request that has been entered to
complicated.
the system is more and higher than the number of
255
N. Hesabian and H. H. S. Javadi / International Journal of Computer Networks and Communications Security, 3 (6), June 2015
bees, determining the amount of nectar, determin-
ing scout bees
and then
sending
them to
food
sources. In the stage of initial initialization, a set of
food sources are randomly selected by the bees and
the amount of their nectar is detected. Then, these
bees come to the hive, and data and informa-tion of
nectar in each hive is subscribed with the bees
waiting in dancing area. In the second stage, after
information subscription, each employed bee goes
to the range of food source this bee has already
visited this range, and has stored that food source in
its
own
memory.
Then,
a
new
food
source
is
selected by using visual information existing in the
neighborhood. In the third stage, a supervisory bee
select
the
food
source
depending
on
nectar
information and data distributed by the employed
bees in dancing area. The nectar amount of food
source increases, and the food source selected by
Fig. 1. the assumptions of cloud system for proposed
algorithm.
supervisory bee increases. Therefore, dancing
employed bees carring much nectar encourage
For example, as it has been shown in figure1, the
cloud system involves scheduling of S and three
cloud nodes of VM={VM1,VM2,VM3}. In cloud
system, we consider S= {VM1, VM2, VM3}. As it
has been already said, the number of cloud node is
equal to the number of schedulers. Bees involve
three groups in bee colony algorithm. They involve
employed bees, supervision. They involve
employed bees, supervisory bees and scout bees.
The bee staying in dancing area to make decision
and select the food source is called explorer bee.
The bee going toward predetermined food source is
called employed bee, while the bee performing
random search is called scout bee. The main steps
of algorithms are as follow:
supervisory bees toward food source with much
nectar. The bee selects a new food source in the
neighborhood depending on visual information
after entering to the selected area. Visual
information is on the basis of comparing food
source. When the nectar of food source is releasd
by the bees, a new food source is randomly
determined by the scout bee. The algorithm is
performed by a set of solutions, and it‘s attempted
to improve it called population.
In this article, some bees have been considered as
scout bees, and some have been taken into account
as supervisory bees, and some of them are
considered as employed bees. After returning to the
hive and after dancing other bees, they go to the
neighborhood of the places where other bees have
Initial initialization
Repetition
moved through observing their dance. Two
methods have been used to simulate it. A random
number is created. If this random number is lesser
than threshold limit, then, the first node occurs. If it
A.
B.
The location of employed bees in food
sources in the memory
The location of search bees in food
is more than threshold limit, the second mode
occurs. It should be mentioned that the set of tasks
and virtual machines are defined as follows:
sources in the memory
VM= {VM1, VM2,...., VM3}
C.
Sending scout bees to search new food
Task= {T1, T2,...Tn}
sources
Each bee is the answer of a problem, so the bee is
shown as follows:
(The considered situation is obtained).
Bee=[X11, X21... Xij]
In ABC algorithm, each search cycle is
composed of three stages.
These stages are sending employed bees toward
food sources and then measuring the amount of
their nectar, selecting food sources they explorer
Xij shows the ith task processing in jth processor.
For instance, X12 shows that the first task is
processed in the second processor. Therefore, in
order to simplify it, the bee has been shown as
follows:
bees after subscribing information they employed
B=2 3 1 2 1 3
HƯỚNG DẪN DOWNLOAD TÀI LIỆU

Bước 1:Tại trang tài liệu slideshare.vn bạn muốn tải, click vào nút Download màu xanh lá cây ở phía trên.
Bước 2: Tại liên kết tải về, bạn chọn liên kết để tải File về máy tính. Tại đây sẽ có lựa chọn tải File được lưu trên slideshare.vn
Bước 3: Một thông báo xuất hiện ở phía cuối trình duyệt, hỏi bạn muốn lưu . - Nếu click vào Save, file sẽ được lưu về máy (Quá trình tải file nhanh hay chậm phụ thuộc vào đường truyền internet, dung lượng file bạn muốn tải)
Có nhiều phần mềm hỗ trợ việc download file về máy tính với tốc độ tải file nhanh như: Internet Download Manager (IDM), Free Download Manager, ... Tùy vào sở thích của từng người mà người dùng chọn lựa phần mềm hỗ trợ download cho máy tính của mình  
7 lần xem

Optimal scheduling in cloud computing environment using the bee algorithm. In this article, we want to use honeybee colony algorithm for resources scheduling. This algorithm is an optimization method based on swarm intelligence and intelligent behavior of honeybee population. Honeybee algorithm involves a group based on search algorithm..

Nội dung

International Journal of Computer Networks and Communications Security VOL. 3, NO. 6, JUNE 2015, 253–258 Available online at: www.ijcncs.org E-ISSN 2308-9830 (Online) / ISSN 2410-0595 (Print) Optimal Scheduling In Cloud Computing Environment Using the Bee Algorithm Msc.NASRIN HESABIAN1, 2 and phD.HAMID HAJ SEYYED JAVADI3 1 Department of Computer Engineering, Science and Research Branch, Islamic Azad University, Boroujerd, Iran 2 Department of Computer Engineering, Boroujerd Branch, Islamic Azad University, Boroujerd, Iran 3 Department of Applied Mathematics, Faculty of Mathematics and Computer Science, Shahed University, Tehran, Iran E-mail: 1, 2n.hesabian@gmail.com, 3hamid.h.s.javadi@gmail.com ABSTRACT Cloud computing has made a fundamental change in the way of strange information and data and implementation of application progress. Everything is hosted on a cloud that is a set of several servers and computer, which can be accessed through internet instead of placing data and application programs on a personal computer. The challenge of cloud computing system is dedicating the resources to the system requests. Dedicating resources to the requests is a NP-complete problem due to requests and resource dynamics. In recent year, one of the most important and promising method to solve such problems is innovatin methods inspired from the nature. These methods are similar to the social or natural system. In this article, we want to use honeybee colony algorithm for resources scheduling. This algorithm is an optimization method based on swarm intelligence and intelligent behavior of honeybee population. Honeybee algorithm involves a group based on search algorithm. Keywords: Cloud Computing, Scheduling, Resource Dedication, Honeybee Algorithm. 1 INTRODUCTION Cloud computing is a general and newfound technology delivering resources and application program to the users. Multi-national companies such as Microsoft, IBM, Google and Oracel companies provide various services of cloud computing to the clients. Currently, cloud com-puting is a commercial field [1],[2]. In terms of cloud computing, companies organizations and individuals don’t pay money for software, hardware or the network; instead, they buy the required software services and computing power. This idea follows high saving and productivity in information technology resources. In fact, cloud computing is a pattern of distributed computing , and it is a combination of many resources and requests with the aim of resource subscription in the form of a service in the internet. The challenge of cloud computing systems is dedicating the resource to the system requests. Dedicating the resource to requests is a NP-complete problem due to requests and resources dynamics [3],[4]. in cloud computing systems, computing resource are presented as virtual machines . In such scenario, scheduling algorithms play a very important role because the aim is scheduling the efficiency of tasks so that time is reduced, and resource exploration can be improved. In recent years, one of the most important and promising methods to solve such problems is innovative methods inspired from the nature. These methods are similar to social and natural systems. These methods are genetic algorithm, ant-colony optimization algorithm and bee-colony optimization problem. In this article, we want to use bee colony algorithm in resource scheduling. This algorithm is an optimization 254 N. Hesabian and H. H. S. Javadi / International Journal of Computer Networks and Communications Security, 3 (6), June 2015 method based on swarm intelligence and intelligent behavior of bee population. Since bee algorithm is a meta-innovative method based on distributed systems and parallel processing technique, it is used for scheduling optimization [15]. 2 CLOUD COMPUTING Distributed systems and parallel processing techniques are the solution for better and quicker utilization of complex information in the present. Nowadays, there are hundreds of computers and supercomputers with various architecture capacities in all over the world, and they are used in scientific, military and commercial fields. Mostly, information subscription among them is necessary. A distributed system as a set of independent comp-uters and users consider it as a coherent system, grid computing is a distributed system and it is infrastructure of cloud computing. This technology provides accessing remote resources by using computer networks and communication infrastruc-ture as well as by using concepts and facilities of distributed systems. Computing resource of heterogeneous software and hardware systems can be connected to each other without geographical limitations, and whole system structure seems as a unique and integrated virtual machine. Then, very large and complex application, program, requiring high processing and much input data can be implemented in this virtual machine. In fact, the purpose is to use computing resources of systems to perform tasks when they are free [5],[6]. Cloud computing has grid properties as well as under properties. In fact, cloud computing as a pattern of distributed computing, and it is combination of many resources and request with the aim of resource subscription in the internet [7, 8]. 3 LITERATURE REVIEW Since tasks scheduling is dynamic in cloud computing environment, it is difficult to obtain a proper solution that can minimize tasks impleme-ntation, the time of executing input tasks and can consider load balance between resources. The number of request that has been entered to the system is more and higher than the number of resource, and it does not stop [9],[10]. * The properties of requests that home been entered to the environment are unclear, and they continuously change such as entering time, executing time and required memory. * Cloud computing environment is a set of heterogynous resources. * Resources home some hardware and software properties such as load volume, the amount of free memory in a system, it has properties of communication network such as bandwidth, traffic and etc. Kanlee and his colleagues (2011) proposed scheduling of cloud tasks on the basis of load balancing ant colony optimization algorithm (LBACO)[17]. The main policy of this method is load balance of whole system. Also, it has been attempted to minimize the time of Make span. In this article, the properties of ant algorithm has been considered to schedule the tasks. Sourav Banerjee and his colleagues (2002) presented scheduling algorithm based on genetic algorithm[11]. Kennedy and his colleagues (2001) proposed an algorithm that has been inspired from the behavior of birds and fish , and it is in unline discrete routs used in ants colony algorithm. This algorithm called particles swarm algorithm searches the solution space by adjusting various factors. Babu and his colleagues (2003) used bee meta-heuristic algorithm to reach load balance in machines. The proposed algorithm provides tasks priority in machines queue on the basis of the least waiting time. They found considerable improvement in execution time and reducing waiting time in queue [12][13][14][16]. 4 THE PROPOSED ALGORITHM As we know, various users send various tasks to the system, so different computing capacities are required. Dedicating the capacity, similar to cloud resources, to various users is undesirable. In addition, it is assumed that cloud resources are heterogeneous, independent and dynamic. Hence, resources properties are temporal and they change. Cloud performance depends on scheduling method because cloud environment is dynamic, independe-nt and heterogeneous, and the space of tasks is complicated. 255 N. Hesabian and H. H. S. Javadi / International Journal of Computer Networks and Communications Security, 3 (6), June 2015 Fig. 1. the assumptions of cloud system for proposed algorithm. For example, as it has been shown in figure1, the cloud system involves scheduling of S and three cloud nodes of VM={VM1,VM2,VM3}. In cloud system, we consider S= {VM1, VM2, VM3}. As it has been already said, the number of cloud node is equal to the number of schedulers. Bees involve three groups in bee colony algorithm. They involve employed bees, supervision. They involve employed bees, supervisory bees and scout bees. The bee staying in dancing area to make decision and select the food source is called explorer bee. The bee going toward predetermined food source is called employed bee, while the bee performing random search is called scout bee. The main steps of algorithms are as follow:  Initial initialization  Repetition A. The location of employed bees in food sources in the memory B. The location of search bees in food bees, determining the amount of nectar, determin-ing scout bees and then sending them to food sources. In the stage of initial initialization, a set of food sources are randomly selected by the bees and the amount of their nectar is detected. Then, these bees come to the hive, and data and informa-tion of nectar in each hive is subscribed with the bees waiting in dancing area. In the second stage, after information subscription, each employed bee goes to the range of food source this bee has already visited this range, and has stored that food source in its own memory. Then, a new food source is selected by using visual information existing in the neighborhood. In the third stage, a supervisory bee select the food source depending on nectar information and data distributed by the employed bees in dancing area. The nectar amount of food source increases, and the food source selected by supervisory bee increases. Therefore, dancing employed bees carring much nectar encourage supervisory bees toward food source with much nectar. The bee selects a new food source in the neighborhood depending on visual information after entering to the selected area. Visual information is on the basis of comparing food source. When the nectar of food source is releasd by the bees, a new food source is randomly determined by the scout bee. The algorithm is performed by a set of solutions, and it‘s attempted to improve it called population. In this article, some bees have been considered as scout bees, and some have been taken into account as supervisory bees, and some of them are considered as employed bees. After returning to the hive and after dancing other bees, they go to the neighborhood of the places where other bees have moved through observing their dance. Two methods have been used to simulate it. A random number is created. If this random number is lesser than threshold limit, then, the first node occurs. If it is more than threshold limit, the second mode occurs. It should be mentioned that the set of tasks and virtual machines are defined as follows: sources in the memory C. Sending scout bees to search new food sources  (The considered situation is obtained). VM= {VM1, VM2,...., VM3} Task= {T1, T2,...Tn} Each bee is the answer of a problem, so the bee is shown as follows: Bee=[X11, X21... Xij] In ABC algorithm, each search cycle is composed of three stages. These stages are sending employed bees toward food sources and then measuring the amount of their nectar, selecting food sources they explorer bees after subscribing information they employed Xij shows the ith task processing in jth processor. For instance, X12 shows that the first task is processed in the second processor. Therefore, in order to simplify it, the bee has been shown as follows: B=2 3 1 2 1 3 256 N. Hesabian and H. H. S. Javadi / International Journal of Computer Networks and Communications Security, 3 (6), June 2015 This bee shows that the first task is dedicated to the second processor, and the second task is dedicated to the third processor and so on. After creating initial bees or population, the fitness of each bee is computed by using fitness function. In order to compute the fitness, two parameters involving loading balance and makespan of the last task are simultaneously used as follows: K1: impact factor of makespan of the last task K2: impact factor of loading balance F ([Bi ])= k1 Makespan([Bi ])+k2 Load.Balancing([Bi ])% 5 EVALUATION AND COMPARISON In order to compare the proposed method with other methods, three parameter involving makespan, flow time and the average of waiting time. In this article, we want to compare the results obtained from our proposed method (PM) with the results obtained from the proposed algorithm (dhinesh BABu, 2013)[16]. Using honey meta-heuristic algorithm to reach the load balance over the machines, it is called HBB-LB. F([Bi]) shows fitness of bee i. [Bi] demonstrates the values of bee Bi. With regard to above definition, the smaller the amount of F, the higher the fitness will be. After computing the fitness of each bee, some of them are selected as scout bee. Other bees move to the neighborhood of the places where these bees have gone. In order to compute the neighborhood location, the method of defining threshold limit is used the third stages. Some bees are randomly sent to the environment again, so that placing them in the local optimally is prevented. After creating all bees, bee’s population is updated. Generally, the stages are as follows: 1. Creating the initial population 2. Computing the fitness function 3. Ordering the population based on fitness function. 4. Defining the threshold limit  If the random number is lesser than threshold Fig. 2. changes of fitness function with the number of replication (1000 tasks-64 virtual machines) limit, then some locations are randomly selected. Random numbers are created by a uniform distribution between [-1, 1]. The selected locations created by random numbers are summed up.  If the random number is lesser than threshold limit, nodes information is extracted from coordinator. A specified percentage of nodes having higher volume of processing is Fig. 3. comparing the flow time in the proposed method determined. A specified percentage of the nodes having the least volume of processing is determined. The load of nodes in moved. 5. Computing fitness function 6. Selecting the best bees 7. Producing the new initial population Table 1 shows the changes of makespan values before and after executing the proposed algorithm for 1000 tasks and 64 resource, as it has been shown, by applying the algorithm , the results obtained from makespan reduces from 54 seconds to 18 seconds; that is, the reduction of 66% is observed. 8. If the number of replications is enough, it ends; otherwise, return to stage 3. 257 N. Hesabian and H. H. S. Javadi / International Journal of Computer Networks and Communications Security, 3 (6), June 2015 Table1: comparing various parameters after and before applying the proposed method. algorithms such as HBB-LB. the proposed method or PM method behaves like HBB-LB method in small environment in terms of makespan. Its performance becomes better by increasing the environment dimensions. It can be due to the nodes awareness of each other’s position. Their task is to cordinate these tasks. 7 REFERENCES Fig. 4. comparing the average of waiting time in the proposed method. Figure 4 demonstrate the average of waiting time in the method proposed in the method in comparison to HBB-LB method .Like makespan , responding time dose not reduce by increasing cloud environment scale. HBB-LB method not only bees weaker performance than PM but also its performance reduce by increasing the network scale. Response time is one of the parameters, showing the service quality in the network. Therefore QOS of the proposed method is better than HBB-LB method. 6 CONCLUSION As we know, various users send various tasks to the system, so different computing capacity is required. Dedicating capacity similar to cloud resources to various users is undesirable. In addition, it’s assumed that cloud resources are heterogeneous, independent and dynamic; therefore, the propertice of cloud resources are temporary, and they change. Cloud performance depends on scheduling method since cloud environment is dynamic, independent and heterogeneous and the space of tasks is complex. In this article, we presented a scheduling method in cloud computing environment on the basis of bee algorithm to dedicate the sources optimally. The proposed method has been compared with other [1] Arkaitz ruiz-Alvarez, Marty Humphery, A Model and Decision Prosedure for Date Storage in Cloud Computing, in Proceedings of the IEEE/ACM International Symposium on Cluster, Ottawa Canada, 2012. [2] S.Selvarani, Dr.G.Sudha Sadhasivam , “IMPROVED COST-BASED ALGORITHM FOR TASK SCHEDULING IN CLOUD COMPUTING” in IEEE, India 2010 . [3] Bhaskar Prasad Rimal, Eunmi Choi, Ian Lamb, " A Taxonomy and Survey of Cloud Computing Systems" in IEEE, Fifth International [4] B. Basturk ; D. Karaboga, "An artificial bee colony(abc) algorithm for numeric function optimization," in IEEE Swarm Intelligence Symposium 2006, Indianapolis,Indiana, USA, May 2006. [5] D. Karaboga , B. Basturk; “ On the performance of artificial bee colony (ABC) algorithm” in Elsevier Erciyes University, Engineering Faculty, Computer Engineering Department, TR-38039 Kayseri, Turkey 2008. [6] Rajkumar Buyya ,Chee Shin Yeo, Srikumar Venugopal, James Broberg, Ivona Brandic, “Cloud com puting and emerging IT platforms: vision, hype, and reality for delivering computing as the 5t h utility “,Future Generation Computer systems, ELSEVIER (2008). [7] Steve Bennett, Mans Bhuller, Robert Covington ,” Architectural Strategies for Cloud computing “ , August (2009) . [8] Brucker P., Drexl A., Mohring R., Neumann K., Pesch E .,“Resource-constrained roject scheduling: Notation, classification, models, and methods”, Elsevier European Journal of Operational Research, Vol. 112, pp. 3-41, 1999 [9] Malathi G, Sarumathi S. 2010. Survey on grid scheduling. [10] Alcaraz J., Maroto C.,“A Robust Genetic Algorithm for Resource Allocation in Project Scheduling”, Annals of Operations Research, Vol. 102, pp. 83-109, 2001 [11] Brucker P., Knust S., Schoo A., Thiele O.,“A branch and bound algorithm for the resource- 258 N. Hesabian and H. H. S. Javadi / International Journal of Computer Networks and Communications Security, 3 (6), June 2015 constrained project scheduling problem”, Elsevier European Journal of Operational Research, Vol. 107, pp. 272-288, 1998 [12]Delavar A.G., Rahmany M., Halaakouie A., Sookhtsaraei ,R “DSQGG: An optimized Genetic-based algorithm for Scheduling in Distributed Grid “,2nd International Conference , Computer Technology and Development (ICCTD), (2010). [13]Pham D.T., Ghanbarzadeh A., Koc E., Otri S., RahimS.,M.Zaidi, "The Bees lgorithm - A Novel Toolfor Complex Optimisation Problems," Proceedings of IPROMS 2006 Conference, pp. 454-461, 2006. [14]D. Karaboga ; B. Basturk, "A powerful and efficient algorithm for numerical function optimizationArtificial bee colony (abc) algorithm," Journal of Global Optimization, p. 459–471, 2007. [15]Sourav Banerjee ,Mainak Adhikari ,Utpal Biswas “Advanced Task Scheduling for Cloud Service Provider Using Genetic Algorithm” IOSR Journal of Engineering (IOSRJEN) ISSN: 2250-3021 Volume 2, Issue 7(July 2012), PP 141-147 [16]Dhinesh Babu L.D, P. Venkata Krishna “Honey bee behavior inspired load balancing of tasks in cloud computing environments” journal home page: www.elsevier.com/l ocate/asoc (Applied Soft Computing 13 (2013) 2292–2303) , 2013 Elsevier . [17]Kun Li, Gaochao Xu, Guangyu Zhao, Yushuang Dong, Dan Wang “Cloud Task scheduling based on Load Balancing Ant Colony Optimization” 2011 Sixth Annual ChinaGrid Conference, 978-0-7695-4472-4/11 $26.00 © 2011 IEEE.

Tài liệu liên quan