Optimum Resource Allocation using Specification Matching and Priority Based Method in Cloud

International Journal on Recent and Innovation Trends in Computing and Communication ISSN: 2321-8169
Volume: 5 Issue: 7 77 – 84
_______________________________________________________________________________________________
77
IJRITCC | July 2017, Available @ http://www.ijritcc.org
______________________________________________________________________________________
Optimum Resource Allocation using Specification Matching and Priority Based
Method in Cloud
Akhil Chaurasia
Department of Computer Science and Engineering
Sant Longowal Institute of Engineering and Technology
Deemed University, Longowal, Sangrur, India
e-mail:akhilchaurasia47@gmail.com
Jaspal Singh
Department of Computer Science and Engineering
Sant Longowal Institute of Engineering and Technology
Deemed University, Longowal, Sangrur, India
e-mail:safrisoft@yahoo.com
Abstract— Cloud computing is summed up as a different model for allowing favorable, network as per demand to use shared devices of
computational resources which are collected and then released with marginal management effort or interaction with any client or any service
provider. Cloud computing is a well-known technology in the pasture of information technology that provides computing as a service. In cloud
computing environment the resources are provisioned on the basis of demand, as and when required. A large number of cloud users can request a
number of cloud services at the same time. Due to increase in the usage of cloud computing there is a need for a efficient and effective resource
allocation algorithm which can be used for proper usage of the resources and also check that the resource is not wastage. In this we propose a
priority based resource allocation algorithm which can be used for proper allocation of resources and also the resources are allocated efficiently
and effectively. In this paper, two strategies are proposed for the purpose of optimum resource allocation in which the first approach uses the
concept of specification matching and second uses the concept of priority based approach. In the first approach, different types of resources
(virtual machine) are allocated by taking three parameters into consideration: processing element, main memory, and network bandwidth. In the
second approach, one parameter is considered namely: Priority. In both strategies, users are allowed to submit the parameters during cloudlet
submission. The user inserted parameters will then be considered while allocating resources to them. The objectives of this research are to
improve utilization of resources and reduce the request loss.
Keywords- cloud computing, specification matching , priority
__________________________________________________*****_________________________________________________
I. INTRODUCTION
Cloud computing “refers to both the applications delivered as
services over the Internet, and the hardware and system
software in the data centres that provide those services”,
according to Armbrust et al.[1], and “is a utility oriented
distributed computing system consisting of a collection of
inter-connected and virtualized computers that are
dynamically provisioned and presented as one or more unified
computing resource(s) based on service-level agreements
established through negotiation between the service provider
and consumers” according to Buyya et al. [2]. Cloud
computing can be considered as an extension of grid
computing. One of the main characteristics of cloud
computing is on-demand self-service. That means Cloud
computing characteristically has provision for on-demand IT
resource allocation and instantaneous scalability. Unlike Grid
computing that typically provides persistent and permanent
use of all available IT resources, the cloud computing is very
specific on the consumer's demand, based on his current
computing requirements and therefore eliminates over-
provisioning of available IT resources.
The organizations can save the huge amount of expense by
avoiding build and manage large data centers for in-house
applications or data storage. The consumers of the cloud
computing do not have to own the IT infrastructure and
therefore need not care about the maintenance of servers and
networks in the cloud. They just pay for services on demand
that is based on the running of application instances normally
varying depending upon the use of Internet bandwidth, a
number of instances of action and amount of data transferred
at a specific time.
There are a variety of background activities in cloud
computing such as allocation of virtual machines (VMs), load
sharing, load balancing, process migration, and shared
memory access etc. which are completely abstracted from the
users view [3]. If the workload is shared by all the available
resources, it is known as load sharing. An even distribution of
workload on available resources is called load balancing [4].
One of the elementary aspects of virtualization technologies
engaged in cloud environments is resource consolidation and
managing. Using hypervisors inside a bunch environment
allows for a number of standalone physical machines to be
consolidated to a virtualized situation, thereby need less
Physical Machine (PM) physical resources than ever before.
The perception of Cloud computing has not only adjusted the
field of distributed systems but also fundamentally changed
how businesses consume computing today. While Cloud
computing offers many advanced features, it still has some
shortcomings such as the moderately high working cost for
both public and private Clouds. Better resource allocation is
depends on maneuver over utilization of physical machines
and virtual machines [5].
In this paper, we present an efficient resource allocation
technique that will help cloud owner to reduce wastage of

_______________________________________________________________________________________________
78
______________________________________________________________________________________
resources and to achieve maximum profit. Efficient resource
allocation in the cloud is a very challenging task as it needs to
satisfy both the user‟s requirements and server‟s performance
equally. Resource allocation in cloud computing environment
is defined as the assignment of available resources such as
CPU, memory, storage, network bandwidth etc in an economic
way. It is the main part of resource management. Yet, an
important problem that must be addressed effectively in the
cloud is how to manage Quality of Services (QoS) and
maintain Service Level Agreement (SLA) for cloud users that
share cloud resources.
The proposed policy based resource allocation strategy
provides maximized resource utilization and reduced
completion time of user requests. Four parameters are used by
the proposed approach for resource allocation to user requests
namely: Processing Element (PE), Main Memory (RAM),
Network Bandwidth (BW) and Priority (P). CloudSim-3.0.3
simulator is used for the purpose of simulation of results and
to analyze the behavior of proposed work.
The rest of this paper is organized as: Section II describes
different related works regarding allocation policies in a cloud
environment and about the parameters which have been used
by the researchers. Section III proposes a policy based
resource allocation approach (PRAA), by taking four
parameters into consideration for allocation of the resources.
Section IV describes the simulation evaluations on CloudSim,
which confirms the effectiveness of the proposed approach.
Section V presents the conclusions and future scope.
II. RELATED WORK
Garima et al .[6] proposed a priority based earliest deadline
first algorithm with task migration. The algorithm that
discussed in this paper used two job scheduling algorithm one
is priority based and second is the earliest deadline first
scheduling algorithm. The tasks were scheduled on the basis of
priority and high priority task gets scheduled first. The earliest
deadline first scheduling algorithm had scheduled the tasks
according to their deadline. The task having earliest deadline
get scheduled first and then the other tasks having the earliest
deadline will be scheduled next. The tasks were scheduled pre-
emptively and preemptable tasks would be migrated to the
other virtual machine.
Pawar and Wagh [7] had been present dynamic resource
allocation for preemptable task execution in the cloud. The
proposed priority based algorithm which considered multiple
SLA parameters such as memory, network bandwidth, and
required CPU time. In order to achieve the agreed SLA
objective, the proposed algorithm dynamically provisioned the
resources by preempting the low priority task with high priority
task.
Natasha and Gill [8] proposed a priority based resource
allocation method for handling a situation where two or more
requests at a particular instance of time had the same priority.
The work flow of this model was discussed in three stages. In
stage I, initial parametric values were generated such as
attaching priority to the request of Cloud user and the user
requests were sorted in descending order based the priority.
After priority assigned, requests with same priority were
grouped into a group known as an open group. A ready queue
was generated for all the requests which were in open group
and had not been executed yet in stage II. In stage III, grouped
requests were executed. A threshold value of available
resources was set and when the load needed by one or more
requests in ready queue exceed the threshold limit; request
should wait in the waiting queue.
Santhosh and Ravichandran [9] proposed a scheduling
algorithm with pre-emptive execution to overcome the non-
pre-emptive scheduling limitations. In non-pre-emptive
scheduling, if any high priority task arrived and wait because
of unavailability of the virtual machine then system
performance degrades. In this work, When a high priority task
arrived in between execution of other task and the deadline of
the task which was about to miss then the task would be
migrated to another virtual machine. This work was compared
with traditional EDF and other non-pre-emptive scheduling
algorithms. The results show that response time was reduced
and overall system performance was improved.
K C Gouda et al. [10] proposed a priority based resource
allocation approach which allocates the resource with
minimum wastage and provides maximum profit in a dynamic
cloud environment. This algorithm used different parameters
like cost, time, no.of processors request etc. This priority
algorithm decides the allocation sequence for different task
requested by the different user after considering the priority
based on some optimum threshold decided by the cloud service
provider.
Walsh et al. [11], proposed a general two-layer architecture
that uses utility functions, adopted in the context of dynamic
and autonomous resource allocation, which consists of local
agents and global arbiter. The responsibility of local agents is
to calculate utilities for given current or forecasted workloads
and range of resources, for each AE and results, are transfer to
the global arbiter. Where global arbiter computer near-optimal
configuration of resources based on the results provided by the
local agents. In global arbiter, the new configurations applied
by assigning new resources to the AEs and the new
configuration computed either at the end of fixed control
intervals or in an event triggered manner or anticipated SLA
violation.
Wazir Y. O. et al. [12] proposed a new approach for dynamic
autonomous resource management in the computing cloud. In
this paper, the author‟s contribution is two-fold. First,
distributed architecture is adopted where resource management
is decomposed into independent tasks, each of which is
performed by Autonomous Node Agents that are tightly
coupled with the physical machines in a data center. Second,
the Autonomous Node Agents carry out configurations in
parallel through Multiple Criteria Decision Analysis using the
PROMETHEE method. This approach is potentially more
feasible in large data centers than centralized approaches.
Savani and Amar [13] proposed a priority based resource
allocation algorithm which can be used for proper allocation of
resources effectively. In this algorithm, the resources in the
cloud are allocated according to the priority which is assigned
to each user request.
Zhen Xiao [14] presents design and implementation of an
automated resource management system that can avoid
overload in the system while minimizing numbers of servers
being used, it introduces the concept of skewness measure the
convent utilization of the server and improves utilization of
servers of multidimensional resource constraints.

_______________________________________________________________________________________________
79
______________________________________________________________________________________
Gaganjot and Sugandhi [15] proposed a preemptive priority
based job scheduling algorithm in green cloud computing
(PPJSGC). In this paper a green energy efficient scheduling
algorithm which makes the use of preemptive priority job
scheduling algorithm in cloud computing. This algorithm
focuses on reducing the power cost. The computing server is
selected on the basis of which satisfies the minimum resource
requirement of a job as per the best fit. Resources are allocated
based on the best allocation scheme. This method creates a
balanced between energy consumption and a load of the server.
This paper focuses on to design such an algorithm that
minimizes the carbon footprints and maximizes the resources
according to the suitability of the servers.
Dorian Minarolli and Bernd Freisleben [16] represents a VM
resource allocation in cloud computing via multi agent fuzzy
control, it focused on line grained dynamic resource allocation
of VM locally on each physical machine of a cloud provider
and consider memory and CPU as a resource that can be
managed. Fuzzy control is used to minimize a global utility
function as n a hill climbing heuristic implemented over fuzzy
rules. The problem considers is how to resource of a cloud
provider should be reallocated to VM dynamically where
workload changes to keep the performance according to SLA's.
Kazuki and Shin-ichi [17] proposed an optimal resource
allocation algorithm with limited energy power consumption.
Three parameters were taken into consideration to allocate
resources: processing element, network bandwidth, and electric
power consumption. The maximum numbers of the request
were processed by this approach. Total consumption of electric
power was reduced by aggregating requests being processed in
multiple areas.
Swachil and Upendra [18] proposed an improvement Priority
based Job Scheduling Algorithm in Cloud Computing using
Iterative Method. Improved priority based job scheduling
algorithm uses an iterative method to find priority of jobs and
resources and also finds priority of jobs to achieve better
performance. The proposed scheduling algorithm consists of
three levels of priorities: scheduling level (goal), resources
level (attributes) and jobs level (alternatives). Scheduling level
is the goal to be achieved by the scheduler, resources level are
the attributes that are available to achieve the desired goal and
the last level is the job level which are the available
alternatives from which the best job should be scheduled first.
This algorithm has better makespan and consistency than the
other algorithm like priority based job scheduling algorithm
and prioritized round robin algorithm.
Satveer and Birmohan [19] proposed an Optimum Resource
Allocation Approach (ORAA) in cloud computing. In this
paper, different types of resources (virtual machine) are
allocated by taking three parameters into consideration:
processing element, main memory, and network bandwidth.
Users are allowed to submit the parameters during job
submission. The user inserted parameters will then be
considered while allocating resources to them. The objective of
this paper is to make optimum resource allocation and achieve
efficient utilization of resources over public cloud.
Jiayin Li [20] presents a resource optimization mechanism in
heterogeneous IaaS federate multi cloud systems, that enable
preemptable task scheduling with resource allotment model,
cloud system model, local mapping, and energy consumption,
and application model. It is suitable for autonomic future in
cloud and VMs. The proposed online dynamic algorithms for
resource allocation and task scheduling. In proposed cloud
resource phenomenal every data center has a manager server,
the communication and resource allotment scheme works
between various servers of each data center for share
workloads among multiple data servers. The workload sharing
makes a large resource pool of flexible and cheaper resources
to resource allocation.
It has been observed from the literature that the strategies used
by the researchers provide a scope for improvement in resource
allocation. To eradicate the limitations of work done by the
researchers, a new priority based policy for resource allocation
has been proposed in this paper.
III. PROPOSED WORK
This paper proposes a method to handle the cloudlet requests
by managing resources using priority based approach for
resource allocation. The proposed approach is used to attain
better utilization of resources and reduced completion time of
user requests by means of optimized allocation of resources at
the virtual machine level. The Cloud parameters basically
represent the different types of resource: four parameters have
been chosen because of its dynamic nature. In the proposed
work, users give the parameter values during request
submission and these parameters are then be considered while
allocating VMs. The demand for any of these resources may
differ from user to user and hence are allocated dynamically.
The information about VMs with different parameters is
maintained by the data center broker in the form of resource
matrices and the VMs are allocated to the user at run time with
the help of these resource matrices. These matrices contain
merely two values either „1‟ or „0‟. Here '1' indicates that the
virtual machine of specific configuration is available and '0'
indicates its unavailability.
At the starting of the approach, to acquire some service, user
sends the request for resource allocation; the broker checks for
the available VM to process that request. When the request
with required parameter values arrives, it starts searching a
VM that fulfils its requirement with the help of the matrices
maintained by the broker. If a virtual instance of required type
exists, the arrived user request will be allocated to that VM.
After allocating all matching VMs to the user requests, the
broker will check unallocated user requests and available VMs
in the datacenter. Here the priority parameter submit by user is
used. The values of priority parameter submitted by user are
assumed. If user gives priority value '1' which represents the
priority of PE type of resource required most by the user. In
this condition, the broker search for the available VM which
has PE value equal or more than required by the user and
allocate to the user request. Accordingly, priority value '2'
represents the RAM type of resource and '3' represents the BW
type of resource. If the user gives priority value '0', then broker
will allocate any available VM to the user request without
checking any specification of the VM. So according to this
allocation of resources, all of the user requests will get
computing resource without waiting in waiting queue if VM
with any configuration is available. To check the availability
of matching VMs in the data center the essential condition is
as follows:
Z1[k1p][VMi]=1 (1)

_______________________________________________________________________________________________
80
______________________________________________________________________________________
p ϵ U and i ϵ J
Where U is a set of PEs having “a” types and J is set of VMs.
The availability of first parameter that is PE is measured with
the help of a matrix of dimension (a × j ). If the value of
corresponding equation 1 is „1‟, that means requisite resource
can be assigned to the request and if that value is „0‟, which
means no resource can fulfil the requirement of the user
request.
Z2[k2q][VMi]=1 (2)
q ϵ V and i ϵ J
Where V is a set of RAMs having “b” types and J is set of
VMs. Similarly, in order to check the availability of the
second parameter that is different types of RAM, matrix of
dimension ( b × j ) is used and
Z3[k3r][ VMi]=1 (3)
r ϵ W and i ϵ J
Where W is a set of BWs having “c” types and J is set of
VMs. To check the availability of the third parameter that is
the BW, a third matrix of dimension ( c × j ) is maintained,
where PE, RAM, and BW are three parameters of a request
that is send by the user to achieve a required type of service.
In the above matrices Z1, Z2 and Z3, the type of PE is
represented by k1, and it may vary from 1 to a, k2 represents
the types of RAM and it may vary from 1 to b, and k3
represents the different kind of BW and it may vary from 1 to
c.
When a user demand for a service with the requisite
parameters, the availability of service is determined with the
help of three resource matrices and each matrix will return a
set of VM Ids namely T1, T2 and T3. The common VM Ids
between T1, T2 and T3 are taken into a set T and these VM
Ids can fulfil the requirement of request of a user. One of the
VM Ids from set T is assigned to the request.
In the proposed approach, if all of three equations must be
satisfied at the same time, then a VM Id returned. If the
service is unattainable, that means VM with the required
configuration does not exist. Once a VM is allocated to a
request, the value of that VM in all three matrices is reset. For
the resource allocation based on the priority, at a time only one
condition must be satisfied like when priority is for PE then,
Equation 1 only checked. Accordingly, Equation 2 checked for
priority of RAM type of resource and Equation 3 will be
checked for the priority of BW type of resource.
In the proposed approach, the space shared policy of
CloudSim-3.0.3 simulator has been chosen over time shared
policy to achieve the concept of load balancing of individual
VM in the data center. The workload on a single virtual
machine is balanced efficiently with the help of fair allocation
of resources and space-shared policy of VM. The proposed
approach which inculcates the above benefits is described in
following algorithm.
Algorithm1: Optimum Resource Allocation (VMs, CRs)
Begin
1. Initialize CloudletRequest-Completed
2. Formulate matrices Z1, Z2, Z3 for parameters PE,
RAM, BW.
3. Update VM-List (based on VM parameters)
4. Update CR-List (based on CR parameters )
5. Repeat for i = 1 to length (CR-List)
a. Call SMA (VM-List, CR-List)
b. Call Update (VM-List, Request-Completed,
CR-List)
6. If (Request-Completed != CR-List && VM-List !=
Null)
Repeat for i = 1 to length (CR-List != Request-
Completed)
a. Call PBRAA (VM-List, CR-List)
b. Call Update (VM-List, Request-Completed,
CR-List)
End
Algorithm2: SMA (VM-List, CR-List)
Begin
1. For Id = 1 to length (VM-List)
if (Z1 [Cloudlet-List.PE] [Id] and
Z2 [Cloudlet-List.RAM] [Id] and
Z3 [Cloudlet-List.BW] [Id] = available)
2. Allocate VM to CR (from VM-List)
3. Call Update (VM-List, CloudletRequest-Completed,
CR-List)
End
Algorithm3: PBRAA (VM-List, CR-List)
Begin
1. For Id = 1 to length (VM-List)
If (CR.P = 1 && Z1 [CR-List.PE] [Id] = Available)
Allocate VM to CR (from VM-List)
Else if (CR.P = 2 && Z2 [CR-List.RAM] [Id] =
Available)
Else if (CR.P = 3 && Z3 [CR-List.BW] [Id] =
Available)
Else if (CR.P = 0 && Z1 [CR-List.PE] [Id] =
Available || Z2 [CR-
List.RAM] [Id] = Available || Z3 [CR-List.BW] [Id]
= Available)
2. Call Update (VM-List, CloudletRequest-Completed,
CR-List)
End
Algorithm4: Update (VM-List, CloudletRequest-
Completed, CR-List)
Begin
1. If (CloudletRequest-Completed = True)
Set (Z1 [CR.PE] [Id] and Z2
[CR-List.RAM] [Id] and Z3
[CR-List.BW] [Id] = 1)
Else
(Z1 [CR-List.PE] [Id] and

_______________________________________________________________________________________________
81
______________________________________________________________________________________
Z2 [CR-List.RAM] [Id] and
Z3 [CR-List.BW] [Id] = 0)
2. Update VM-List
3. Update CR-List
End
Figure 1. Dataflow Diagram of Proposed Method
The proposed algorithm in this chapter takes O (m2
* n2
) time
and O (n2
) space to process all the Cloudlets in the virtual
Cloud in worst case where n is the number of virtual machines
and m is the number of Cloudlets.
IV. RESULT AND DISCUSSION
For the purpose of simulation, the CloudSim-3.0.3 simulation
toolkit has been used in the proposed work as the primary
objective of this toolkit is to provide a generalized and
extensible simulation framework that enables seamless
modeling, simulation, and experimentation of emerging Cloud
computing infrastructures and application [26].
The Cloud service system with proposed policy is shown in
figure 4.1. Only one data center is considered in the public
Cloud. There are different types of user requests and VMs in
this CloudSim-3.0.3 environment. A new broker policy is
implemented for resource allocation in which space-shared
policy is used by Cloudlets on VMs. The value of parameters
PE, RAM, and BW that a user can submit in its request are
listed in table 4.1 and the values of Priority parameter is listed
in table 4.2
TABLE 4.1 The values of Three Parameters
S.N. Type PE RAM BW
(No. of
Cores) ( in MB) ( in Mbps)
1 1 1 Core 1024 1000
2 2 2 Core 2048 2000
3 3 4 Core 4096 4000
4 4 8 Core 8192 8000
TABLE 4.2 The Values of Priority Parameter
S. No. Value of Priority (P) Demand for
Parameter
(Resource Type)
1 0 Any available VM
2 1 PE type of resource
3 2 RAM type of resource
4 3 BW type of resource
In this simulation environment, three different VM-sets in
which VM-set 1 consist of ten VMs and VM-set 2 consist of
eleven VMs and VM-set 3 consist of twelve VMs. Three
different data-sets in which each data-set consist of ten or
twelve user requests are created. All three data-sets are
executed on VM-sets separately which confirms the simulation
results. The specifications of the VMs in VM-sets and
specifications of cloudlet requests in data-sets are listed in
table 4.3 and table 4.4 respectively
The proposed approach for allocation of resources has been
verified with the help of three datasets. The Cloudlets
requested by the users of dissimilar types due to variation in
input parameters and these are allocated to the three VM-sets.
Table 4.5, Table 4.6 and table 4.7 show the number of
requests that has been accepted using matching method and
using priority separately, to provide the services, and this is
because a machine with required configuration has been
found.
As a result of this exact matching and by using priority given
by user, a mapping has been performed between the requests
and the virtual machines. In other words, user requests have

_______________________________________________________________________________________________
82
______________________________________________________________________________________
been assigned to the virtual machine so as to get its required
service. Sometimes, if there is no available virtual machine
and in case a virtual machine is available, which does not
fulfils the need of user request, then only a request will be
unallocated. Using this information about user requests, the
average utilization of virtual machines has been calculated.
TABLE 4.3 The Specifications of Virtual Machines
S. N.
VMs
VM Set 1 VM Set 2 VM Set 3
PE
Type
RAM
Type
BW
Type
PE
Type
RAM
Type
BW
Type
PE
Type
RAM
Type
BW
Type
1 VM 0 Type 1 Type 1 Type 1 Type 1 Type 2 Type 1 Type 1 Type 2 Type 1
11 VM10 - - - Type 4 Type 1 Type 8 Type 4 Type 1 Type 8
12 VM 11 - - - - - - Type 8 Type 4 Type 8
TABLE 4.4 The Specification of Cloudlet Request
CR Id Data Set 1 Data Set 2 Data Set 3
PE
Type
RAM
Type
BW
Type
P PE
Type
RAM
Type
BW
Type
P PE
Type
RAM
Type
BW
Type
P
0 1 2 1 0 1 2 1 0 1 2 1 1
1 1 2 2 0 1 2 3 2 1 2 1 0
2 2 1 2 0 2 2 2 1 1 2 2 3
3 2 3 2 0 2 2 3 3 2 1 2 0
4 2 3 3 0 2 3 3 2 2 3 2 2
5 3 4 2 0 3 4 2 1 2 3 3 0
6 3 3 3 0 3 3 3 0 2 3 3 1
7 3 2 3 0 3 2 4 1 3 3 3 0
8 4 3 4 0 4 3 4 0 3 4 2 3
9 4 4 3 0 4 3 3 2 4 4 4 0
10 4 4 4 0 4 4 1 3 4 4 2 2
11 4 4 2 0 4 4 3 1 4 1 2 2
TABLE 4.5 The Result on VM Set 1
Data Set No. of CR No. of VMs Allocation of User Requests Average Utilization (Pu)
(In %)Using
Matching
Method
Using
Priority
Unallocated
Requests
01 12 10 10 00 02 100.00
02 12 10 05 05 02 90.00
03 12 10 08 02 02 95.00

_______________________________________________________________________________________________
83
______________________________________________________________________________________
TABLE 4.6 The Results on VM Set 2
(In %)Using
Matching
Method
Using
Priority
Unallocated
Requests
01 12 11 09 01 01 90.00
02 12 11 05 06 00 100.00
03 12 11 08 03 00 95.00
TABLE 4.7 The Results on VM Set 3
(In %)Using
Matching
Method
Using
Priority
Unallocated
Requests
01 12 12 09 03 00 100.00
02 12 12 06 05 01 85.34
03 12 12 07 05 00 90.00
It has been observed, the proposed approach has achieved an
average utilization of 100% for dataset 1 on VM Set 1, VM
Set 2 and on VM Set 3. Dataset 2 and Dataset 3 also
achieved nearly 90% to 95 % average utilizations on VM
sets. In the table 4.6 and 4.7, there is case in which one
cloudlet request is unallocated when a VM is available in
VM set. It is because; the resource priority given by user
can‟t fulfill using available VM.
V. CONCLUSION
The optimum resource allocation using configuration
matching and priority approach is used for the purpose to
improve the resource utilization and to reduce the
completion time of user requests. Basically, this approach is
proposed to eradicate the limitations of scheduling of
resources of optimum allocation. At the time when priority
given to the resources, there has a situation where there are
available computing resources with miss-matched
configuration to serve the user request, but the resource is
not allocated to the request. In this situation, the request loss
is more and utilization of resources is somehow not
maximized as well. So, to eliminate this limitation of that
situation this method of optimum allocation of resource by
using SMA and priority is proposed.
VI. REFERENCES
[1] R. Buyya, C. S. Yeo, and S. Venugopal, “Market-Oriented
Cloud Computing: Vision, Hype, and Reality for
Delivering IT Services as Computing Utilities”, 10th
IEEE
International Conference on High Performance Computing
and Communications (HPCC 2008), Sept. 25-27, 2008.
[2] M. Armbrust, A. Fox, R. Griffith, A. D. Joseph, R. Katz, A.
Konwinski, G. Lee, D. Patterson, A. Rabkin, I. Stoica, and
M. Zaharia, “A view of Cloud Computing”,
Communications of the ACM, Apr. 2010.
[3] Gouda, K. C., T. V. Radhika, and M. Akshatha, “Priority
based resource allocation model for cloud computing”,
International Journal of Science, Engineering and
Technology Research , pp-215, 2013.
[4] Ren, Haozheng, Yihua Lan, and Chao Yin, “The load
balancing algorithm in cloud computing environment”,
Computer Science and Network Technology (ICCSNT),
2012.
[5] Dixit, Rituraj, Prashant Buyan, and Vijay Kumar. "Priority
Based Dynamic Resource Allocation for Green Cloud
Environment: A Survey” International Journal of
Engineering and Management Research, 2014.
[6] Gupta, Garima, Vimal Kr Kumawat, P. R. Laxmi,
Dharmendra Singh, Vinesh Jain, and Ravinder Singh, “A
simulation of priority based earliest deadline first
scheduling for cloud computing system” In Networks &
Soft Computing (ICNSC) International Conference on, pp.
35-39, 2014.
[7] Pawar, Chandrashekhar S., and Rajnikant B. Wagh,
“Priority based dynamic resource allocation in cloud
computing” In Cloud and Services Computing (ISCOS),
International Symposium, pp. 1-6, 2012.
[8] N. Gill, “Enhanced priority based resource allocation in
Cloud computing,” In The Next Generation Information
Technology Summit (4th International Conference) , pp.
121-126, September 2013.
[9] Santhosh, R., and T. Ravichandran, “Pre-emptive
scheduling of on-line real time services with task migration
for cloud computing” In Pattern Recognition, Informatics
and Mobile Engineering (PRIME), pp. 271-276, 2013.
[10] Gouda, K. C., T. V. Radhika, and M. Akshatha, “Priority
based resource allocation model for cloud computing”
International Journal of Science, Engineering and
Technology Research, pp-215, 2013.

_______________________________________________________________________________________________
84
______________________________________________________________________________________
[11] Walsh, William E., Gerald Tesauro, Jeffrey O. Kephart,
and Rajarshi Das, “Utility functions in autonomic systems”
In Autonomic Computing International Conference, pp.
70-77, 2004.
[12] Yazir, Yagiz Onat, Chris Matthews, Roozbeh Farahbod,
Stephen Neville, Adel Guitouni, Sudhakar Ganti, and
Yvonne Coady, “Dynamic resource allocation in computing
clouds using distributed multiple criteria decision analysis”
3rd International Conference IEEE, pp. 91-98, 2010.
[13] Savani Nirav M, Amar Buchade, “Priority Based Resource
Allocation in Cloud Computing” International Journal of
Engineering Research & Technology (IJERT), 2014.
[14] Xiao, Zhen, Weijia Song, and Qi Chen, “Dynamic resource
allocation using virtual machines for cloud computing
environment” IEEE transactions on parallel and distributed
systems, 2013.
[15] Kaur, Gaganjot, and Sugandhi Midha, “A Preemptive
Priority Based Job Scheduling Algorithm in Green Cloud
Computing” 6th International Conference IEEE, 2016, pp.
152-156.
[16] Minarolli, Dorian, and Bernd Freisleben, “Utility-based
resource allocation for virtual machines in cloud
computing” In Computers and Communications (ISCC),
IEEE Symposium, pp. 410-417, 2011.
[17] K. Mochizuki and S. I. Kuribayashi, “Evaluation of optimal
resource allocation method for Cloud computing
environments with limited electric power capacity,” In
Network-Based Information Systems (NBiS), 14th
International Conference on IEEE, pp. 1-5, September
2011.
[18] Patel, Swachil, and Upendra Bhoi, “Priority based job
scheduling techniques in cloud computing: a systematic
review” International journal of scientific & technology
research, 2013.
[19] S Satveer, and Birmohan Singh. "Optimum resource
allocation approach in cloud." In Advanced
Communication Control and Computing Technologies
(ICACCCT), International Conference on IEEE, pp. 600-
605, 2016.
[20] Li, Jiayin, Meikang Qiu, Zhong Ming, Gang Quan, Xiao
Qin, and Zonghua Gu, “Online optimization for scheduling
preemptable tasks on IaaS cloud systems” Journal of
Parallel and Distributed Computing, 2012.
.

Optimum Resource Allocation using Specification Matching and Priority Based Method in Cloud

More Related Content

What's hot (15)

Similar to Optimum Resource Allocation using Specification Matching and Priority Based Method in Cloud (20)

More from rahulmonikasharma (20)

Recently uploaded (20)

Optimum Resource Allocation using Specification Matching and Priority Based Method in Cloud