Skip to main content

Full text of "Reducing Energy Consumption in Smart System through Mobilouds Framework"

See other formats



International Journal of Trend in Scientific 
Research and Development (IJTSRD) 
International Open Access Journal 



♦. 

ISSN No: 2456 - 6470 | www.ijtsrd.com | Volume - 2 | Issue -1 


♦ 

♦ 


Reducing Energy Consumption in Smart System 
through Mobilouds Framework 


Mr. Muneshwara M.S 

Dept, of CS&E, BMSIT&M, 
Yelahanka, Bengaluru-566064, 
Karnataka, India 


Dr. Anil G.N 

Dept, of CS&E, BMSIT&M, 
Yelahanka, Bengaluru-566064, 
Karnataka, India 


Dr. Thungamani M 

Department of CS&E , COH UHS 
Campus GKVK, Yelahanka, 
Bengaluru -560065, Karnataka 


ABSTRACT 

Mobile cloud computing (MCC) it’s related energy 
implication are seen everywhere in large-scale. 
Offloading computationally raised endeavor’s to the 
cloud datacentres being the basic thought driving 
MCC (Mobile Cloud Computing), a vast segment of 
the mobile terminal assets partaking in the MCC 
(mobile Cloud Computing) aggregate execution are 
wasted as they remain sit out of gear till the mobile 
terminals get response from the datacentres.. This is 
an additional wastage of assets near to the cloud 
assets are starting at now being tended to as colossal 
vitality customers. Despite the fact that the assets 
consumed of the site without moving mobile 
resources are unimportant in contrast with the cloud 
partner, such utilizations impact sly affects the mobile 
devices bringing about superfluous battery channels. 
Mobilouds which consolidate a multi-level processing 
architecture with various phases of process cluster 
limits and a product application to supervise vitality 
utilization. Mobilouds framework energizes the 
mobile device co-operation in the MCC (Mobile 
Cloud Computing) synchronized effort execution, 
there by lessens the weight of idle mobile resource 
and uses such idle resource in the actual job 
execution. 

Keywords'. Cloud computing, mobile cloud 
computing, mobilouds 

I. INTRODUCTION 

The cloud is an arrangement of various sort of 
software and hardware that works join to conveyance 
numerous things of registering to the end user as an 


online services. Cloud computing (CC) is the 
utilization of hardware and software to reaction an 
administration over a system with cloud computing, 
users can get to any record and any application from 
any gadget that can get to the web. Mobile Cloud 
computing (MCC) is the consolidated approach of 
cloud computing (CC), Mobile Computing (MC) and 
wireless network (WN) to convey better 
computational asset to mobile user, network operators 
and in addition cloud computing suppliers (CCS). 
Mobile Cloud Computing (MCC) is an incorporated 
system that consolidates cloud datacenters, mobile 
devices and correspondence framework. Mobile 
Cloud Computing (MCC) organizations are broadly 
utilized as a part of different applications, for 
example, e-leaming, Tele-checking, Tele-surgery, IT, 
business administrations and so forth. 

Mobile Cloud Computing (MCC) advantage models 
assemble a complicated association between 
foundation suppliers, application and specialist 
organizations, designers and end-clients. Framework 
suppliers generally it will incorporate some additional 
components for equipment and programming 
organizations; application and specialist co-ops are 
responsible for executing customer requested 
organizations; engineers are all around who make 
applications being encouraged on the cloud 
datacenters; and end-clients are the shoppers of the 
Mobile Cloud Computing (MCC) organizations. The 
end-clients of the Mobile Cloud Computing (MCC) 
administrations they don't have the advantage of 
control over the subordinate foundations, for instance, 


@ IJTSRD | Available Online @ www.ijtsrd.com | Volume-2 | Issue-1 |Nov-Dec2017 


Page: 627 







International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470 


equipment, arrange, servers and so forth.., yet they 
can have an aggregate control over the applications 
sent by them. The achievement of such a Mobile 
Cloud Computing (MCC) structure depends on upon 
the predictable blend of committed equipment and 
programming assets of the three core technologies. 

II. LITERATURE REVIEW 

Alzharani et al [1] discussed outline of Mobile Cloud 
Computing (MCC) rewards, drawbacks. The authors 
have likewise talked about significance of versatile 
cloud application and highlights of the portable 
distributed computing open difficulties. Mathew et al 
[2] investigates a portion of the specifics of these 
wellbeing and wellness application. And furthermore 
author presented advancement calculation as an 
instrument to effectively handle information point 
locally by mobile devices. This calculation can take 
focal points of the nearby preparing energy of mobile 
devices and diminishes correspondence taken a toll 
between versatile endpoint and cloud based long haul 
information administrations. Kitanov et al [3] 
discussed about a method to solve the problem related 
to throughput, low-latency (delay), high mobility 
(speed) and high capacity. The author have also 
discussed about importance of mobile cloud 
application, services and current research trends. 
Sanaei et al [4] discussed about a method to solve the 
problem related to heterogeneity in convergent 
figuring and systems administration (wired and 
remote systems) and separation it into two 
measurements to be specific vertical and horizontal. 
The author have also focused on issue related to 
impact of heterogeneity in MCC are researched and 
overwhelming heterogeneity dealing with approach 
like virtualization, middleware and benefit situated 
design. Abolfazli et al [5] discussed about problem 
related to mobile augmentation domain and present 
taxonomy of CM A approach and also main objective 
like effect of remote resource on the quality and 
dependability of increase prepare and utilizing 
changed cloud based asset in expanding mobile 
devices. The author likewise break down the 
condition of workmanship CMA approach. Lose et al 
[6] discussed about problem related to mobile device 
do not need high end resource (e.g., processing speed, 
storage and memory capacity) since all the data and 
complex computing can be offloaded to the cloud and 
cloud will perform action on that data and response 
will give back to user. The author also focused on 
cloud computing infrastructure to augment the use of 


mobile phones in information and communication 
technology for development and also services that 
cloud can offer to improve and support the use of 
mobile phone. Satyanarayanan et al [7] discussed 
about problem related to specialized snags to this 
change and versatile client misuses virtual machine 
innovation to quickly instantiate redid benefit 
programming on a close-by cloudlet and afterward 
utilizes that administration over a remote LAN the 
cell phone ordinarily work as a thin customer 
regarding the administration. Creator additionally 
talked about cloudlet is a put stock in, asset rich PC or 
group of PC that is very much associated with the web 
and accessible for use by adjacent cell phones and 
furthermore utilizing cloudlet likewise disentangles 
the difficulties of taking care of the pinnacle transfer 
speed demand of different clients intuitively 
producing and getting media, for example, superior 
quality video and high determination image. Oureshi 
et al [8] discussed about the problem related to the 
information handling, stockpiling and other escalated 
operation. The creator additionally centered on best in 
class versatile distributed computing and its execution 
strategies. Ravi et al [9] discussed about the problem 
related to the correspondence overhead, offloading of 
utilization execution to cloud customer more vitality 
than executing in the gadget itself. The creator 
additionally centered around the structure for vitality 
proficient consistent administration with highlight 
like, associating heterogeneous cell phone to frame 
portable impromptu cloud. Administration disclosure 
in versatile specially appointed cloud and offloading 
choices. Li et al [10] discussed about the problem 
related to the key security challenges confronted by 
green distributed computing condition and outline a 
virtualization security confirmation design named 
cyber guarder to address the security problem with 
consideration of energy efficient. The author also 
focused on Virtual Machine Security Service (VMSS) 
incorporating a numeral of novel technique including 
VMM-based integrity measurement approach for a 
Netapp isolation mechanism for as user isolation, VM 
(virtual machine) separation and virtual network 
separation of multiple Netapps according to dynamic 
energy efficiency and security needs. Al-Aqrabi et al 
[11] discussed about a method to solve the problem 
related to Business Intelligent (BI). The author also 
focused on cloud facilitating of BI has been proposed 
with the assistance of reenactment on Op-net which 
involving a cloud show with different OLAP 
application servers apply parallel question stacks a 
variety of servers facilitating social database. Li et al 


@ IJTSRD | AvailableOnline@www.ijtsrd.coml Volume-2 | Issue-1 |Nov-Dec2017 


Page: 628 




International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470 


[12] discussed about the method to solve the problem 
related to protection and safeguard of Iaas cloud 
environment and also resource usage. The author also 
focus on these challenges to overcome this problem, 
author proposed a new system called Cloud-Mon 
which empowers dynamic asset allocation. Cloud- 
Mon provides two kind of technique to maintain high 
resource efficiency. The first technique is by making 
use of fuzzy model it establish a complex relation 
amongst execution and asset requests of a NIDS-VA 
and builds up an online Fuzzy controller to organize 
asset distribution for NIDS-VA under shifting system 
activity. Second one is worldwide asset booking 
approach for improving the asset productivity of the 
cloud conditions. Lee et al [13] discussed about a 
method to solve the issue identified with pivot time 
and asset squander in cell phones. The author likewise 
centered on the proposed system can be connected to 
a more practical distributed computing and 
furthermore creator explained the calculation of 
structure and directed a broad arrangement of 
reproduction with different situations. Liu et al [14] 
discussed about a method to solve the problem related 
to extend battery lifetime, increase processing speed 
and approximately a few critical issue, for example, 
issue of dead spots or scope openings. The author 
additionally centered around the errand offloading 
utilizing self-composed criticality (TOSOC) utilized 
this technique to take care of the issue identified with 
dead spots or scope gaps and furthermore issue 
identified with benefit postpone imperatives. Nir et al 
[15] discussed about a method to solve the problem 
related to minimize computational time and energy 
consumption and also user defined constraints. The 
author also focused on centralized broker node 
approach, this approach will solve the problem related 
to undertaking task to limit the aggregate vitality 
utilization over. 

Mobile Cloud Computing (MCC) can be viewed as an 
extension that fills the crevice between the limited 
computing resources of SMD (Storage Module 
Device) and processing necessities of escalated 
applications on SMDs. The Mobile Cloud Computing 
(MCC) Forum characterizes Mobile Cloud 
Computing (MCC) as takes after: "mobile Cloud 
Computing (MCC) in any occasion troublesome shape 
intimates a framework where both the information 
stockpiling and the information handling occur 
outside of the framework. Versatile cloud applications 
move the figuring power and information stockpiling 
far from framework and into the cloud, bringing 


applications and portable registering to not simply 
mobile devices clients yet rather a liberally more 
expansive degree of portable endorsers". Mobile 
Cloud Computing (MCC) has pulled in the 
consideration of business specialists as a useful and 
valuable business solution that limits the development 
and execution costs of mobile applications, enabling 
mobile user to obtain most recent innovation 
advantageously on an on-demand basis. Fig. 4.1 
demonstrates the general perspective of Mobile Cloud 
Computing which is made out of three primary parts: 
the mobile device, wireless communication implies, 
and a cloud infrastructure that contains data centers. 
These last give storage services, processing, and 
security instruments for both the cloud environment 
and mobile device. 



Mobile 

Devices 



Wireless 

Communication 

Means 



Computational 

Cloud 


Figure-1: Mobile Cloud Computing 


Computation offloading is the errand of sending 
computation intensive application components to a 
remote server. As of late, various computation 
offloading structures have been proposed with a few 
methodologies for applications on mobile devices. 
These applications are partitioned at various 
granularity levels and the components are sent 
(offloaded) to remote servers for remote execution 
keeping in mind the end goal to extend and improve 
the SMD's abilities. Be that as it may, the computation 
offloading mechanism are as yet confronting a few 
difficulties 





local execullon 

remote execution 


Figure 4.2: computational Offloading 


III. PROPODED METHODOLOGY 

The proposed solution addresses the issues of vitality 
consumption brought about by the sit out of gear 


@ IJTSRD | AvailableOnline@www.ijtsrd.coml Volume-2 | Issue-1 |Nov-Dec2017 


Page: 629 














International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470 


mobile assets and designed a Hybrid Energy Efficient 
MCC (HEE-MCC) framework, named Mobilouds, 
with the end goal of expanding the involvement of the 
mobile devices in the collaboration Mobile Cloud 
Computing MCC job completion at last to minimize 
the undesirable vitality utilizations of the sit out of 
gear mobile resource with reduced service delays. 

The major contributions of this paper include, 

1. A new multi-level process configuration named 
Mobilouds, made out of various sizes of process 
cluster for vitality productive Mobile Cloud 
Computing (MCC) consolidated execution. This 
Mobilouds framework can be both climbed to a 
higher limit cluster in the midst of asset deficiency 
and limited when there are wealth assets in the 
process cluster with the ultimate objective of 
limiting vitality usage. 

2. The Mobilouds application which is a product 
procedure passed on to support the functionalities 
of the Mobilouds framework. This product 
procedure continues running in the mobile system 
for processing the asset accessibility in the 
portable terminals. This Mobilouds methodology 
serves to picks the ideal process cluster from the 
Mobilouds framework, and a vitality efficient 
Mobile Cloud Computing (MCC) collective 
execution is proficient in the picked process 
cluster by the technique for a passing on a 
dispersed offloading system among the available 
asset in the cluster. 

IV. SYSTEM ARCHITECTURE 

The architecture identifies the major modules and the 
functional interfaces between them. 

The System architecture is shown below. 



Figure 5.2: System Architecture 


Offloader: This module accesses the task request for 
offloading and sent to Target cluster module. Also 
when the result of offloading arrives, it notifies to the 
User. 

Target Cluster Selection: This class identifies the 
target cluster where the task has to offloaded and 
executed. It then transfers the code to the target 
machine for execution. 

Offload Handler: It does the work of packing the 
task and sending to task machine for execution. 

In this architecture mobile user will offload the task 
into offload handler and that task will processed by 
target cluster selection this will select the particular 
cluster and in that cluster there will be N number of 
nodes. Offload handler will select the particular node 
based on the node availability and then it will assign a 
task to the node and node will perform the task and 
then result will gives back to the offload handler and 
finally the offload handler will send the results to off¬ 
loader. The result will be notify to user with simple 
notification and finally user can view the results. 

V. DATA FLOW DIAGRAM 

Level 0 Data Flow Diagram (DFD) gives the 
overview of the data flow of the work. The above 
DFD provides us the flow of the data and gives us 
only the brief detail about the flow of data. 

Offloading is overall process in the system. 



Figure 5.8: Level 0 Data flow Diagram 


Offload process is split to sub process and drawn in 
the level 1 data flow. 



Figure 5.9: Level 1 Data flow diagram 


@ IJTSRD | AvailableOnline@www.ijtsrd.coml Volume-2 | Issue-1 |Nov-Dec2017 


Page: 630 




































International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470 


VI. PSEUDO CODE 

Pseudo Code is a partial code which explains the main 
content (explanation) of any program or any 
algorithm. It includes the aim of the function which 
we are designing. 

The explanation behind using pseudo code is that it is 
less requesting for people to grasp than common 
programming language code, and that it is a gainful 
and condition free portrayal of the key norms of a 
count. It is routinely used as a piece obviously 
readings and consistent creations that are recording 
distinctive computations, and moreover in organizing 
of PC program change, for depicting out the structure 
of the program before the real coding happens. 

The pseudo code for offloading at various modules is 
given 

Node 

Function: SubmitJob 

Input: Task T 

If resource need of T is available in system 
Execute Job Locally 

Else 

Send offload request to cluster manager 

End 

Node 

Function: Execute offload Job 

Input: Task T, offloading Node X 

Results C-Execute T in local 

Send Result to Node X 

Cluster Manager 

Function: SelectTargetNode 

Input: Task T, offloading Node X 

Node Y = Select Best Node in Local Cluster; 

If Y==null 

Forward Request to Next Level Cluster 
Else 

Send Task Execution Request to Node Y. 

End 

Function: connect 

Input: Server IP, Port P, Guiist G 

If server IP and Port P, Guiist G is matches with the 

cluster 

Results <-Node will connect to cluster 
Else 

Node will not connect to cluster 
end 

Function: Noderegister 

Input: type 3 

If the message type is 3 


Results <-Node is registered in the cluster manager 
Else 

Node is not registered in the cluster manager 
End 

Function: OffloadReq 

Input: type 1 

If the requesting to offload task into the cluster 
Results <- node is requesting to offload task into the 
cluster 
Else 

Node is not requesting 
end 

Function: Offloadres 

Input: type 2 

If the message type 2 

Results^- Node is responses to the Offloading task 
Else 

Cluster is not responding 
end 

VII. INTERPRETATION OF RESULT 

The following snapshots describe the results or 
outputs that we will get after step by step execution of 
all the modules of the system. 

Interpretation: 

The results are categorized in three categories 


Case 1 

Offloading within same Node (Do- 
Local) 

Case 2 

Offloading within sane cluster 

Case 3 

Offloading to Next Level Cluster 


> Case 1 Screen Shots: Offloading within same 
Node (Do-Local) 



Figure 8.1: cluster port value 


@ IJTSRD | AvailableOnline@www.ijtsrd.coml Volume-2 | Issue-1 |Nov-Dec2017 


Page: 631 























International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470 


In this snap shot 8.1 we can see that Listen port as 
5000 that means that in cluster, cluster manager will 
configures server at port value as 5000. Through this 
port value all node will get connected with the cluster 
and once if you give listen port as 5000 and press the 
start button if will starts the cluster manager and we 
can see either the cluster manager is started or not in 
Log window. And let’s suppose if you want to add 
one more cluster we have go next level CM there we 
can add many number of cluster. And the Registered 
Nodes we can view the how many number of nodes 
are registered in the cluster. 


O Node • NetBrani IDE 74X1 



resource need to compute that task in the form of KB. 
After that we can see that two option that is offload 
button and one more is Do-Local in this case let’s go 
for Do-local if you click that button it will compute 
task in that particular node itself. 

O Node • NetOeans IO€ 7.011 - e ^9 



Figure 8.4: Result view in log 





Figure 8.2: Node registration 


In this snapshot 8.4 we can see response time to 
compute that task. To compute that task in that same 
node it took 5093 mile second to compute that task 
and also we can see that one statement that is job is 
executed in locally. In other window we can view 
registered node in the cluster manager. 


In this window 8.2 we can see the node details that is 
before node is connected to the cluster we need to 
give node id and Node cluster manager IP address and 
also cluster manager port value so that node will get 
registered in that cluster manager and also Resource 
available in that particular node in this snapshot we 
can see available resource as 10KB and also we can 
see that one pop-up that tells that whether the node is 
connected to cluster or not. 



Figure 8.3: Do-local process 

In this snapshot 8.3 we can see the registration of 
node ID along with IP address and port value and 
available resource in that node. In the pop-up that is 
node window we can see how exactly we will offload 
that task that is firstly we browse that task in the local 
disk and then we select the task. The task will be in 
the form of jar extension and also we need to give the 


> Case 2 Screen Shots: Offloading within sane 
cluster 



Figure 8.5: Two Node Registration 


In this snapshot 8.5 we can view the two node are get 
connected to same cluster manager with same port 
value. And each node will have the unique node Id 
that is node number 1 and 2 followed by cluster 
manager IP address followed by cluster manager port 
value that is 5000 and available resource in that node 
in the node 1 it has 10KB of resource and node 2 has 
20KB of resource. The available resource tells that it 


@ IJTSRD | Available Online @ www.ijtsrd.com | Volume-2 | Issue-1 |Nov-Dec2017 


Page: 632 


















































































International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470 


can allocate 10KB of resource if any task comes to 
that node if any task come more than that available 
resource it cannot perform task. 


0 

Node- NetBeans IDE 7.1X1 

- 0 

v 

JU Node 

- ° 1 Jb Node 

. dQ 

| HooeCofflgurciNn ’ ouotfJoD [ loo | 

J NodtCortguiaon | OdoMJoe I Log ] 

130© 



08 11 1 I 



Figure 8.6: File offload to cluster 


In this snap shot 8.6 we can see that we can see that 
node 1 is offloading task and we can see two pop-up 
that first one is offloading request is send to the 
cluster manger and other one is check the result. This 
how offloading of task will be done in cluster 
manager level. The offloaded task will be in the form 
of jar format that java executable form. 


Node -NetBeans IDE 7.(11 



i 19 K 


Figure 8.7: Result view in log 


In this snap shot 8.7 we can see the result in log 
window. In the node 1 window we can see response 
time to compute that task and also we can see that 
result will be send to local disk of the system there we 
can view the results. And in the node 2 window we 
can see that two statement that is it got offload request 
form node 1 with task file name that is test-app.jar. 
Once the node is received request node will perform 
the task and then it will sends results to node 1. In the 
cluster manager window we can see that one 


statement that is found eligible node as 2. The cluster 
manager will scheduling job it first check whether the 
node has enough resource once if it got to know it has 
enough resource then only it assigns task to that node 
if not it will displays no eligible node are present in 
the cluster. 

> Case 3 Screen Shots: Offloading to Next Level 
Cluster 




be 



Figure 8.8: Multi level cluster 


• 30® 


In this snapshot 8.8 we can see that how exactly 
multi-level cluster will create. Initially cluster 
manager 1 with port value 5000 and cluster manager 2 
values with port value 6000. Once if you give port 
value and then press start button cluster manger will 
get started. 

o ClusterManager NetBeans IDE 7.0.1 - O 

** M \M Ouster Manager - ° ifc Cluster Manager - ° I 



Figure 8.9: Connecting cluster levels 


This snapshot 8.9 show how exactly two cluster 
manager that is cluster manager 1 and 2 will get 
connected that is in the cluster manager window we 
can see one option that is next-level CM in that 
window if you give cluster manager IP address 
followed by next-level cluster manager port value if 
you press add button it will add the next level cluster 
manager. 


@ IJTSRD | AvailableOnline@www.ijtsrd.coml Volume-2 | Issue-1 |Nov-Dec2017 


Page: 633 





















































































International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470 



Figure 8.10: Results view in log 


In this snapshot we can view how exactly task will 
execute in the next-level cluster manager. Initially job 
will offloaded from the node 1 into the cluster 
manager 1 and cluster manager 1 will check for 
available resource in same cluster if any resource 
present in the cluster then job will assigned to 
available resource if not cluster manager will forward 
to the next level cluster manager that is cluster 
manager 2 once again cluster manager will check for 
the available resource in the cluster manager 2 if it 
found resource in the cluster manager 2 the it will 
assigns that job to particular node based on that 
availability of resource in that node. We can see result 
in the node 1 log window to compute that job it took 
352 mile seconds and also how exactly cluster 
manager will forward offload request to the next level 
cluster and we can see IP address of cluster manager 
and also node port value. This is how task will 
performed in the next level cluster manager. 

CONCLUSION 

In this work we implemented a proof of concept 
offloading server. The offloading based on the 
resource usage. The offload decision was made in 
hierarchy till the best resource matching to the task 
request is available. Through this work, we 
demonstrated that if the task is executed in place 
where resource is not available, it takes much longer 
time or may also fail, but due to offload and matching 
to best resource the time was able to execute 
successfully in less time. We also offloaded to live 
Microsoft azure cloud and proved that the concept 
was able to resort to cloud at top level hierarchy. 


REFERENCES 

1. A. Alzahrani, N. Alalwan and Mohamed Sarrab, 

"Mobile Cloud Computing: Advantage, 

Disadvantage and Open Challenge," Proc. 7th 
Euro American Conf. Telematics and Information 
Systems, Chile, 2014. 

2. G. Mathew and Z. Obradovic, "Improving 
Computational Efficiency for Personalized 
Medical Applications in Mobile Cloud Computing 
Environment," IEEE Int. Conf. Healthcare 
Informatics (ICHI), pp. 535540, Philadelphia, 
Spetember 2013. 

3. S. Kitanov and T. Janevski, "State of the Art: 
Mobile Cloud Computing," 6th Int. Conf. 
Computational Intelligence, Communication 
Systems and Networks, Tetova, pp. 153-158, May 
2014. 

4. Z. Sanaei, S. Abolfazli, A. Gani and R. Buyya, 
"Heterogeneity in Mobile Cloud Computing: 
Taxonomy and Open Challenges," IEEE 
Communications Surveys & Tutorials, vol. 16(1), 
pp. 369-392, Feb 2014. 

5. S. Abolfazli, Z. Sanaei, E. Ahmed, A. Gani and R. 
Buyya, "CloudBased Augmentation for Mobile 
Devices: Motivation, Taxonomies, and Open 
Challenges," IEEE Communication Surveys & 
Tutorials, vol. 16 (1), pp. 337-368, February 2014. 

6. T. Lose and M. Thinyabe, "A Cloud Computung 
Platform to Augment Mobile Phone Use in 
Marginilized Rural Areas," SATNAC, Western 
Cape, 2012. 

7. M. Satyanarayanan, P. Bahl, R. Caceres and N. 
Davies, "The Case for VM-Based Cloudlets in 
Mobile Computing," IEEE Pervasive Computing,, 
vol. 8 (4), pp. 14-23, October 2009. 

8. S-S. Qureshi , T. Ahmad, K. Rafique and Shuja- 
ul-islam, "Mobile cloud computing as future for 
mobile applications - Implementation methods 
and challenging issues," IEEE Int. Conf. Cloud 
Computing and Intelligence Systems (CCIS), pp. 
467-471, Beijing 2011. 

9. A. Ravi and S-K. Peddoju, "Energy Efficient 
Seamless Service Provisioning in Mobile Cloud 
Computing," IEEE 7th Int. Symp. Service- 
Oriented System Engineering, pp. 463-471, 
Redwood City, March 2013. 

10. J. Li, B. Li, T. Wo, C. Hu, J. Huai, L. Liu and K- 
P. Lam, CyberGuarder: A Virtualization Security 


@ IJTSRD | AvailableOnline@www.ijtsrd.coml Volume-2 | Issue-1 |Nov-Dec2017 


Page: 634 




























International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470 


Assurance Architecture for Green Cloud 
Computing, Future Generation Computer 
Systems, Elsevier Science, Vol. 28(2), pp 379- 
390, February 2012. 

11. H. Al-Aqrabi, L. Liu, R. Hill, and N. 
Antonopoulos, Cloud BI: Future of Business 
Intelligence in the Cloud, Journal of Computer 
and System Sciences, Elsevier, Vol. 81 (1), pp. 
85-96. February 2015. 

12. B. Li, J. Li and L. Liu, CloudMon: A Resource- 
Efficient IaaS Cloud Monitoring System Based on 
NIDS Virtual Appliances, Concurrency and 
Computation: Practice and Experience, Vol. 27(8), 
Wiley, June 2015, pp 1861-1885. 

13. W. Lee J. Jung, and H. Kim, "Analyzing Extent 
and Influence of Mobile Device’s Participation in 


Mobile Cloud Computing," Int. Conf. ICT 
Convergence (ICTC), pp. 14-16, Jeju, October 
2013. 

14. X. Liu, C. Yuan, Z. Yang, and Z. Hu, "An Energy 
Saving Algorithm based on User-Provided 
Resources in Mobile Cloud Computing," IEEE 
78th Vehicular Technology Conference (VTC 
Fall), pp. 1-5, Las Vegas, Spet 2013. 

15. M. Nir, A. Matrawy and M. St-Hilaire, "An 
Energy Optimizing Scheduler for Mobile Cloud 
Computing Environments," IEEE Conf. on 
Computer Communications Workshops 
(INFOCOM WKSHPS), pp. 404-409, Toronto, 
May 2014. 


@ IJTSRD | AvailableOnline@www.ijtsrd.coml Volume-2 | Issue-1 |Nov-Dec2017 


Page: 635