Opportunistic job sharing for mobile cloud computing
Architecture and Programming Models for Mobile Cloud Computing
1. Architecture and Programming Model for Mobile
Cloud Computing
Brijendra Pratap Singh
Registration No. - 2011SW21
Motilal Nehru National Institute Of Technology Allahabad
Allahabad - 211004, India
November 9, 2012
Brijendra Pratap Singh Registration No. - 2011SW21 Architecture and Programming Model for Mobile Cloud Computin
2. Outline
Introduction
Architecture for mobile cloud computing
Programming model for mobile cloud computing
Semi-Markov decision process (SMDP) based mobile cloud
computing model
Performence of SMDP-based mobile cloud computing model
Brijendra Pratap Singh Registration No. - 2011SW21 Architecture and Programming Model for Mobile Cloud Computin
3. Introduction
Cloud computing
Cloud computing infrastructure models
Public, private, and hybrid cloud
Cloud computing service models
Infrastructure as a service (IaaS)
Platform as a service (PaaS)
Software as a service (SaaS)
Mobile cloud computing or mCloud
Brijendra Pratap Singh Registration No. - 2011SW21 Architecture and Programming Model for Mobile Cloud Computin
4. Architecture Of mobile cloud computing
1. A remote cloud server providing services to mobile devices
Mobile device is connected to the cloud through Internet
Brijendra Pratap Singh Registration No. - 2011SW21 Architecture and Programming Model for Mobile Cloud Computin
5. 2. Mobile devices act as a resource provider of the cloud
Brijendra Pratap Singh Registration No. - 2011SW21 Architecture and Programming Model for Mobile Cloud Computin
6. 3. A cloudlet is between the cloud and mobile device
Mobile device is connected to the cloudlet and the cloudlet is
connected to cloud
Brijendra Pratap Singh Registration No. - 2011SW21 Architecture and Programming Model for Mobile Cloud Computin
7. Cloudlet
Mobile device’s resource poverty address via cloudlet
A cloudlet is a resource rich computer or cluster
Simplify the challenges of meeting high bandwidth
Dynamic virtual machine synthesis on cloudlet
Brijendra Pratap Singh Registration No. - 2011SW21 Architecture and Programming Model for Mobile Cloud Computin
8. Programming model of mobile cloud computing
1. To offload computation of mobile devices, the communication
between the mobile device and cloud is done by the protocols such
as Remote Procedure Call (RPC), Remote Method Invocation
(RMI), etc
Example: Hyrax
Brijendra Pratap Singh Registration No. - 2011SW21 Architecture and Programming Model for Mobile Cloud Computin
9. Hyrax
Hadoop is used as a basis for Hyrax
Hadoop is an open source implementation of MapReduce
programming model and google file system
Figure: Execution overview of MapReduce programming model
Brijendra Pratap Singh Registration No. - 2011SW21 Architecture and Programming Model for Mobile Cloud Computin
10. Porting Hadoop to run on mobile device
Name node and job tracker run on cloud
Data node and task tracker run on mobile device
Brijendra Pratap Singh Registration No. - 2011SW21 Architecture and Programming Model for Mobile Cloud Computin
11. 2. To offload computation of mobile devices, virtual machine is
used as the computing unit
Example: Cloudlet, MAUI, CloneCloud, MobiCloud
Brijendra Pratap Singh Registration No. - 2011SW21 Architecture and Programming Model for Mobile Cloud Computin
12. CloneCloud
Automatically transforms a single machine execution into a
distributed executation
Brijendra Pratap Singh Registration No. - 2011SW21 Architecture and Programming Model for Mobile Cloud Computin
13. 3. To offload computation of mobile devices mobile agent is used
Example: Scavenger
Brijendra Pratap Singh Registration No. - 2011SW21 Architecture and Programming Model for Mobile Cloud Computin
14. Scavenger
Consist of two independent software component
The daemon running on surrogate enabling to receive and
perform task
Client library
Both written in Python
Client application must use the client library
A manual mode, where the application may itself ask for
available surrogate
Automatic mode
Brijendra Pratap Singh Registration No. - 2011SW21 Architecture and Programming Model for Mobile Cloud Computin
15. Semi-markov decision process (SMDP) based mobile cloud
computing model
A mobile cloud system has multiple cloud domains. There are K
virtual machine resource available in one cloud domain. A service
can use c virtual machine, where c ∈ {1, 2, ..., C }, C ≤ K
Brijendra Pratap Singh Registration No. - 2011SW21 Architecture and Programming Model for Mobile Cloud Computin
16. There are two types of service requests
New service request
Interdomain transfer service request
The system state of the cloud domain is defined by the
current services having different numbers of virtual machines
and event such as arrival or departure of services
The action can be accept, reject, or transfer upon receiving a
request
The reward of a mobile cloud system is calculated by system
state and corresponding action
Brijendra Pratap Singh Registration No. - 2011SW21 Architecture and Programming Model for Mobile Cloud Computin
17. Brijendra Pratap Singh Registration No. - 2011SW21 Architecture and Programming Model for Mobile Cloud Computin
18. Performence of SMDP-based mobile cloud computing
model
SMDP-based model have lower dropping probability than the
greedy approach
Expected system reward of SMDP-based system is better than
greedy scheme
Brijendra Pratap Singh Registration No. - 2011SW21 Architecture and Programming Model for Mobile Cloud Computin
19. References I
Niroshinie Fernando, Seng W. Loke, Wenny Rahayu, “Mobile
Cloud Computing: A survey,” Future Generation Computer
System 29(2013) 84-106, June 2012
J. Carolan, S. Gaede, J. Baty, G. Brunette, A. Litch, J.
Rammell, L. Tucker, J. Weise, “Introduction to Cloud
Computing Architectue,” white paper, 2009
Jeffrey Dean, Sanjay Ghemawat, “MapReduce: Simplified
Data Processing on Large Clusters,” Communications of ACM,
Junary 2008
Paramvir Bahl, Richard Y. Han, Li Erran Li, Mahadev
Satyanarayanan, “Advancing the State of Mobile Cloud
Computing,” Proceedings of the third ACM workshop on
Mobile cloud computing and services, June 2012
Brijendra Pratap Singh Registration No. - 2011SW21 Architecture and Programming Model for Mobile Cloud Computin
20. References II
Byung-Gon Chun, Sungwan Ihm, Petros Maniatis, Mayur
Naik, Ashwin Patti, “CloneCloud: Elastic Execution between
Mobile Device and Cloud,” In EuroSys 11: Proceedings of
sixth conference on computer system, ACM, 2011
E. E. Marinelli, “Hyrax: Cloud Computing on Mobile Devices
Using MapReduce,” Masters Thesis, Carnegie Mellon
University, 2009
M. Satyanarayanan, P. Bhal, R. Caceres, N. Davies, ”The Case
for VM-based Cloudlet in Mobile Computing,” IEEE Pervasive
Computing, 2009
Mads Daro Kristensen, “Scavenger: Transparent Development
of Efficient Cyber Foraging Applications,” Proceedings of the
IEEE International Conference, 2009
Brijendra Pratap Singh Registration No. - 2011SW21 Architecture and Programming Model for Mobile Cloud Computin
21. References III
Hongbin Liang, Lin X. Cai, Dijiang Huang, Xuemin (Sherman)
Shen, and Daiyuan Peng, “An SMDP-Based Service Model for
Interdomain Resource Allocation in Mobile Cloud Networks,”
IEEE Transactions on Vehicular Technology, June 2012
Brijendra Pratap Singh Registration No. - 2011SW21 Architecture and Programming Model for Mobile Cloud Computin