An Autonomous Availability Computation Resource ... - IEEE Xplore

0 downloads 0 Views 101KB Size Report
computation resource allocation platform for Internet of Things in the Fog computing environment. Danilo Reis de Vasconcelos. MDCC - Mestrado e Doutorado ...
2015 International Conference on Distributed Computing in Sensor Systems

Smart Shadow - An autonomous availability computation resource allocation platform for Internet of Things in the Fog computing environment Danilo Reis de Vasconcelos

Rossana Maria de Castro Andrade

Jose Neuman de Souza

MDCC - Mestrado e Doutorado em Ciencia da Computacao UFC -Universidade Federal do Ceara Fortaleza, Ceara 30332–0250 Email: [email protected]

MDCC - Mestrado e Doutorado em Ciencia da Computacao UFC -Universidade Federal do Ceara Fortaleza, Ceara 30332–0250 Advisor Email: [email protected]

MDCC - Mestrado e Doutorado em Ciencia da Computacao UFC -Universidade Federal do Ceara Co-Advisor Email: [email protected]

Abstract—In the ecosystem of Internet of Things, many applications require external computational resources with quick response. In this context, a possible solution is to use computational resources from other devices that are located nearby the host device. This new environment is named Fog computing and creates several opportunities and challenges. One these challenges, which is the focus of this work, is how to allocate available resources in a simple and effective way for a client device that you want to use these features.The proposal ”Smart Shadow” is an autonomous platform availability allocation of computer resources in a Fog computing infrastructure based on learning from previous event to support mobile hosts devices.



How to restore information needed to resolve the main issue whether the framework of mist is computing Dynamics?



How to learn how past experiences of resource allocation considering the different scenarios that the device host this application experiencing?



How address the great heterogeneity of devices, wireless media, protocols, connectivity cost, restrictions on processing power, power consumption, connectivity media and location of the application host device?

Keywords—Internet of Things, Fog computing, web of things, resource allocation.

I.

III.

The work aims to propose a collaborative resource allocation algorithm of distributed computation resources in order to comply with the restrictions of smart devices available in Fog Computing; application host device and information learned from previous avaibility allocations. The main idea of the proposal is that the environment constituted by local fog computing devices learn to allocate avaibility of computational resources autonomously. The Learning is based on events of past allocations, availability of computational resources on local fog and goal of the host device application. The learning parameters are stored in local fog computing structure. The data will be stored in a distributed and should be able to restore when some component of the infrastructure is not available at the time and was part of the data.

I NTRODUCTION

The emergence of the Internet of things allows a large number of geographically distributed and intelligent devices with wireless connections. Many applications in this environment require a low response time or a low latency of the network, which can hardly be met with the conventional architecture of cloud computing. A new approach to solving this problem is to use computational resources at the edges of the network, what means providing resources to devices locally without many layers of communication. In this context, a new infrastructure platform called Fog computing uses computational resources of smart devices located at the edge of the network and close physically from the application execution environment. One of the challenges that emerges in this new architecture is how dynamically allocate in an autonomous way computational resources (processing, storage or even internet connectivity) available locally on the environment. II.

IV.

M AIN R ESEARCH Q UESTION

V. •

How to allocate the resources quickly without much consumption of computer resources?

978-1-4799-8856-3/15 $31.00 © 2015 IEEE DOI 10.1109/DCOSS.2015.25

O BJECTIVES

The main goal of the research is to propose algorithms and an architecture that allow mobile device hosts in a cloud computing environment, using computational resources in a simple and efficient way with low latency and short response time leveraging the uses of resources in Fog Computing.

The main research question is how to allocate in an autonomous way computational resources of a fog computing infrastructure to support applications of mobile devices. This question addresses several other issues such as: •

P ROPOSAL

216

M ETHODOLOGY

Initially search the State of the art in the area doing a bibliographic survey on the research issue.



Identify the main problems seeking divide them into simpler problems;



Propose solutions to the new arises problems;



Validate individually through simulation or experiment the proposed solutions;



Integrate individual solutions to compose the main research issue.



Validate through simulation the proposed solution;



Analyze the results to identify strengths and weaknesses;



Compare with similar solutions; VI.

R EFERENCES [1]

[2]

[3]

[4]

B ENEFITS

Search results can enable a better use of computational resources locations in fog computing, enabling the development of close loop control applications that normally require answer in short time. The work also aims to provide a computational resource infrastructure in Fog computing enabling a better performance in applications hosted on mobile devices allowing them to use external computational resources in a ubiquitous way. VII.

S HORT B IOGRAPHY

Danilo Reis de Vasconcelos - Graduated in Electronic Engineering from the Technological Institute of Aeronautics ITA(1988), specialist in economics from UFC- CAEN (1994) , master degree in applied informatics - UNIFOR (2006). Currently, he is a PhD student in the MDCC-UFC, working on the computer networks research area, specifically on Internet of things. Chief technology officer of Dwa Tecnology Import & Export Ltd. since 1991, professor of Unifor (University of Fortaleza) of the following disciplines: Advanced Digital Circuits, Electronic Circuits II, web technology, programming’s logic and embedded systems. He has extensive experience in embedded systems, language programming C / C ++ and Java, embedded operating systems, RTOS, smartcards, RFID, computational intelligence, pattern recognition, web programming with Java, object-oriented programming, design patterns, UML, projects with programmable Logic Devices (FPGAs ), Verilog and data security. VIII.

C URRENT RESEARCH INTERESTS

Internet of Thing, machine learning, Fog Computing, Embedded systems, crowd sensing. IX.

P LAN FOR FUTURE RESEARCH / CAREER

Manager R&D projects in Internet of Things with focus on smart appliance and home IoT. Participate in international cooperation projects in the area of Internet of things (IoT). X.

O BJECTIVES FOR PARTICIPATING IN DCOOS S TUDENT P ROGRAM



Build a relationship network with researchers;



Meet work that could be used in my research;



Search suggestions in my area of research;



Meet work related to my area of research;

217

1 -John A. Stankovic, Life Fellow, IEEE - Research Directions for the Internet of Things - IEEE INTERNET OF THINGS JOURNAL, VOL. 1, NO. 1, FEBRUARY 2014. 2 -Andrea Zanella, Senior Member, IEEE, Nicola Bui, Angelo Castellani, Lorenzo Vangelista, Senior Member, IEEE, and Michele Zorzi, Fellow, IEEEInternet of Things for Smart Cities- IEEE INTERNET OF THINGS JOURNAL, VOL. 1, NO. 1, FEBRUARY 2014. L. Wang and G. S. Kuo, Mathematical modeling for network selection in heterogeneous wireless networks - a tutorial, IEEE Communications Surveys and Tutorials, vol. 15, pp. 271292, 2013. bibitembib4:Xu Li Da Xu, Senior Member, IEEE, Wu He, and Shancang Li - Internet of Things in Industries: A Survey - IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, VOL. 10, NO. 4, NOVEMBER 2014 Ali Jose Mashtizadeh, Andrea Bittau, Yifeng Frank Huang, David Mazi eres - Replication, History, and Grafting in the Ori File System - Stanford University

Suggest Documents