Mobility-Aware Application Scheduling in Fog Computing

TY - GEN

T1 - Mobility-Aware Application Scheduling in Fog Computing

AU - Bittencourt, Luiz F.

AU - Diaz-Montes, Javier

AU - Buyya, Rajkumar

AU - Rana, Omer F.

AU - Parashar, Manish

N1 - Publisher Copyright: © 2014 IEEE.

PY - 2017

Y1 - 2017

N2 - Fog computing provides a distributed infrastructure at the edges of the network, resulting in low-latency access and faster response to application requests when compared to centralized clouds. With this new level of computing capacity introduced between users and the data center-based clouds, new forms of resource allocation and management can be developed to take advantage of the Fog infrastructure. A wide range of applications with different requirements run on end-user devices, and with the popularity of cloud computing many of them rely on remote processing or storage. As clouds are primarily delivered through centralized data centers, such remote processing/storage usually takes place at a single location that hosts user applications and data. The distributed capacity provided by Fog computing allows execution and storage to be performed at different locations. The combination of distributed capacity, the range and types of user applications, and the mobility of smart devices require resource management and scheduling strategies that takes into account these factors altogether. We analyze the scheduling problem in Fog computing, focusing on how user mobility can influence application performance and how three different scheduling policies, namely concurrent, FCFS, and delay-priority, can be used to improve execution based on application characteristics.

AB - Fog computing provides a distributed infrastructure at the edges of the network, resulting in low-latency access and faster response to application requests when compared to centralized clouds. With this new level of computing capacity introduced between users and the data center-based clouds, new forms of resource allocation and management can be developed to take advantage of the Fog infrastructure. A wide range of applications with different requirements run on end-user devices, and with the popularity of cloud computing many of them rely on remote processing or storage. As clouds are primarily delivered through centralized data centers, such remote processing/storage usually takes place at a single location that hosts user applications and data. The distributed capacity provided by Fog computing allows execution and storage to be performed at different locations. The combination of distributed capacity, the range and types of user applications, and the mobility of smart devices require resource management and scheduling strategies that takes into account these factors altogether. We analyze the scheduling problem in Fog computing, focusing on how user mobility can influence application performance and how three different scheduling policies, namely concurrent, FCFS, and delay-priority, can be used to improve execution based on application characteristics.

KW - Middleware

KW - distributed computing

KW - distributed management

KW - scheduling algorithms

UR - http://www.scopus.com/inward/record.url?scp=85019246676&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85019246676&partnerID=8YFLogxK

U2 - 10.1109/MCC.2017.27

DO - 10.1109/MCC.2017.27

M3 - Article

SN - 2325-6095

VL - 4

SP - 26

EP - 35

JO - IEEE Cloud Computing

JF - IEEE Cloud Computing

ER -