Authors:
Mayssa Trabelsi
;
Nadjib Bendaoud
and
Samir Ben Ahmed
Affiliation:
LIPSIC Laboratory, Faculty of Sciences of Tunis, University of Tunis El Manar, Tunis 2092, Tunisia
Keyword(s):
Internet of Things (IoT), Fog Computing, Service Placement, Real-Time Application, Quality-of-Service (QoS), iFogSim.
Abstract:
The Internet of Things (IoT) is a key technology that improves the connectivity between applications and devices over different geographical locations. However, IoT devices, particularly those used for monitoring, have stringent timing requirements that Cloud Computing might not be able to satisfy. Fog computing, which uses fog nodes close to IoT devices, can solve this problem. In this paper, we propose a Dynamic Service Placement (DSP) algorithm for Fog infrastructures. DSP’s objective is to dynamically place the services emitted by applications one at a time and in real time on Fog nodes. The algorithm chooses the fog node with the least response time over the infrastructure and dynamically places the incoming service in it. The algorithm is implemented in the iFogSim simulator, and its performances were evaluated and compared to other algorithms. DSP showed very encouraging results, as it proceeded to minimize the average response times and the application placement rate, thus lo
wering the infrastructure usage and energy consumption.
(More)