Fog-Server Placement Technique Based on Network Edge Area Traffic for a Fog-Computing Environment

Min-Sik Son, Sang-Hwa Chung, Won-Suk Kim

http://doi.org/10.5626/JOK.2018.45.6.598

Cloud computing is now widely used. However, cloud computing alone cannot adequately respond to the traffic patterns generated by various objects in the new network environment of the internet of things (IoT). Fog computing is one method for overcoming and solving the problems of cloud computing. A fog-server placement method is needed, including a mechanism for determining the location for a fog-server that can provide services. However, most studies only consider the fog device’s computing resources, and the locations of the client and the data sources are not considered; therefore, the network situation becomes worse after the fog-server placement. In this paper, we propose a technique for fog-server placement that considers traffic generation in relation to the locations of the clients and data sources. In the experiment, clients and data sources are concentrated or distributed in the network topology, and their corresponding network-traffic patterns are considered. Experimental results show that, in terms of reducing core network traffic and memory usage, placing the fog server according to the proposed network traffic conditions is more efficient than placing fog servers in all of the fog devices.