Placement of IoT Microservices in Fog Computing Systems: A Comparison of Heuristics
Abstract
Fog computing has been recognized as a promising approach to support modern IoT applications based on microservices, where geographically distributed sensors or mobile end users act as data sources. Relying purely on cloud computing introduces non-negligible latency between data sources and distant data centers — an issue for real-time, latency-sensitive IoT applications.
Placing tasks such as preprocessing or data aggregation in a layer of fog nodes close to the sensors can decrease response times and reduce traffic towards cloud data centers. However, the fog scenario is far more complex and heterogeneous than a cloud data center, so placing microservices efficiently over distributed fog nodes requires novel solutions.
In this paper we propose and compare different heuristics for placing application microservices over the nodes of a fog infrastructure, testing their ability to minimize application response times and satisfy Service Level Agreements across a wide set of operating conditions.