Mobility-Aware Controller Orchestration in Multi-Tier Service-Oriented Architecture for IoT

Abstract

This work addresses the problem of controller orchestration in Internet of Things (IoT)-based Service-Oriented Architecture. It focuses on the fundamental issues of heterogeneity and dynamism of hybrid edge-cloud networks involving both static and mobile IoT devices for provisioning IoT-based services. Due to the limited capacity of IoT devices, IoT-cloud services require the support of edge networks, termed as ‘Service Edge,’ for meeting their ultra-low latency requirements. In such complex and dynamic networks, software-defined networking (SDN) paradigm is utilized for efficient resource utilization and service-oriented network management. A significant problem in SDN is the controller placement problem. Although researchers proposed several controller placement schemes for SDN, the problem of multi-tier controller placement in hybrid cloud-edge networks has not yet been addressed. In this work, this problem is formulated mathematically and shown to be a variant of the well-known <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">two-level capacitated facility location problem</i> , which is NP-hard. Thereafter, COMET, a scheme based on coalitional game and social choice theory, is proposed to solve the problem in polynomial time. COMET is designed while considering varying loads on the switches due to the presence of mobile IoT devices. Simulations depict that COMET reduces network delay by 49.02% with a significant increase in resiliency compared to the state-of-the-art.