HINC - Harmonizing Diverse Resource Information across IoT, Network Functions, and Clouds

Abstract

Effective resource management in IoT systems must represent IoT resources, edge-to-cloud network capabilities, and cloud resources at a high-level, while being able to link to diverse low-level types of IoT devices, network functions, and cloud computing infrastructures. Hence resource management in such a context demands a highly distributed and extensible approach, which allows us to integrate and provision IoT, network functions, and cloud resources from various providers. In this paper, we address this crucial research issue. We first present a high-level information model for virtualized IoT, network functions and cloud resource modeling, which also incorporates software-defined gateways, network slicing and data centers. This model is used to glue various low-level resource models from different types of infrastructures in a distributed manner to capture sets of resources spanning across different sub-networks. We then develop a set of utilities and a middleware to support the integration of information about distributed resources from various sources. We present a proof of concept prototype with various experiments to illustrate how various tasks in IoT cloud systems can be simplified as well as to evaluate the performance of our framework.