Architecture and Algorithm for IoT Autonomic Network Management
Luis Zavala, Alfonso Ordonez Garcia, Mario Siller · 2019
The Internet of Things focuses on the configuration, interaction and connectivity of physical objects which are not commonly connected to the Internet. This opens challenges such as interoperability between devices (objects), data collection, management, among others. Nowadays, the number of IoT devices amounts up to billions and from a diversity of applications domains challenging architectural designs in terms of interoperability, scalability, supported services and functionality, etc. We believe that this can be addressed by a transit from current semi-autonomic properties present in IoT systems to full autonomic ones. In this paper we propose new architectural and functional blocks to the IoT model to enable the autonomic behavior. We focus on the Self-Star* autonomic property of self-configuration. This property is key in IoT systems management because of the scale and diversity involved. We propose a self-management algorithm in the IoT domain of urban agriculture based on Utility Theory. Simulation results show that this approach is suitable for such IoT application domain in which low algorithmic complexity is desired given the limitations of low cost and processing capacity.