A Scalable and Adaptive Tasks Orchestration Platform for IoT

Edge computing is an emerging computing paradigm that aims to solve the cloud limitations by bringing its applications closer to the Internet of Things (IoT) devices. Thanks to its horizontal scalability, this paradigm leverages from the rapid growth of connected devices and makes it in its favor. As a result, it improves the scalability and reduces the latency. However, the adoption of this paradigm alone does not guarantee to meet the quality of service (QoS). Due to the heterogeneity of those devices and their requirements, the QoS is more influenced by the nature of the devices that are responsible for offloading the task, and by their location, which complicates the offloading process. To address this issue, in this paper, we present a tasks orchestration platform for IoT. It focuses on the role of edge computing in order to guarantee a high scalability and enable the self-capabilities of IoT. We also present a tasks orchestration algorithm that is based on Fuzzy Decision Tree. It leverages from reinforcement learning which enables it to adapt to the unpredictable environmental changes. As opposed to the existing solutions, the proposed architecture has provided more scalability and low delays regardless of the number of devices. On the other hand, the proposed algorithm has reduced the tasks completion delay by nearly 32%, the energy consumption by 52%, and the failure rate by 44%, as compared to the state of the art algorithm.