Coordination Models for Internet of Things

Abstract

In constrained environments, there is a variety of devices like sensors and actuators with limited computation power or energy that form an Internet of Things (IoT) system. When processing complex tasks is required, those devices send the data to the cloud and obtain the result later. However, the IoT system could process complex task if more devices work together, sharing computational resources and cooperating. This cooperation can be achieved using a coordination model that distributes the load among the different devices based on a set of parameters, laws and defined entities. This research implements and evaluates a data-oriented coordination model with three variations for Internet of Things (IoT). It also presents, implements and evaluates a new process-oriented coordination model that can make constrained environments much more effective and allow the processing of more complex tasks closer to the network. The development of all the coordination models was focused on using the system’s computational resources effectively. As IoT is a heterogeneous field, devices with more power can process more complex tasks, creating an uneven but adequate load distribution. Various experiments were conducted to evaluate the performance of each model using one and two workers. The results showed that every coordination model works effectively when distributing the load among more workers. For the process-oriented model, implementing some CoAP features allowed the system to perform better when repetitive tasks are required.