Unified data management in collaborative IoT systems

Abstract

Data has become a central focus for industries and businesses in delivering services. Furthermore, efficient collaboration solutions for IoT systems are increasingly needed to optimize IoT data usage across various domains. This work provides a comparative analysis between data models for handling heterogeneous IoT data. The analysis shows that the document data model outperforms its competitors with a throughput rate of 597.2/sec and an average end-to-end execution time of 0.54 seconds in CRUD operations on IoT data. Furthermore, the work introduces a data unification framework that standardizes IoT device messages sent to the MQTT broker, eliminating the need to read extensive documentation. The framework includes a data interpreter approach that unifies heterogeneous IoT data into a well-defined JSON document the sensing information along with the metadata for the corresponding IoT device, storing it in data modelling collections. The results show an optimization of 88% in the IoT devices profile more than state-of-the-art solutions like Iotivity and SensorML. Also, Interoperability tests have been carried out in the area of smart intersections and their collaboration with weather and healthcare systems. The tests confirm the framework’s versatility and reliability thus proving its suitability for adoption by the developing community.