Heterogeneous Interoperable Sensors Integrating Cognitive Knowledge for IOT-Based Cross-Domain Applications

The Internet of Things (IoT) technology has enabled a proliferation of linked devices that produce a massive amount of heterogeneous data. Massive amounts of data from the IoT devices can cause interoperability issues, including issues with communication protocols, device compatibility, and open standards adoption. To overcome these drawbacks, a novel heterogeneous INTeroperable sensors (HintSense) technique has been proposed, which improves semantic interoperability by enabling deep knowledge in the IoT link layer. The proposed model integrates heterogeneous sensors, such as temperature and humidity sensors, that are connected with various operations and data management systems on different applications, including weather and healthcare. The sensor data will be organized in an identical form using the SenML model, and semantic modeling converts relational data into RDF format. The proposed model uses the Gannet optimization with Squeeze and Excitation-based ResNeXt (GO-SER) blocks to build deep knowledge and facilitate data reuse. By linking domain ontologies and datasets, an M3 reasoning engine makes data integration and transmission more efficient. The HintSense strategy is evaluated using the Cooja simulator based on metrics, such as accuracy (ACC), latency, throughput, and energy efficiency, and it yields better results than other strategies. The proposed HintSense achieves higher throughput of 33.33%, 20.00%, and 9.09% than the existing techniques, such as HeDI, fuzzy logic technique, and MDSS techniques, respectively. Providing real-time data interpretation and interoperability for the IoT ecosystems, this work offers scalable and effective solutions for weather monitoring and healthcare applications.