Anomalous data detection in cognitive IoT sensor network

Recent research in the internet of things (IoT) focuses on the insertion of cognition into its system architecture and design, which introduces the new discipline known as cognitive IoT (CIoT). The cognitive internet of things sensor network defines a new paradigm for bridging the gap between the virtual and the real world. Sensors integrated into the CIoT network serve as the primary data collectors. These sensors are used in hazardous or unmanaged a situation, which makes sensor readings prone to errors and abnormalities. Since sensor data are essential to the system's operation, the quality of various data-centric CIoT services will ultimately depend on the accuracy of sensor readings. However, detecting anomalies in sensor data is a complex process because CIoT sensor networks are frequently resource-constrained devices with limited computation, networking, and storage power. To fulfil the objectives, an effective and affordable cognitively-inspired detecting method is required. Therefore, this research proposed a novel technique to identify the anomaly in sensor node data. The experimental evaluation is conducted on the environmental data of 21.25 years, and detection accuracy reveals the efficacy of the proposed method over competing approaches.