Sensor Self-Declaration of Numeric Data Reliability in Internet of Things

Abstract

Since diverse noises and irregularities impact on sensor data, self-declaration of sensor data reliability is crucial for advancing Internet of Things applications and industrial automation. Relevant works on reliability include sensor self-attribution of data confidence, and self-diagnosis of sensor faults using temporal data redundancy or neighboring sensor data. Models are built on edge devices and then transferred to sensors. Overall, the existing methods are computationally expensive, require real-time data from other sensors and incur considerable transmission overhead. Therefore, they are not suitable for independent sensor data reliability assessment. Addressing these issues, we introduce an independent reliability self-declaration method for sensors. Two Kalman filter-inspired, block-based lightweight algorithms are designed that handle isolated and burst noises and estimate block data reliability. Moreover, a conceptual model to dynamically adjust block size is proposed leveraging noise level and maximum TCP/IP packet size to reduce data transmissions. The reliability levels are conveyed using TCP header reserved bits to avoid communication overhead. The approach was tested using water quality monitoring (WQM) and healthcare application datasets. Results show, for burst noise, our lightweight and scalable approach attains superior accuracy in WQM (89.06%) and healthcare (82.63%) for five-level reliability estimation. A real-world deployment using an Arduino-based sensor node demonstrates the feasibility of the approach for in-sensor operation.