Real-Time High-Speed High Dimension Data Streaming and Feature Extraction on Edge Computing Devices in Industrial Internet of Things (IIoT)
Year: 2023
Type: article
Abstract: Industrial Internet of Things (IIoT) and Artificial Internet of Things (AIoT) attracted technologist due to its economical impact and advancement in prediction and decision making using collected data. Data collection and data processing in the IIoT and AIoT is experiencing its limitations in computational devices and storage devices. IIoT and AIoT pumps large amount of data to the network in the form of high dimension and high speed, the pumped data is collected processed and feature extracted. Key Challenge of the collected data is to extract insights (Features) about it, huge amount of data collection and data processing has become a most interesting and open problem statement for both Industrialist and Researchers. IIoT system is becoming a primary area for data mining research, most sophisticated data analysis tools are required to understand the stress of data coming for the IIoT sensors and the devices connected to the network. More the devices connected to the network, more the data we need to process at a given time. This makes the data mining more complex in the IIoT environment. As the data is collected for longer duration, there is huge chance of data drift; this is one of the core issues in the IIoT data streaming and mining system. This paper proposes an efficient data mining and feature extraction technique for IIoT, The proposed technique reduce the computation significantly and increase the feature extractions on Edge computing. The author proposes a frame work for IIoT data processing named as Information Flow of IIoT (IFIIoT). The Proposed Framework is assessed using both synthetic and real-world data, encompassing various streaming speeds and data drift scenarios. The evaluation takes into account the overall performance of existing state-of-the-art algorithms in the literature, while the generated data provides valuable insights into the clustering system. Consequently, the proposed framework demonstrates superior performance compared to the current state-of-the-art algorithms described in the literature.
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W7217640634
