Yavari Ali, Korala Harindu, Georgakopoulos Dimitrios, Kua Jonathan, Bagha Hamid
6G Research and Innovation Lab, Swinburne University of Technology, Melbourne, VIC 3122, Australia.
School of Science, Computing and Engineering Technologies, Swinburne University of Technology, Melbourne, VIC 3122, Australia.
Sensors (Basel). 2023 May 30;23(11):5211. doi: 10.3390/s23115211.
The Internet of Things (IoT) plays a fundamental role in monitoring applications; however, existing approaches relying on cloud and edge-based IoT data analysis encounter issues such as network delays and high costs, which can adversely impact time-sensitive applications. To address these challenges, this paper proposes an IoT framework called Sazgar IoT. Unlike existing solutions, Sazgar IoT leverages only IoT devices and IoT data analysis approximation techniques to meet the time-bounds of time-sensitive IoT applications. In this framework, the computing resources onboard the IoT devices are utilised to process the data analysis tasks of each time-sensitive IoT application. This eliminates the network delays associated with transferring large volumes of high-velocity IoT data to cloud or edge computers. To ensure that each task meets its application-specific time-bound and accuracy requirements, we employ approximation techniques for the data analysis tasks of time-sensitive IoT applications. These techniques take into account the available computing resources and optimise the processing accordingly. To evaluate the effectiveness of Sazgar IoT, experimental validation has been conducted. The results demonstrate that the framework successfully meets the time-bound and accuracy requirements of the COVID-19 citizen compliance monitoring application by effectively utilising the available IoT devices. The experimental validation further confirms that Sazgar IoT is an efficient and scalable solution for IoT data processing, addressing existing network delay issues for time-sensitive applications and significantly reducing the cost related to cloud and edge computing devices procurement, deployment, and maintenance.
物联网（IoT）在监控应用中发挥着基础性作用；然而，现有的依赖基于云和边缘的物联网数据分析方法存在网络延迟和成本高等问题，这可能会对时间敏感型应用产生不利影响。为应对这些挑战，本文提出了一种名为Sazgar IoT的物联网框架。与现有解决方案不同，Sazgar IoT仅利用物联网设备和物联网数据分析近似技术来满足时间敏感型物联网应用的时间限制。在此框架中，物联网设备上的计算资源用于处理每个时间敏感型物联网应用的数据分析任务。这消除了将大量高速物联网数据传输到云或边缘计算机所带来的网络延迟。为确保每个任务满足其特定应用的时间限制和准确性要求，我们对时间敏感型物联网应用的数据分析任务采用近似技术。这些技术会考虑可用的计算资源并相应地优化处理过程。为评估Sazgar IoT的有效性，已进行了实验验证。结果表明，该框架通过有效利用可用的物联网设备成功满足了COVID-19公民合规监测应用的时间限制和准确性要求。实验验证进一步证实，Sazgar IoT是一种用于物联网数据处理的高效且可扩展的解决方案，解决了时间敏感型应用中现有的网络延迟问题，并显著降低了与云及边缘计算设备采购、部署和维护相关的成本。
