Effective task scheduling algorithm with deep learning for Internet of Health Things (IoHT) in sustainable smart cities

• Published in: Sustainable cities and society Vol. 71; p. 102945
• Main Authors: Nagarajan, Senthil Murugan, Deverajan, Ganesh Gopal, Chatterjee, Puspita, Alnumay, Waleed, Ghosh, Uttam
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01-08-2021
• Subjects: Deep learning; Fog computing; Health data analysis; Internet of Health Thing (IoHT); Sustainable; Task scheduling; Deep learning; Fog computing; Task scheduling; Sustainable; Health data analysis; Internet of Health Thing (IoHT)
• ISSN: 2210-6707; 2210-6715
• DOI: 10.1016/j.scs.2021.102945

•Proposed novel IoT based FoG assisted cloud network architecture.•Proposed task priority and load balancing algorithms.•Wearable sensor devices are used for collecting the health related data.•Real-time analysis in smart sustainable environment.
In the recent years, important key factor for urban planning is to analyze the sustainability and its functionality towards smart cities. Presently, many researchers employ the conservative machine learning based analysis but those are not appropriate for IoT based health data analysis because of their physical feature extraction and low accuracy. In this paper, we propose remote health monitoring and data analysis by integrating IoT and deep learning concepts. We proposed novel IoT based FoG assisted cloud network architecture that accumulates real-time health care data from patients via several medical IoT sensor networks, these data are analyzed using a deep learning algorithm deployed at Fog based Healthcare Platform. Furthermore, the proposed methodology is applied to the sustainable smart cities to evaluate the process for real-time. The proposed framework not only analyses the healthcare data but also provides immediate relief measures to the patient facing critical conditions and needs immediate consultancy of doctor. Performance is measure in terms of accuracy, precision and sensitivity of the proposed DHNN with task scheduling algorithm and it is obtained 97.6%, 97.9%, and 94.9%. While accuracy, precision and sensitivity for deep CNN is 96.5%, 97.5% and 94% and for Deep auto-encoder is 92%, 91%, and 82.5%.