Energy-Efficient Delay-Aware Task Offloading in Fog-Cloud Computing System for IoT Sensor Applications

• Published in: Journal of network and systems management 2022, Vol.30 (1), Article 14
• Main Authors: Singh, Parvinder, Singh, Rajeshwar
• Format: Article
• Language: English
• Subjects: Algorithms; Cloud computing; Communications Engineering; Computation offloading; Computer Communication Networks; Computer Science; Computer Systems Organization and Communication Networks; Data processing; Delay; Depletion; Edge computing; Energy consumption; Information Systems and Communication Service; Intelligent and Trustworthy Internet Edge; Internet of Things; Networks; Operations Research/Decision Theory; Optimization; Servers; Smart sensors
• ISSN: 1064-7570; 1573-7705
• DOI: 10.1007/s10922-021-09622-8

Internet of Things (IoT) system comprises of many interrelated computing devices and smart sensors with limited battery, processing, and storage capabilities. Due to its nature and area of operation, IoT systems always work in a constrained environment, with battery depletion, hardware malfunction and harsh wireless channel conditions. For application processing, Cloud computing provided an absolute solution, but its viability is limited by the exorbitant costs and high transmission delays associated with it. In such a scenario, Fog computing is fast emerging as an attractive solution. It focuses on shifting the data processing activities to the edge of the network. However, fog computing has its own share of challenges that needs to be overcome for efficient and effective designs. The computational resources of fog server are so scarce that it cannot respond quickly to the high computational requirements that can cause an unacceptable queuing delay among the IoT applications. Hence, the optimal solution lies in the convergence of the two technologies. In this paper, the energy efficient and delay-aware task allocation problem in an IoT-fog-cloud system is investigated. We formulate a delay-based task allocation problem which suggests the optimal task allocation among local IoT devices, edge server and the cloud toward the minimum energy consumption and end to end delay. The problem is then solved using Energy-efficient task offloading strategy (EETOS) based on Levy-flight moth flame optimization (LMFO) algorithm. The EETOS reduces energy consumption by 22%, 25% and 29% in comparison to the Online Job Dispatching (OJD), Multi-tier Fog Computing (MFC) and Computation Offloading Game (COG) algorithms, respectively.