MSSAMTO-IoV: modified sparrow search algorithm for multi-hop task offloading for IoV

• Published in: The Journal of supercomputing Vol. 79; no. 18; pp. 20769 - 20789
• Main Authors: Alseid, Marya, El-Moursy, Ali A., Alfawaz, Oruba, Khedr, Ahmed M.
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.12.2023; Springer Nature B.V
• Subjects: Communication; Compilers; Computation offloading; Computer Science; Convergence; Edge computing; Efficiency; Genetic algorithms; Greedy algorithms; Internet of Things; Internet of Vehicles; Interpreters; Mobile computing; Network latency; Optimization; Processor Architectures; Programming Languages; Search algorithms; Vehicles; Wireless communications; Modified sparrow search algorithm; Multi-hop task offloading; Internet of vehicle; Mobile edge computing
• ISSN: 0920-8542; 1573-0484
• DOI: 10.1007/s11227-023-05446-2

Mobile edge computing (MEC) is an emerging technology that can be integrated with the Internet of vehicles (IoV) to enhance vehicle services with low latency and optimize task offloading efficiency. MEC technology enhances the battery life of mobile devices by bringing storage and computational services closer to the edge of the network. However, challenges arise in areas with limited MEC server coverage. To address these challenges, this paper proposes a multi-hop task offloading model in MEC-IoV utilizing the modified sparrow search algorithm (MSSAMTO-IoV). The MSSAMTO-IoV model consists of two phases: candidate vehicle selection and task offloading. In the candidate vehicle selection mechanism, the model takes into account the k-hop wireless communication range. It identifies and selects candidate vehicles from neighboring vehicles for task offloading. The task offloading problem is then formulated as an optimization problem, aiming to minimize the delay. To solve this problem, MSSA is utilized, with modifications introduced to enhance the initial population quality and diversity of the standard SSA using the logistic map. Furthermore, an inertia weight is introduced to improve search speed, convergence rate, and exploration capabilities. The mean termination criterion is employed to avoid unnecessary iterations and minimize run-time. Additionally, a mutation strategy is employed to prevent falling into local optima. The simulation results demonstrate that MSSAMTO-IoV achieves faster convergence and outperforms BAT and SSA by approximately 2 times and 3 times, respectively. Moreover, MSSAMTO-IoV effectively reduces latency across different task sizes. For a 500 KB task, it reduces latency by approximately 5, 9, and 13% compared to basic SSA, BAT, and greedy algorithms, respectively.