Skoči na glavni sadržaj

Prethodno priopćenje

https://doi.org/10.17559/TV-20221207032927

A Novel Task of Loading and Computing Resource Scheduling Strategy in Internet of Vehicles Based on Dynamic Greedy Algorithm

Huiyong Li orcid id orcid.org/0000-0003-0487-9078 ; School of Intelligent Manufacturing and Information, Jiangsu Shipping College, Nantong 226010, China
Shuhe Han ; School of Intelligent Manufacturing and Information, Jiangsu Shipping College, Nantong 226010, China
Xiaofeng Wu ; School of Intelligent Manufacturing and Information, Jiangsu Shipping College, Nantong 226010, China
Furong Wang ; School of Intelligent Manufacturing and Information, Jiangsu Shipping College, Nantong 226010, China


Puni tekst: engleski pdf 720 Kb

str. 1298-1307

preuzimanja: 349

citiraj


Sažetak

Focus on the scheduling problem of distributed computing tasks in Internet of Vehicles. Firstly, based on the computing-aware network theory, a distributed computing resource model of the Internet of Vehicles is established, and the seven-dimensional QoS attributes of the computing resources in the Internet of Vehicles (reliability between computing resources, communication costs, computing speed and computing costs of the computing resources themselves , computing energy consumption, computing stability, and computing success rate) are grouped and transformed into two-dimensional comprehensive attribute priorities: computing performance priority and communication performance priority. Secondly, the weighted directed acyclic graph model of distributed computing tasks in the Internet of Vehicles and the seven-dimensional QoS attribute weighted undirected topology graph model of distributed computing resources in the Internet of Vehicles are respectively established. Moreover, a dynamic greedy algorithm-based task of loading and computing resource scheduling algorithm is proposed. Finally, the example analysis shows that the overall performance of this dynamic greedy algorithm-based task of loading and computing resource scheduling algorithm is better than the classic HEFT scheduling algorithm and round robin scheduling algorithm.

Ključne riječi

computing-aware networks; edge computing; fog computing; greedy algorithm; internet of vehicles; scheduling algorithm

Hrčak ID:

305496

URI

https://hrcak.srce.hr/305496

Datum izdavanja:

28.6.2023.

Posjeta: 822 *