Research on Application of Auction Algorithm in Internet of Vehicles Task Scheduling under Fog Environment

Abstract

The demand for real-time computing services in the Internet of Vehicles (IOV) is becoming increasingly prominent. Aiming at the delay requirements of real-time computing, a task scheduling model based on M/M/b queuing theory is established with minimum network latency as the goal. When formulating a task scheduling strategy, a CPU resource pricing method that emphasizes network latency sensitivity is designed, where RAM memory, I/O interface bandwidth, hard disk capacity, and latency requirements are used as constraints for task scheduling, which can be solved based on the mechanism of Vickrey-Clark-Groves auction algorithm. Simulation and analysis results show that compared with integer linear programming and genetic algorithms, the auction algorithm is the optimal solution when solving the task scheduling strategy, and the complexity of algorithm is lower than that of the other two algorithms. The proposed auction algorithm can effectively reduce network delay.