Energy-Delay Tradeoff in Parallel Task Allocation and Execution for Autonomous Platooning Applications

Abstract

Aiming to guarantee the Makespan of data-intensive tasks in autonomous platooning applications and improve the system energy utilization, we proposed a parallel task allocation and execution algorithm using divisible load theory for energy-delay tradeoff in autonomous platooning applications in this paper. Assuming that the autonomous vehicles in the platoon support the dynamic voltage and frequency scaling technology, the autonomous vehicles adjust the processing capacity and communication capacity of the vehicles to the optimal level without exceeding the maximum completion time based on dynamic programming theory, so as to minimize the total energy consumption of the autonomous platoon. In order to calculate the allocation strategy conveniently in large-scale autonomous platoon,the Markov chain model is used to analyze the parallel task allocation and execution process in autonomous platooning applications, which is equivalent to queuing model. Then the Little theorem is used to analyze the total Makespan in the parallel task allocation and execution process, which modelled as a serially divisible load scheduling process. The simulation results show that the proposed parallel task allocation and execution algorithm for autonomous platooning applications can adjust the voltage and frequency of the autonomous vehicles to the most suitable level under the premise of meeting the total Makespan constraint, so as to reduce the total energy consumption in the task allocation and execution process. At the same time, the equivalence of the parallel task allocation and execution process and Markov chain model is verified.