Intelligent Pricing Model for Task Offloading in Unmanned Aerial Vehicle Mounted Mobile Edge Computing for Vehicular Network Asrar Ahmed Baktayan, Student, IEEE, Ibrahim Ahmed Al-Baltah, and Abdul Azim Abd Ghani Abstract—In the fifth-generation (5G) cellular network, the Mobile Network Operator (MNO), and the Mobile Edge Computing (MEC) platform will play an important role in providing services to an increasing number of vehicles. Due to vehicle mobility and the rise of computation-intensive and delay- sensitive vehicular applications, it is challenging to achieve the rigorous latency and reliability requirements of vehicular communication. The MNO, with the MEC server mounted on an unmanned aerial vehicle (UAV), should make a profit by providing its computing services and capabilities to moving vehicles. This paper proposes the use of dynamic pricing for computation offloading in UAV-MEC for vehicles. The novelty of this paper is in how the price influences offloading demand and decides how to reduce network costs (delay and energy) while maximizing UAV operator revenue, but not the offloading benefits with the mobility of vehicles and UAV. The optimization problem is formulated as a Markov Decision Process (MDP). The MDP can be solved by the Deep Reinforcement Learning (DRL) algorithm, especially the Deep Deterministic Policy Gradient (DDPG). Extensive simulation results demonstrate that the proposed pricing model outperforms greedy by 26% and random by 51% in terms of delay. In terms of system utility, the proposed pricing model outperforms greedy only by 17%. In terms of server congestion, the proposed pricing model outperforms random by 19% and is almost the same as greedy. Index terms—Computation Offloading, Dynamic Price, System Utility, Deep Reinforcement Learning (DRL), Unmanned Aerial Vehicles (UAVs), MEC. I. INTRODUCTION The drones assist cellular networks to provide a quick, efficient, and cost-effective solution. The Mobile Edge Computing (MEC) server, assisted by Unmanned Aerial Vehicles (UAV), provides services to mobile users or vehicles if they are in the same cell as the UAV-MEC server [1]. Manuscript received November 3, 2021; revised February 2, 2022. Date of publication April 6, 2022. Date of current version April 6, 2022. The associate editor prof. Vladan Papić has been coordinating the review of this manuscript and approved it for publication. A. A. Baktayan and I. A. Al-Baltah are with the Department of Information Technology, FCIT, Sana'a University, Sana’a, Yemen (e-mails: asrar@yemenmobile.com.ye, albalta2020@gmail.com). A. A. Abd Ghani is with the Department of Software Engineering Information Systems, Faculty of Computer Science and Information Technology, Universiti Putra Malaysia, Selangor, Malaysia (e-mail: azim@upm.edu.my). Digital Object Identifier (DOI): 10.24138/jcomss-2021-0154 As long as the latency between the vehicle and the serving UAV-MEC server satisfies the service delay requirements, the UAV-MEC server that serves the vehicle cannot be too far away from the vehicle, i.e. the delay spread between them should be at a minimum. However, due to obstructions such as dense buildings, a lack of infrastructure in some zones, and even the mm-wave Line of Sight (LOS), the quality of the connection is difficult to ensure. As a result, due to their terrain-agnostic and flexible deployment capabilities, UAVs have become one of the primary techniques for establishing communication linkages between the two ends [2]. In actuality, the UAV with MEC is unable to manage the amount of data offloaded by vehicles, and the UAV-MEC is unable to observe the vehicle's offloading profit function. Furthermore, vehicles are hesitant to report profile functions as they contain sensitive information such as battery states [3]. Given that, various vehicles place varying negative values on delay, as some are more delay-sensitive than others [4]. Thus, Mobile Network Operators (MNOs) find themselves intertwined in this profit-generating sector with competition for cached content in shared storage controlled by numerous MNOs, as well as pricing concerns [5]. Furthermore, typical optimization approaches cannot effectively optimize the computing resources and energy consumption of the UAV- assisted MEC system due to its complexity [6]. Vehicular computation offloading is a well-received approach for executing vehicles' delay-sensitive and/or computing-intensive tasks. The response time of vehicular computation offloading can be reduced by employing MEC, which has a high processing capacity and brings these computation tasks closer to the end vehicles. Deep Reinforcement Learning (DRL), a significant extension of the Reinforcement Learning (RL) technique, has been linked to several complex online optimization problems involving enormous arrangement spaces. Traditional network optimization techniques, such as greedy linear programming and greedy search, meet the network's immediate needs, whereas RL algorithms filter the entire system by considering every possible scenario. Vehicles determine the best strategy for real-time allocation of resources for dynamic networks, such as wireless networks, where conditions change regularly [7]. DRL is a powerful sequential decision-making approach to maximize long-term framework execution in an unknowable JOURNAL OF COMMUNICATIONS SOFTWARE AND SYSTEMS, VOL. 18, NO. 2, JUNE 2022 111 1845-6421/06/2021-0154 © 2022 CCIS