Mobile edge computing (MEC) and cloud computing (CC) have been considered as the key technologies to improve the task processing efficiency for Internet of Vehicles (IoV). In this article, we consider a random traffic flow and dynamic network environment scenario where MEC and CC are collaborated for processing delay-sensitive and computation-intensive tasks in IoV. We study the joint optimization of computation offloading and resource allocation (CORA) with the objective of minimizing the system cost of processing tasks subject to the processing delay and transmission rate constraints. To attack the challenges brought by the dynamic environment, we use the Markov decision process model for formulating the dynamic optimization problem, and apply a deep reinforcement learning (DRL) technique to deal with high-dimensional and continuous states and action spaces. Then, we design a CORA algorithm, which is able to effectively learn the optimal scheme by adapting to the network dynamics. Extensive simulation experiments are conducted, in which we compare the CORA algorithm with both non-DRL algorithms and DRL algorithms. The experimental results show that the CORA algorithm outperforms others with excellent training convergence and performance in processing delay and processing cost.

Due to the limited computing resource and battery capability at the mobile devices, the computation-intensive tasks generated by mobile devices can be offloaded to edge servers or cloud for processing. In this paper, we study the multi-user task offloading problem in an end-edge-cloud system, in which all user devices compete for the limited communication and computing resources. Particularly, we first formulate the offloading problem with the goal of maximizing the Quality of Experience (QoE) of the users subject to resource constraints. Since each user focuses on maximizing its own QoE, we reformulate the problem as a Multi-User Task Offloading Game (MUTO-Game). We then identify an important property that for any device, both the communication interference and the degree of computing resource competition can be upper bounded. Based on the property, we further theoretically prove that there exists at least one Nash Equilibrium offloading strategy in the MUTO-Game. We propose the Game-based Decentralized Task Offloading (GDTO) approach to obtain the Nash Equilibrium offloading strategy. Finally, we analyze the upper bound for the convergence time and characterize the performance guarantee of the obtained offloading strategy for the worst case. A series of experimental results are presented, in comparison with both the centralized optimal approach and the approximate approaches.

In many remote areas lacking ground communication infrastructure support, such as wilderness, desert, ocean, etc., an integrated edge computing network in the air with edge computing nodes is an effective solution. It can provide over-the-air computing services for ground devices (GDs) with limited computing resources and battery life. In this paper, we study task offloading and resource allocation in the aerial-based mobile edge computing (MEC) system supported by a high altitude platform (HAP) and unmanned aerial vehicles (UAVs), with the goal of minimizing the GD's energy consumption. Considering that the task arrival of GDs and wireless communication quality are both stochastic and dynamic, we apply stochastic optimization techniques to transform this task offloading and resource allocation problem into two subproblems, i.e., 1) a subproblem for local computation resource allocation and 2) a subproblem for offloading resource allocation. For the first subproblem, we use convex optimization methods to address it. For the second subproblem, we use game theory to formulate the competition of offloading resources among GDs and propose the Distributed Game-theoretical Multi-server Selection (DGMS) algorithm and the Transmission Power Allocation (TPA) algorithm. Finally, we propose a Distributed Online Task Offloading and Resource Allocation (DOTORA) algorithm and give the theoretical performance analysis of the algorithm. We perform extensive experiments, including the comparison experiments with the UAV-Only and HAP-Only framework, and the comparison experiments with other algorithms under our HAP-UAV framework. The experimental results validate our proposed framework and the DOTORA algorithm.

With more computational intensive applications deployed involving mobile edge computing (MEC), the collaboration among mobile devices, edge and cloud servers becomes an effective mechanism to fully utilize all available distributed computing resources. However, two main challenges have yet to be addressed to enable this three-way collaboration for securing necessary computational resources and further guaranteeing the quality of service (QoS) of task handling. The first challenge is related to the partitioning of an application task into several dependent subtasks and schedule them among the collaborating device-edge-cloud (DEC). The second is focused on the allocation of necessary computing resources of device, edge and cloud servers for effective subtask handling. To this end, we study the joint task offloading and resource allocation for DEC collaboration in this paper by formulating a new optimization problem with the objective of minimizing the task handling latency. To solve this problem, we decompose the original problem into two subproblems, which include the first one of calculating the optimal task partitioning ratio by mathematical analytical method, as well as the second on using the Lagrangian dual (LD) method for obtaining the optimal task offloading and resource allocation policy. Finally, we conduct simulation experiments on a real-life dataset obtained from the central business district (CBD) of Melbourne, Australia, and the experimental results validate the efficacy of our approach in minimizing latency.

Mobile Edge Computing (MEC), as a highly promising technology, effectively processes computation-intensive tasks by offloading them to edge servers. Utilizing the advantages of Unmanned Aerial Vehicles (UAVs) in deployment flexibility and broad coverage, UAV-assisted edge computing can significantly enhance system efficiency. This paper studies a scenario where a UAV-MEC system serves multiple Mobile Users (MUs) with random task arrivals and movements. We minimize the energy consumption of MUs by jointly optimizing UAV trajectory and resource allocation for MUs subjected to the UAV energy limit. The problem is formulated as a multi-stage Mixed-Integer Nonlinear Programming (MINLP) problem. To address this, we propose an algorithm called JTORA integrated Deep Reinforcement Learning (DRL) and Lyapunov optimization techniques. Specifically, we initially transform the multi-stage MINLP problem into a deterministic optimization problem utilizing Lyapunov techniques and decompose the original problem into two sub-problems in parallel. Through DRL, we solve the first sub-problem of trajectory and communication resources optimization. For the second sub-problem involving computing resource allocation, convex optimization is employed to get the optimal solution. Theoretical analysis and experimental results demonstrate that the JTORA algorithm can effectively reduce the energy consumption of MUs while ensuring UAV endurance.