22 February 2023 in IEEE Transactions on Vehicular Technology
In this paper, we investigate the joint impact of task priority and mobile computing service on the mobile edge computing (MEC) networks, in which one unmanned aerial vehicle (UAV) provides mobile computing service to help compute the tasks from users in multiple hotspots where the task priority is time-varying. For such a system, we firstly measure the system performance by the computing utility of multiples users, where the effect of a wide-range task priority is incorporated. We then analyze the impact of the network wireless bandwidth and UAV computational capability on the system performance, from which we optimize the system through UAV hotspot selection and user task offloading. To solve the optimization problem, we further employ deep Q-learning algorithm to learn an effective solution by continuous interaction between the UAV agent and system environment. Simulations are finally conducted to verify the superiority of the proposed studies in this paper.
11 May 2023 in IEEE Transactions on Vehicular Technology
In this paper, we investigate a multiuser cache-enabled vehicular mobile edge computing (MEC) network, where one edge server (ES) has some caching and computing capabilities to assist the task computing from the vehicular users. The introduce of caching into the MEC network significantly affects the system performance such as the latency, energy consumption and profit at the ES, which imposes a critical challenge on the system design and optimization. To solve this challenge, we firstly design the vehicular MEC network in a non-competitive environment by maximizing the profit of the ES with a predetermined threshold of user QoE, and jointly exploit the caching and computing resources in the network. We then model the optimization problem into a binary integer programming problem, and adopt the cross entropy (CE) method to obtain the effective offloading and caching decision with a low complexity. Simulation results are finally presented to verify that the proposed scheme can achieve the near optimal performance of the conventional branch and bound (BnB) scheme, while sharply reduce the computational complexity compared to the BnB.
2 November 2021 in Cluster Computing
5 August 2024 in IEEE Communications Letters
This letter investigates a video streaming system enabled by mobile edge computing (MEC), where we aim to maximize the user quality of experience (QoE) under dynamic mobile environments, through user association, bitrate adaptation and bandwidth allocation. To tackle this hybrid optimization problem, we first decompose it into discrete inter-frame and continuous intra-frame problems. In further, an in-depth analysis of intra-frame bandwidth allocation is conducted, guiding the design of a modified equal bandwidth allocation (MEBA) algorithm. Furthermore, we adopt a deep reinforcement learning (DRL) approach to cope with inter-frame user association and bitrate adaptation in dynamic mobile environments, while embedding MEBA in the DRL algorithm to jointly tackle the hybrid problem. Simulations are finally conducted to verify the superiority of the proposed method under the dynamic environment.
25 February 2026 in IEEE Transactions on Network and Service Management
With the rapid evolution of 5G/6G networks, deterministic services such as the Industrial Internet, vehicular communications, and immersive applications impose stringent end-to-end (E2E) latency guarantees. Efficient resource allocation for E2E network slicing (NS) is therefore required to simultaneously satisfy strict latency constraints and improve overall resource efficiency. However, orchestrating cross-domain and multidimensional resources across the radio access network (RAN) and core network (CN) remains challenging due to strong coupling and scalability issues. To address this problem, this paper develops a profit-driven optimization framework for static E2E NS, which jointly accounts for deterministic service guarantees and system profitability. On the RAN side, a hybrid genetic algorithm (HGA) is developed to preserve the global exploration capability of genetic algorithms, while incorporating a gradient-guided resource scaling (GGRS) operator to refine elite solutions and improve solution quality. On the CN side, we design a capacity-aware workload balancing algorithm (CA-WLB). For each candidate path, CA-WLB couples capacity adaptive virtual network function (VNF) placement with demand responsive CPU allocation to jointly optimize latency and resource consumption, and then selects the path that yields the maximum profit. Extensive simulations across a wide range of scenarios demonstrate that the proposed framework consistently outperforms conventional heuristic and meta-heuristic approaches, while achieving performance close to high-complexity optimization benchmarks with significantly reduced computational overhead.