关键词： Streaming media   Rendering (computer graphics)   Servers   Energy consumption   Wireless communication   Resource management   Quality of experience   Multiuser virtual reality   rate splitting   mobile edge computing   precoding design   resource allocation  

摘要： With the growing demand for virtual reality (VR) applications, mobile wireless networks should support the connections of a massive number of VR users. To support ultra-high data rates of multiple simultaneously transmitted VR streamings, we propose a mobile edge computing (MEC)-assisted rate splitting (RS) VR streaming transmission system. In the proposed system, RS technology exploits the shared interests of multiple VR users and MEC offloads the partial rendering tasks to achieve a better quality of experience (QoE) for VR users. Aiming to minimize the total weighted energy consumption, we formulate a joint communication and computing resource optimization problem while ensuring the required distortion and latency of VR users. To deal with the formulated intractable problem, we propose a joint rendering offloading and resource allocation algorithm that alternately solves the subproblems of quantization parameters selection, rendering offloading decision, transmit precoding design, rate allocation of RS transmission, and computing resource allocation. The simulation results demonstrate the effectiveness of the proposed algorithm in saving energy consumption. Specially, the performance of the proposed algorithm is 22.1% higher than that of the multicast-unicast scheme and can achieve 95.9% of the exhaustive search based algorithm.

关键词： Mobile edge computing   unmanned aerial vehicle   computation offloading   computation offloading   resource allocation   resource allocation   trajectory scheduling   trajectory scheduling   trajectory scheduling  

摘要： In Mobile Edge Computing (MEC), Unmanned Aerial Vehicles (UAVs) equipped with communication and computation capabilities can be used as an edge node, which can not only satisfy the user's demand for high computation power and low latency, but also extend the range of computation services and enhance the mission quality in environments with limited communication facilities. In this study, we investigate a UAV-assisted MEC system with stochastic computing tasks. The system seeks to minimize overall energy consumption by optimizing computation offloading, resource allocation, and the UAV's trajectory scheduling. This objective corresponds to a stochastic optimization problem. Given the problem's non-convex nature and the temporal coupling of variables, the Lyapunov optimization method is employed to analyze the task queue, breaking down the original optimization problem into three independent and more manageable subproblems. A joint optimization algorithm with iterative solving is proposed for solving the three sub-problems and then obtaining the stochastic computation offloading, resource allocation, and trajectory scheduling strategy. The simulation results show that the proposed strategy is able to achieve an effective compromise between the system's energy consumption and queue stability by adjusting the Lyapunov parameters, and significantly reduce energy consumption compared to the baseline strategies.

关键词： Mobile edge computing   SDN   Edge caching   visible light communication   millimeter Wave  

摘要： The integration of 5G networks has revolutionized wireless communication with higher data rates and ultra-low delay, enabling new services and applications, especially in mobile edge computing (MEC). MEC aims to minimize delay and enhance user experience by placing computational power and storage capabilities closer to end-users. A key challenge in MEC is efficiently managing network resources, particularly in environments with hybrid Radio Access Technologies (RATs). This paper explores a collaborative video caching strategy focused on delay and energy efficiency for software-defined hybrid Visible Light Communication (VLC) and millimeter Wave (mmWave) networks. These technologies offer high data rates but face challenges like varying channel conditions and blockages. The paper introduces a framework using Software-Defined Networking (SDN) for dynamic resource management and optimized content caching. It integrates a Long Short-Term Memory (LSTM) network to forecast user preferences and use a collaborative communication framework to improve edge caching. Optimization problems are formulated to reduce delay and energy consumption, considering limited storage at small-cell base stations (SBSs). The proposed algorithm significantly improves delay, energy consumption, and cache hit rates.

关键词： Deep Q-Networks   Graph Convolutional Networks   Mobile Edge Computing   performance consistency across networks   Service Function Chains   traffic steering optimization  

摘要： With the rise of applications such as the Internet of Things (IoT) and Virtual Reality (VR), there is an increasing demand for stringent service latency and quality of service requirements, which has led to a shift in network service deployment from cloud to edge, giving rise to Mobile Edge Computing (MEC) architectures. In MEC environments, network infrastructure is distributed near users, allowing access to local networks in real time. However, dynamically orchestrating Service Function Chains (SFCs) presents a significant challenge, especially in resource-constrained settings where maximizing SFC deployments while maintaining low latency is essential for service providers’ revenue optimization. To address this challenge, this paper proposes an intelligent SFC orchestration strategy, termed GCN-DQN, which combines Graph Convolutional Networks (GCNs) and Deep Q-Networks (DQNs). The GCN-DQN framework is designed to optimize the request acceptance rate while ensuring compliance with stringent low-latency requirements. To achieve this, the GCN-DQN strategy is designed to perceive network structure, resource availability, and SFC-specific information, enabling optimized decision-making for SFC traffic steering paths. Performance evaluations demonstrate that the proposed algorithm outperforms existing methods in SFC request acceptance rates under various network loads. © 2025 The authors. This article is published by IIETA and is licensed under the CC BY 4.0 license (http://***/licenses/by/4.0/).

关键词： Smart agriculture   Resource management   Autonomous aerial vehicles   Internet of Things   Blockchains   Optimization   Real-time systems   Delays   Dynamic scheduling   Data collection   Mobile edge computing   blockchain   reinforcement learning   unmanned aerial vehicles (UAV)   Internet of Things (IoT)   smart farms  

摘要： Unmanned aerial vehicle-to-everything (U2X) communications have become increasingly essential technologies to support multiple delay and computation-sensitive services, including augmented reality (AR), Virtual Reality (VR), digital twins, smart farms, unknown geological exploration, smart manufacturing, etc. Therefore, it is necessary to propose effective privacy-preserving communication and computation optimisation algorithm schemes to improve the quality of service (QoS) of multiple UAV-assisted Internet of Things (IoT) systems. In recent years, blockchain, U2X communications, and mobile edge computing will gradually become the main technologies of future academia and industry to overcome the aforementioned challenges for the UAV-assisted IoT system networks, while guaranteeing user data collection, data processing, and privacy-preserving requirements. In this paper, we consider a UAV-assisted IoT smart farm network systems, where blockchain technology is applied at each UAV server to guarantee real-time user data collection, data processing, data encryption, system access authorisation, and immutability. To minimise the total service delay (which includes the transmission delay and processing delay) among the user equipment (UEs) and UAVs, we propose jointly optimising the communication resource allocation for all UEs and UAVs, the computation resource allocation at each UAV, and the transmission power. This is achieved by devising a Q-learning-based optimisation algorithm, which is called a Q-learning-based latency reduction optimisation algorithm (QLROA). Simulation results illustrate that our proposed algorithm outperforms the benchmarks in terms of the long-term average delay among all the UEs.

关键词： Mobile edge computing  

摘要： Service migration is crucial in mobile edge computing (MEC) to ensure seamless service provision as users move. Although some migration schemes have been proposed, they fail to efficiently support the migration of microservices in a directed acyclic graph (DAG)-based service across different edge servers, resulting in high service latency. This article focuses on the DAG-based service migration problem and proposes a collaborative microservice migration framework for MEC, aiming to minimize the service migration latency while efficiently distributing the migration workload across edge servers. We divide edge servers into clusters and formulate the DAG-based service migration problem as a two-stage optimization problem. In the first stage, a deep reinforcement-learning-based service premigration algorithm is developed to identify the optimal cluster of edge servers for hosting the migrated service. In the second stage, a microservice migration algorithm is devised, utilizing topological sorting and network flow techniques to further determine the target edge server for each microservice. Our design addresses the inherent dependencies among microservices within a DAG task and adapts well to dynamic network environments. Experimental results on real-world datasets demonstrate that our approach significantly reduces service migration latency. © 2014 IEEE.

关键词： Mobile edge computing  

摘要： Edge Intelligence (EI) has been instrumental in delivering real-time, localized services by leveraging the computational capabilities of edge networks. The integration of Large Language Models (LLMs) empowers EI to evolve into the next stage: Edge General Intelligence (EGI), enabling more adaptive and versatile applications that require advanced understanding and reasoning capabilities. However, systematic exploration in this area remains insufficient. This survey delineates the distinctions between EGI and traditional EI, categorizing LLM-empowered EGI into three conceptual systems: centralized, hybrid, and decentralized. For each system, we detail the framework designs and review existing implementations. Furthermore, we evaluate the performance and throughput of various Small Language Models (SLMs) that are more suitable for deployment on edge devices. This survey provides researchers with a comprehensive vision of EGI, offering insights into its vast potential and establishing a foundation for future advancements in this rapidly evolving field. © 1986-2012 IEEE.

关键词： Mobile edge computing  

摘要： Applying artificial intelligence (AI) to mobile networks can significantly enhance inference services, driving the rapid evolution towards smarter, more efficient networks. As next-generation mobile networks are deployed, they are expected to provide higher data rates and lower latency, which presents challenges for traditional large-model inference. These challenges necessitate that inference architectures and models quickly adapt to meet growing demands and handle increasingly complex tasks. To tackle these issues, the cloud-edge-end collaborative inference framework has emerged as a highly effective solution. This framework facilitates intelligent collaboration across the cloud, edge, and end devices, optimizing computational resources at all levels. Building on this approach, this article proposes a collaborative inference system specifically designed for mobile networks, addressing critical issues like computational bottlenecks and inference latency in traditional model deployments. The system integrates an innovative task offloading strategy and advanced model acceleration techniques, resulting in reduced inference latency. Experimental results show that the system not only significantly lowers latency but also maintains high-quality output. Finally, the article discusses future development trends in cloud-edge-end collaborative inference systems. © 1986-2012 IEEE.

关键词： Mobile edge computing  

摘要： The rapid development of the Internet of Vehicles (IoV) and sixth-generation (6G) technology has made traditional cloud computing architectures inadequate for vehicular networks. With ultra-low latency, high bandwidth, and massive connectivity, 6G networks provide essential support for task offloading in mobile edge computing. Task offloading transfers computational tasks from vehicles to edge base stations or cloud servers. However, traditional methods struggle to adapt to diverse user tasks and dynamic network environments. These limitations increase offloading delays. Digital twin (DT) offers real-time simulation to reflect system dynamics and address these challenges. In this paper, we propose a task offloading framework that combines DTs with deep reinforcement learning. We propose a multi-layer dynamic framework with an error-reward feedback mechanism to handle complexity, dynamic changes, and errors. This approach enables the conclusion of task offloading strategies that minimize latency. The framework includes a threshold-based warning mechanism to effectively manage edge base station loads. By monitoring load in real time, the system evaluates base station status and adjusts task allocation to maintain stability. We propose a multi-dimensional state encoder architecture, comprising environment encoders, base station state encoders, and task state encoders to effectively extract critical features. The proposed architecture enables the conclusion of task offloading strategies that minimize latency. Experimental results demonstrate that the proposed algorithm can accurately and quickly conclude the task offloading strategy to minimize energy consumption and latency. © 2014 IEEE.

关键词： Mobile edge computing  

摘要： Edge service caching can significantly mitigate latency and reduce communication and computing overhead by downloading and caching service (application) data from clouds. The freshness of cached service data is critical when providing satisfactory services to users, but has been overlooked in existing research efforts. In this paper, we study the online low-latency and fresh service provisioning in mobile edge computing (MEC) networks. Specifically, we jointly optimize the service caching, task offloading, and resource allocation. To solve the formulated joint online long-term optimization problem, we design a Lyapunov-based online framework that decouples the problem at temporal level into a series of per-time-slot subproblems. For each subproblem, we propose an online integrated optimization-deep reinforcement learning (OIODRL) method, which consists of an optimization stage and a learning stage. In the optimization stage of OIODRL, a quadratically constrained quadratic program (QCQP) transformation and a semidefinite relaxation (SDR) method are utilized;in the learning stage of OIODRL, a deep reinforcement learning (DRL) algorithm is applied. Extensive simulations show that the proposed OIODRL method achieves a near-optimal solution and outperforms benchmark methods. © 1972-2012 IEEE.