关键词： mobile edge computing   dynamic matching   D2D   delay constraint   multi-agent reinforcement learning  

摘要： Device-to-device (D2D) is a pivotal technology in the next generation of communication, allowing for direct task offloading between mobile devices (MDs) to improve the efficient utilization of idle resources. This paper proposes a novel algorithm for dynamic task offloading between the active MDs and the idle MDs in a D2D-MEC (mobile edge computing) system by deploying multi-agent deep reinforcement learning (DRL) to minimize the long-term average delay of delay-sensitive tasks under deadline constraints. Our core innovation is a dynamic partitioning scheme for idle and active devices in the D2D-MEC system, accounting for stochastic task arrivals and multi-time-slot task execution, which has been insufficiently explored in the existing literature. We adopt a queue-based system to formulate a dynamic task offloading optimization problem. To address the challenges of large action space and the coupling of actions across time slots, we model the problem as a Markov decision process (MDP) and perform multi-agent DRL through multi-agent proximal policy optimization (MAPPO). We employ a centralized training with decentralized execution (CTDE) framework to enable each MD to make offloading decisions solely based on its local system state. Extensive simulations demonstrate the efficiency and fast convergence of our algorithm. In comparison to the existing sub-optimal results deploying single-agent DRL, our algorithm reduces the average task completion delay by 11.0% and the ratio of dropped tasks by 17.0%. Our proposed algorithm is particularly pertinent to sensor networks, where mobile devices equipped with sensors generate a substantial volume of data that requires timely processing to ensure quality of experience (QoE) and meet the service-level agreements (SLAs) of delay-sensitive applications.

关键词： Mobile edge computing   Deep reinforcement learning   Directed acyclic graph   Deep neural network  

摘要： With the rapid development of innovative applications, lots of computation-intensive and delay-sensitive tasks are emerging. Task offloading, which is regarded as a key technology in the emerging mobile edge computing paradigm, aims at offloading the tasks from mobile devices (MDs) to edge servers or the remote cloud to reduce system delay and energy consumption of MDs. However, most existing task offloading studies either didn't consider the dependencies among tasks or simply designed heuristic schemes to solve dependent task offloading problems. Different from these studies, we propose a deep reinforcement learning (DRL) based task offloading scheme to jointly offload tasks with dependencies. Specifically, we model the dependencies among tasks by directed acyclic graphs and formulate the task offloading problem as minimizing the average cost of energy and time (CET) of users. To solve this NP-hard problem, we propose a deep Q-network learning-based framework that creatively utilizes deep neural networks to extract system features. Simulation results show that our proposed scheme outperforms the existing DRL scheme and heuristic scheme in reducing the average CET of all users and can obtain near-optimal solutions.

关键词： Deep reinforcement learning   Lyapunov optimization   Mobile edge computing   Task offloading  

摘要： With the arrival of 5G technology and the popularization of the Internet of Things (IoT), mobile edge computing (MEC) has great potential in handling delay-sensitive and compute-intensive (DSCI) applications. Meanwhile, the need for reduced latency and improved energy efficiency in terminal devices is becoming urgent increasingly. However, the users are affected by channel conditions and bursty computational demands in dynamic MEC environments, which can lead to longer task correspondence times. Therefore, finding an efficient task offloading method in stochastic systems is crucial for optimizing system energy consumption. Additionally, the delay due to frequent user-MEC interactions cannot be overlooked. In this article, we initially frame the task offloading issue as a dynamic optimization issue. The goal is to minimize the system's longterm energy consumption while ensuring the task queue's stability over the long term. Using the Lyapunov optimization technique, the task processing deadline problem is converted into a stability control problem for the virtual queue. Then, a novel Lyapunov-guided deep reinforcement learning (DRL) for delay-aware offloading algorithm (LyD2OA) is designed. LyD2OA can figure out the task offloading scheme online, and adaptively offload the task with better network quality. Meanwhile, it ensures that deadlines are not violated when offloading tasks in poor communication environments. In addition, we perform a rigorous mathematical analysis of the performance of Ly2DOA and prove the existence of upper bounds on the virtual queue. It is theoretically proven that LyD2OA enables the system to realize the trade -off between energy consumption and delay. Finally, extensive simulation experiments verify that LyD2OA has good performance in minimizing energy consumption and keeping latency low.

关键词： detection task   railway internet of things   mobile edge computing   computation offloading   deep reinforcement learning  

摘要： Online detection devices, powered by artificial intelligence technologies, enable the comprehensive and continuous detection of high-speed railways (HSRs). However, the computation-intensive and latency-sensitive nature of these detection tasks often exceeds local processing capabilities. Mobile Edge Computing (MEC) emerges as a key solution in the railway Internet of Things (IoT) scenario to address these challenges. Nevertheless, the rapidly varying channel conditions in HSR scenarios pose significant challenges for efficient resource allocation. In this paper, a computation offloading system model for detection tasks in the railway IoT scenario is proposed. This system includes direct and relay transmission models, incorporating Non-Orthogonal Multiple Access (NOMA) technology. This paper focuses on the offloading strategy for subcarrier assignment, mode selection, relay power allocation, and computing resource management within this system to minimize the average delay ratio (the ratio of delay to the maximum tolerable delay). However, this optimization problem is a complex Mixed-Integer Non-Linear Programming (MINLP) problem. To address this, we present a low-complexity subcarrier allocation algorithm to reduce the dimensionality of decision-making actions. Furthermore, we propose an improved Deep Deterministic Policy Gradient (DDPG) algorithm that represents discrete variables using selection probabilities to handle the hybrid action space problem. Our results indicate that the proposed system model adapts well to the offloading issues of detection tasks in HSR scenarios, and the improved DDPG algorithm efficiently identifies optimal computation offloading strategies.

关键词： Mobile edge computing  

摘要： Handling long tail corner cases is a major challenge faced by autonomous vehicles (AVs). While large language models (LLMs) hold great potentials to handle the corner cases with excellent generalization and explanation capabilities and received increasing research interest on application to autonomous driving, there are still technical barriers to be tackled, such as strict model performance and huge computing resource requirements of LLMs, which are difficult to be met locally at AVs. In this paper, we investigate a new approach of applying remote or edge LLMs to support autonomous driving. With this approach connected autonomous vehicles (CAVs) send driving assistance requests to the LLMs. LLMs deployed at the edge of the networks or remote clouds process the requests and generate driving assistance instructions for the CAVs. A key issue for such LLM assisted driving system is the assessment of LLMs on their understanding of driving theory and skills, ensuring they are qualified to undertake safety critical driving assistance tasks for CAVs. As there is no published work on assessing LLM of driving theory and skills, we design and run driving theory tests for several proprietary LLM models (OpenAI GPT models, Baidu Ernie and Ali QWen) and open-source LLM models (Tsinghua MiniCPM-2B and MiniCPM-Llama3-V2.5) with more than 500 multiple-choices theory test questions. These questions are close to the official UK driving theory test ones. Model accuracy, cost and processing latency are measured from the experiments. Experiment results show that while model GPT-4 passes the test with improved domain knowledge and Ernie has an accuracy of 85% (just below the 86% passing threshold), other LLM models including GPT-3.5 fail the test. For the test questions with images, the multimodal model GPT4-o has an excellent accuracy result of 96%, and the MiniCPM-Llama3-V2.5 achieves an accuracy of 76%. While GPT-4 holds stronger potential for CAV driving assistance applications, the

关键词： Task analysis   Training   Multitasking   Costs   Computational modeling   Data models   Resource management   Artificial intelligence (AI)   federated learning (FL)   mobile edge computing   multitask learning (MTL)  

摘要： The wide applications of artificial intelligence (AI) on massive Internet of Things or smartphones raises significant concerns about privacy, heterogeneity, and resource efficiency. Correspondingly, federated learning (FL) emerges as an effective way to enable AI over massively distributed nodes without uploading the raw data. Conventional works mostly focus on learning a single unified model for one solitary task. Multitask learning (MTL) outperforms single-task learning by training multiple models concurrently, leading to reduced model sizes and increased flexibility. However, existing FL efforts often face challenges in efficiently managing MTL scenarios, particularly with the presence of stragglers, without incurring prohibitive computation and communication *** this article, inspired by the natural cloud-base station (BS)-terminal hierarchy of cellular networks, we provide a viable resource-aware hierarchical federated MTL (RHFedMTL) solution to meet the task heterogeneity corresponding to different nonindependent and identically distributed (IID) training data sets. Specifically, a primal-dual method has been leveraged to effectively transform the coupled MTL into some local optimization subproblems within BSs. Therefore, it enables solving different tasks within a BS and aggregating the multitask result in the cloud without uploading the raw data. Furthermore, compared with existing methods that reduce resource costs by simply changing the aggregation frequency, we dive into the intricate relationship between resource consumption and learning accuracy, and develop a resource-aware learning strategy for adjusting the iteration number on local terminals and BSs to meet the resource budget. Extensive simulation results demonstrate the effectiveness and superiority of RHFedMTL in terms of improving the learning accuracy and boosting the convergence rate.

关键词： Mobile edge computing  

摘要： With rapid increase of computing demands from users in mobile edge computing (MEC), the traditional mobile edge servers are no longer capable of providing flexible and timely services. Unmanned aerial vehicle (UAV)-assisted mobile edge computing is envisioned as a potential solution to provide high quality services for large-scale user devices. In UAV-assisted mobile edge computing, there are many complex applications such as obstacle detection, image recognition and path planning. These applications are characterized by huge computational resources and extremely low-latency requirements. Most of complex applications in the real-world can be expressed as a set of interdependent tasks, also known as application workflows. Due to the limited battery capacity and computing capabilities of UAVs, and the complex dependencies between workflow tasks, it is challenging to jointly optimize the UAV path planning and the workflow offloading under deadline constraints. To address this challenge, this paper proposes a distributed deep reinforcement learning algorithm based on the graph convolutional network (GCN) and asynchronous-advantage-actor-critic (A3C) learning that minimizes the energy consumption of UAVs while meeting the deadlines of workflows. We conduct simulation experiments to evaluate the performance of the proposed algorithm under complex workflow applications. The results show that our algorithm improves the performance by about 33.6% compared with the other optimization approaches. © 2024, The Authors. All rights reserved.

关键词： Mobile edge computing  

摘要： On-device large language models (LLMs), referring to running LLMs on edge devices, have raised considerable interest since they are more cost-effective, latency-efficient, and privacy-preserving compared with the cloud paradigm. Nonetheless, the performance of on-device LLMs is intrinsically constrained by resource limitations on edge devices. Sitting between cloud and on-device AI, mobile edge intelligence (MEI) presents a viable solution by provisioning AI capabilities at the edge of mobile networks. This article provides a contemporary survey on harnessing MEI for LLMs. We begin by illustrating several killer applications to demonstrate the urgent need for deploying LLMs at the network edge. Next, we present the preliminaries of LLMs and MEI, followed by resource-efficient LLM techniques. We then present an architectural overview of MEI for LLMs (MEI4LLM), outlining its core components and how it supports the deployment of LLMs. Subsequently, we delve into various aspects of MEI4LLM, extensively covering edge LLM caching and delivery, edge LLM training, and edge LLM inference. Finally, we identify future research opportunities. We hope this article inspires researchers in the field to leverage mobile edge computing to facilitate LLM deployment, thereby unleashing the potential of LLMs across various privacy- and delay-sensitive applications. Copyright © 2024, The Authors. All rights reserved.

关键词： network slicing   zero-touch network   resource slicing   mobile edge computing   service function chaining   network function virtualization   blockchain   SDN  

摘要： Zero-touch networks (ZTNs) can provide autonomous network solutions by integrating software-based solutions for various emerging 5G and 6G applications. The current literature does not provide any suitable end-to-end network management and resource-slicing solutions for service function chaining (SFC) and user intent-based (time and cost preference) 6G/non-6G application execution over ZTNs enabled by mobile edge computing, network function virtualization, and software-defined networking. To tackle these challenges, this work initiates an end-to-end network management and user intent-aware intelligent network resource-slicing scheme for SFC-based 6G/non-6G application execution over ZTNs, taking into account various virtual and physical resources, task workloads, service requirements, and task numbers. The results depicted that at least 25.27% average task implementation delay gain, 6.15% energy gain, and 11.52% service monetary gain are realized in the proposed scheme over the compared schemes.

关键词： Mobile edge computing  

摘要： A novel paradigm of mobile edge generation (MEG)enabled digital twin (DT) is proposed, which enables distributed on-device generation at mobile edge networks for real-time DT applications. First, an MEG-DT architecture is put forward to decentralize generative artificial intelligence (GAI) models onto edge servers (ESs) and user equipments (UEs), which has the advantages of low latency, privacy preservation, and individual-level customization. Then, various single-user and multi-user generation mechanisms are conceived for MEG-DT, which strike trade-offs between generation latency, hardware costs, and device coordination. Furthermore, to perform efficient distributed generation, two operating protocols are explored for transmitting interpretable and latent features between ESs and UEs, namely sketch-based generation and seed-based generation, respectively. Based on the proposed protocols, the convergence between MEG and DT are highlighted. Considering the seed-based image generation scenario, numerical case studies are provided to reveal the superiority of MEG-DT over centralized generation. Finally, promising applications and research opportunities are identified. Code is available at https://***/xiaoxiaxusummer/MEG_ DT. Copyright © 2024, The Authors. All rights reserved.