关键词： Computational modeling   Heuristic algorithms   Delays   Collaboration   Long short term memory   Vehicle dynamics   Predictive models   Prediction algorithms   Decision making   Resource management   Mobile edge computing   task offloading   deep reinforcement learning   internet of vehicles  

摘要： Roadside units (RSUs) with strong sensing abilities enhance the viability of the RSU-to-Everything (R2X) paradigm, offering crucial infrastructure support for Mobile Edge Computing (MEC) that enables real-time data processing and reduced delay. Since RSUs collect a large volume of data but have limited computing capability, data analysis tasks are usually offloaded to other network nodes, such as the cloud, other RSUs, or even vehicles. The multi-hop distributed collaborative task offloading scheme is expected to achieve high resource utilization efficiency and low task delay in this scenario, despite increasing energy consumption in data transmission. However, the highly dynamic nature of the R2X network topology makes it challenging for a node to independently select the next hop and collaboratively allocate tasks to neighbors in a multi-hop transmission path. Specifically, offloading decisions made by an individual node are influenced not only by its immediate neighbors but also by other nodes along the multi-hop path, referred to in this paper as the effect value. Additionally, the heterogeneity in computing resources and link delays among network nodes further increases the difficulty. To address these challenges, we first apply a Long Short-Term Memory (LSTM) model to predict and update the neighbors for each node while considering effect values, allowing them to independently adapt to environmental changes. Then, we design a two-layer Deep Reinforcement Learning (DRL) algorithm for network nodes to make decisions. The first-layer DRL algorithm is implemented by RSUs to determine task offloading modes. When an RSU decides to offload tasks to multiple vehicles for collaborative computing, the second-layer DRL algorithm is used by a vehicle to select its next hop vehicle and allocate tasks. Simulation results show that our proposed approach effectively adapts to topology changes in complex and highly dynamic network environments. Compared with existing methods,

关键词： Architecture   Deep Reinforcement Learning   Computation offloading   Task offloading   Metrics   Multi-Access Edge Computing   Mobile Edge Computing   Unmanned Aerial Vehicles   Drones   Internet of Things   Wireless sensor networks  

摘要： Unmanned Aerial Vehicle (UAV)-assisted Internet of Things (IoT) applications vary widely including data monitoring, data collection and analysis, intelligent navigation and object tracking, surveillance and emergency response, vehicular and intelligent transport, and agricultural, marine, and photogrammetry. Mobile Edge Computing (MEC)-based UAV-assisted IoT networks enable resource-constrained mobile or IoT devices to offload computationally demanding tasks to UAVs or edge nodes with more computing power in order to improve battery consumption, performance, or Quality of Service. UAV-assisted IoT applications generally require a lot of precision for efficient UAV control involving UAV movement and position optimization and Deep Reinforcement Learning (DRL) has recently been identified as one of the most prominent techniques for facilitating this and optimizing of the terrestrial network performance, thus improving the operation of these applications. This paper aims to answer twelve important research questions relating to the research on DRL for Mobile Edge Computing (MEC)-based UAV-assisted sensor and IoT applications from 47 systematically selected articles. The questions cover a variety of topics including the UAV-assisted MEC IoT applications studied, variants of deployed DRL, the purpose of DRL, Markov Decision Processes (MDPs) components, unique network architectural features, environments and integrated technologies, role of UAVs, optimization constraints, joint optimization frameworks, energy-management techniques, metrics examined, benchmark algorithms and performance results as well as identified probable future considerations based on the review. Lastly, the challenges and future directions of DRL application in UAV-assisted MEC systems are discussed. This paper aims to provide both communication generalists and optimization specialists with a comprehensive understanding of the research landscape in this field.

关键词： Mobile edge computing   Heterogeneous cellular networks   Stackelberg game   Computation offloading   Resource pricing  

摘要： The explosive growth of mobile users (MUs) has led to a substantial increase in computing tasks being offloaded to MEC, bringing tremendous pressure to the sustainable development of MEC. To address this problem, the integration of multi-base station (BS) with MEC has attracted widespread attention. In this paper, we propose the joint optimization problem of resource allocation, computation offloading, and resource pricing for multi-BS cooperative MEC in heterogeneous multicell networks. We model the interactions between MEC server, macro base station (MBS), femto base station (FBS), and MUs as a four-stage Stackelberg game, and derive the equilibrium resource pricing and allocation strategies. In the first three stages, we focus on devising optimal resource allocation and pricing strategies to maximize the utility of all stakeholders. In the fourth stage, MUs make an offloading decision to maximize their utility based on the different pricing strategies. The optimization problems in each stage are simplified and proved by derivation using backward induction method. Our approach offers a novel strategy by effectively integrating multiple BSs, enabling a more flexible and efficient resource distribution. The simulation results demonstrate that the proposed multi-BS model can improve the utilities of MBS and MUs by 53% and 50% respectively by introducing FBS.

关键词： Delays   Industrial Internet of Things   Optimization   Heuristic algorithms   Energy consumption   Dynamic scheduling   Resource management   Industrial IoT   mobile edge computing   computation offloading   multi-objective optimization   deep reinforcement learning  

摘要： With the increasing scale of industrial equipments, delay and energy consumption have emerged as critical concerns within the Industrial Internet of Things (Industrial IoT). Mobile edge computing (MEC) offloads tasks to nearby edge servers to meet the demands of delay-sensitive applications. However, the limitations of edge computing resources can lead to significant processing delays or even task failures when offloading numerous tasks. Furthermore, it is difficult for existing centralized algorithms to acquire global information within large-scale industrial environment. To tackle these challenges, the computation offloading problem is transformed into a decentralized partially observable Markov decision process (Dec-POMDP) with rewards for delay and energy consumption, and a decentralized multi-objective computation offloading method is proposed to achieve the long-term reward maximization. Specifically, two deep neural networks, namely the delay and energy network, are designed to estimate the expected rewards for each offloading decision in terms of delay and energy consumption. Meanwhile, to achieve dynamic offloading, gated recurrent unit (GRU) is introduced to predict the occupancy of edge computing resources, and an adaptive weight network is devised to dynamically adjust the weights of optimization objectives based on historical information. Comprehensive experiments demonstrate that the proposed method effectively meets the requirements of delay-sensitive tasks, as well as minimizing long-term delay and energy consumption.

关键词： Evolutionary multitasking   Multifactorial evolutionary algorithm   Multi-form evolutionary optimization   Service migration optimization   Mobile edge computing  

摘要： In high-speed mobile networks, mobile edge computing is tasked with service migration optimization, i.e., assigning mobile users to the right servers to minimize the response time. Service migration optimization is a complex problem posing significant challenges to conventional optimization methods. To tackle this problem, we develop a multifactorial memetic algorithm with adaptive auxiliary tasks or MFMA-AAT for short. MFMA-AAT solves the target service migration optimization problem and an adaptively selected auxiliary task simultaneously, where the auxiliary task is a simplified version of the target problem to guide the search towards promising regions faster via knowledge transfer. Multiple auxiliary tasks are pre-constructed based on the distribution of the mobile users and the one with the best improvement at each generation is selected for knowledge transfer. A community detection-based memetic operator is also introduced to accelerate the local convergence of the proposed algorithm. Experimental results on test problems demonstrate that MFMA-AAT is more efficient than traditional service migration approaches and other state-of-the-art multifactorial evolutionary algorithms.

关键词： Mobile edge computing   heterogeneous internet of things   big data analysis   subchannel   terminal  

摘要： The big data heterogeneous Internet of Things (IoT) requires mobile edge computing (MEC) to process some data, and the data analysis system of MEC often has the problem of excessive terminal energy consumption (ECS) or long delay. So this study designed an energy-saving optimization algorithm for the task offloading processing module in the big data heterogeneous IoT analysis system, and designed and conducted simulation experiments to verify the application performance of the algorithm. The experimental results show that the #04 scheme of the designed algorithm has the lowest terminal ECS under the same conditions. Choosing the #04 scheme to build the algorithm, comparative analysis shows that when the edge server (ES) computing rate is 10 cycles/s, the weighted sum values of terminal ECS for EOPU, MPCO, exhaustive search, and local computing methods are 23.6 J, 23.9 J, 28.5 J and 84.5 J, respectively. Moreover, the algorithm possesses a significantly higher percentage of remaining time under different conditions of total SMD devices and total subchannels compared to other methods. This indicates that the designed algorithm can markedly enhance the processing performance of the task offloading model of the big data heterogeneous IoT data analysis system, and can also effectively reduce terminal ECS and system latency. The research results can provide reference for improving the processing ability of heterogeneous IoT big data analysis systems. The contribution of this study to the academic field lies in providing a model that can effectively reduce the operational ECS and time consumption of heterogeneous IoT big data analysis systems containing mobile animal networking devices. Moreover, from an industrial perspective, the results of this study contribute to improving the efficiency of information exchange and processing in the field of IoT computing, thereby promoting the promotion of IoT technology.

关键词： Internet of vehicles   Mobile edge computing   Deep reinforcement learning   Task offloading   Resource allocation  

摘要： In vehicular networks, the increasing demand for computational resources often exceeds the capabilities of in-vehicle devices. To address these challenges, we propose a cloud-edge-device collaborative framework integrated with a Multi-Agent Deep Reinforcement Learning (MADRL) algorithm for dynamic optimization of task offloading and resource allocation. Experimental evaluations demonstrate the proposed algorithm's superiority over traditional methods, achieving an 11% reduction in energy consumption and a 23% increase in task completion rate compared to local processing-only strategies, while reducing average task delay by 50% relative to static offloading approaches. The MADRL-based framework not only ensures efficient task distribution but also adapts to fluctuating network conditions, achieving a resource utilization rate of 85%. These findings underscore its potential to enhance performance in intelligent transportation systems by balancing computational efficiency, energy consumption, and task latency.

关键词： UAV   mobile edge computing   dynamically processes   computing resources   communication technology  

摘要： With the advancement of communication technology, mobile edge computing (MEC) is considered a key technology for handling computation-intensive and latency-sensitive tasks. However, in scenarios such as disaster response and emergency rescue, edge servers cannot be quickly deployed to provide task processing services. The impressive mobility and simplicity of deployment of unmanned aerial vehicles (UAVs) have garnered a lot of attention for their potential to help MEC. However, allocating resources is made more difficult by the fact that UAVs have limited energy and computational capabilities. A suggested resource-efficient technique for multi-UAV collaboration (LUAVs-Cor) aims to tackle this issue. Tasks are dynamically processed by multi-UAV collaboration in this technique. Using a search for the ideal task combination, it finds task transmission mechanisms to make the most of UAVs' idle processing capabilities. Additionally, by estimating UAV processing capabilities, task quantity, and processing status, it simplifies the process of deploying UAVs to lower power usage. We observed that the LUAVs-Cor method boosts service capacity by around 6.8% and decreases total UAV energy consumption by 10.3% through comprehensive simulated trials. More customers may be served at reduced energy costs with the suggested LUAVs-Cor method, which involves minimum cooperation expenses for UAVs.

关键词： UAV   Mobile edge computing   Task offloading   Multi-cache placement   Deep reinforcement learning  

摘要： Unmanned aerial vehicles (UAVs) are being developed as a promising technology to assist mobile edge computing (MEC) systems due to their reliable wireless communication, flexible computing service capabilities, and flexible deployment. However, in the face of huge information and demanding task delay, it is a challenging problem to reduce the system cost. This paper studies task offloading and cache space placement for ground users, proposes a multi-UAV assisted computing framework, which is a four-layer transmission system composed ground users (UE), UAVs, edge data centers (EDC) and remote clouds. By jointly optimizing UAV cache flight path, offloading decision, channel ratio, and battery power, we formulate the problem to minimize the term average weighted cost of the system under the constraint of cache space and computing resources. Since problem is a mixed integer variable problem, we design a task offloading and cache placement algorithm on deep reinforcement learning, namely the Cooperative Long-term Average Cost Minimization Optimization Algorithm (CLACMO). Firstly, we transform the mixed action variable space by using embedded tables conditional variational autoencoders (VAE) combined with latent space, and map the mixed action variable latent action space. This approach effectively unifies discrete and continuous actions, addressing the challenge of mixed action spaces that traditional deep reinforcement learning algorithms struggle with. Secondly, on the deep reinforcement learning (DRL), we achieve the long-term system average weighted cost minimization more efficiently under the constraints of task offloading and cache placement. The results show that compared with the PER-UOS-RL, MASAC, and MADDPG algorithms, the average reward has increased by 54.5%, 66.7%, 69.7% respectively, and the average task completion rate has increased by 12.9%, 38.1%, and 9.11% respectively, demonstrating the effectiveness of our novel method.

关键词： Mobile networks   5G communications   Vehicle-to-Everything   Mobile edge computing   Mobile Netzwerke   5G-Kommunikation   Vehicle-to-Everything-Kommunikation   Mobile Edge Computing  

摘要： In the context of the growing demand for ultra-reliable and low-latency communications (URLLC) in Vehicle-to-Everything (V2X) applications and edge computing offloading, there is a growing imperative to understand the end-to-end Quality of Service (QoS) performance in real-world cellular networks. This study presents an empirical evaluation of the performance of the 4G LTE and 5G Non-Standalone (NSA) network performance, based on extensive drive tests conducted in Bratislava, Slovakia. The analysis of the empirical cumulative distribution function (ECDF) of key Quality of Service (QoS) metrics, specifically latency and throughput, has highlighted the variability of end-to-end performance in different mobility and coverage scenarios. The results obtained have provided insights into the suitability of current cellular infrastructures to support latency-sensitive applications, particularly in vehicular environments where seamless connectivity and real-time data processing are essential. The need for field measurements to complement theoretical and simulation-based studies is emphasized in the findings, which offer valuable input for network optimization and future 5G Standalone (SA) deployments. This work contributes to ongoing efforts to improve the reliability of cellular networks for mission-critical applications, ensuring robust connectivity for next generation intelligent transportation systems. Angesichts der wachsenden Nachfrage nach ultra-zuverl & auml;ssiger und latenzarmer Kommunikation (URLLC) in Vehicle-to-Everything-Kommunikation (V2X)-Anwendungen sowie bei der Auslagerung von Rechenprozessen an das Edge Computing, wird ein vertieftes Verst & auml;ndnis der End-to-End Quality of Service (QoS) in realen Mobilfunknetzen zunehmend unerl & auml;sslich. Diese Studie pr & auml;sentiert eine empirische Bewertung der Leistungsf & auml;higkeit von 4G LTE- und 5G Non-Standalone (NSA)-Netzen, basierend auf umfangreichen Fahrtests in Bratislava, Slowakei. Die Analyse