关键词： Path planning   Collision avoidance   Reinforcement learning   Mobile edge computing   Unmanned aerial vehicle   Disaster management  

摘要： Unmanned aerial vehicles (UAVs) are significantly used in a various field, including disaster relief and rescue operations. The goal of path planning is to determine the shortest or most efficient route while using the least amount of energy and resources possible. In environmental disaster scenarios, multiple UAVs can be deployed in search and rescue operations to look for individuals that are trapped in the disaster areas. To address the issue of multi-UAV path planning, numerous academics have recently proposed heuristic, sampling, and machine learning-based techniques. However, deploying UAVs in dynamic environments with moving obstacles and interUAV interaction presents considerable problems. This paper presents the framework for path planning optimization for multi-UAV-assisted Mobile Edge Computing (MEC) network using Reinforcement Learning (RL) strategy during disaster relief operations in an urban environment. To tackle these challenges, we provide a novel multi-UAV Mobile Edge Computing architecture that uses reinforcement learning to improve Quality-of-Service and path planning. Our primary contributions include: 1) maximizing the quality of service for UAV-assisted mobile edge computing and path planning strategy in the reinforcement learning framework;2) illustrating the computational demand of terminal users to secure higher level quality-of-service using a sigmoid-like function;3) applying a reinforcement learning incentive matrix that incorporates synthetic considerations of the geometric distance, risk level, and demand of the terminal users to guarantee cost savings, risk avoidance, and the service quality. In this paper, the proposed framework considers the configuration of urban surroundings, including critical operational constraints, such as obstacle mobility and obstacle avoidance. From the simulation testing results, it shows that the proposed strategy is more appropriate in the disaster scenario since it is effective and performs better than

关键词： Post-disaster rescue   Mobile edge computing   Unmanned aerial vehicles (UAVs)   Path planning   Reinforcement learning  

摘要： This study addresses critical challenges in urban disaster rescue operations, such as fires, including communication failures, complex environments, and information scarcity. We propose a novel Dual-layer UAV-assisted Mobile Edge Computing (DUAMEC) system, leveraging an air-space-ground collaborative communication framework and intelligent task scheduling to overcome traditional limitations like information blind spots, decision-making delays, and inefficient response. DUAMEC innovatively combines a high-altitude upper-layer UAV (U-UAV) for wide-area coverage and a low-altitude down-layer UAV (D-UAV) for task processing, achieving strong coverage, low latency, and high energy efficiency. The core innovations of the DUAMEC system are manifested in the following aspects: First, we propose a grid-based adaptive multi-stage greedy optimization algorithm for optimal UAV deployment, dynamically generating multi-level candidate grids and employing adaptive step-size contraction. An uncovered-point compensation mechanism ensures continuous area coverage. Second, we design a Multi-Agent TD3 with Hindsight Priority Experience Replay (MATD3-HP) algorithm, utilizing a multi-dimensional state space and compound reward mechanism to optimize resource allocation, path planning, and task offloading in dynamic obstacle environments. Experimental results demonstrate that compared to conventional single-layer UAV-MEC systems and fixed path planning schemes, the DUAMEC system achieves an 66.78% reduction in system overhead while maintaining 98% user coverage. Simultaneously, it sustains stable performance with low task processing latency and energy consumption even in scenarios with dense user distribution and highly dynamic obstacles, thereby providing an efficient and reliable intelligent solution for urban disaster rescue operations.

关键词： Mobile edge computing   Wireless power transfer   Reinforcement learning   Binary offloading   Mobile edge computing   Wireless power transfer   Reinforcement learning   Binary offloading   Mobile edge computing   Wireless power transfer   Reinforcement learning   Binary offloading  

摘要： Most Internet of Things (IoT) devices face challenges in handling complex computational tasks due to their limited computing capabilities. To address this issue, Mobile Edge Computing (MEC) has been introduced, which significantly enhances computational efficiency and response speed by offloading tasks to the cloud or the network edge. Additionally, by integrating Wireless Power Transfer (WPT) technology, IoT devices can harvest energy wirelessly, thereby alleviating energy constraints. This paper investigates a WPT-enabled MEC network with the goal of minimizing the system's overall energy consumption. First, we formulate the energy minimization problem as a mixed-integer nonlinear programming (MINLP) problem. Then, we propose a Deep Reinforcement Learning (DRL)-based algorithm to jointly optimize offloading decisions and time allocation. Simulation results demonstrate that the proposed approach not only converges quickly but also achieves performance comparable to that of the exhaustive search method. Furthermore, it significantly reduces energy consumption compared to baseline schemes.

关键词： Computation offloading   Resource allocation   Ultra-dense networks   Mobile edge computing  

摘要： To ensure gains from the ultra-dense networks (UDNs) with multi-slope channels, we cluster small base stations (SBSs) with intra-cluster orthogonal bands and inter-cluster shared bands, and establish some new stipulations tailored for UDN deployment. Under such proposed interference management strategies and proportional remote computation allocation, we jointly optimize computation offloading and power control to minimize local energy consumption. The problem is decomposed into: (1) an offloading subproblem solved via an improved coalitional game-based algorithm, and (2) a power control subproblem transformed into single-variable polynomial optimization, solved via a two-tier iterative algorithm using Newton-like and fixed-point iteration approaches. Simulations demonstrate that the designed algorithm, integrating proposed interference management strategies and proportional remote computation allocation, outperforms existing methods in minimizing local energy consumption under strict latency requirements.

关键词： mobile crowdsensing   mobile edge computing   lyapunov optimization   bipartite graph matching  

摘要： Mobile CrowdSensing (MCS) has become a powerful sensing paradigm for information collection recently. As sensing becomes more complicated, it is beneficial to deploy edge servers between users and the cloud center with a so-called mobile edge computing. Instead of directly offloading the sensing data to the cloud center, mobile users offload the sensing data to the edge servers. Then, the edge server processes and transmits the data to the cloud center in a distributed and parallel manner. It's however critically important to balance cost, such as energy consumption, and the stability of the queues on both mobile users and edge servers. Therefore, to minimize the data offloading cost while maintaining system stability, we should carefully design the sensing data offloading strategy for edge-based crowdsensing. To this end, we formulate a double-queue Lyapunov optimization problem and propose a sensing data offloading strategy. We analyze the upper bounds of the total offloading cost and queue backlog. We further formulate the heterogeneous sensing data problem as the minimum weight bipartite graph matching problem and develop an approach that is based on Kuhn-Munkres algorithm. Finally, we conduct simulations based on three mobility sets. Simulation results show that the proposed techniques outperform several state-of-art algorithms in overall cost, system stability, and other performance metrics.

关键词： Mobile edge computing   UAV   Computational task offloading   Nature-inspired algorithms  

摘要： With the rise of Mobile Edge Computing (MEC) and the increasing number of User Equipments (UE), traditional MEC systems struggle with high UE density. UAVs can assist in offloading tasks from base stations, but their limited resources make deployment and offloading strategies critical. This paper investigates UAV deployment strategies and task offloading policies in scenarios where UE density varies over time. First, we introduce the Maximum Clique in Weighted Graph (MCWG) algorithm, which is designed to calculate the UAV deployment coordinates within a weighted graph. In the task offloading phase, considering the resource constraints of UAVs, we develop a collaborative computation offloading framework involving UE, UAVs, MEC, and cloud servers. Secondly, we propose an Improved Hybrid Nature-Inspired Optimization (IHNIO) algorithm under delay and energy consumption constraints. This algorithm aims to minimize the average delay and energy consumption of UEs in UAV-assisted Mobile Edge Computing. Simulation results show that our approach significantly enhances energy efficiency compared to baseline solutions, with potential improvements in efficiency reaching up to 11%.

关键词： Mobile edge computing   Partial computation offloading   Load-balancing   Multi-agent reinforcement learning  

摘要： Mobile edge computing (MEC) can enhance the computation performance of end-devices by providing computation offloading service at the network edge. However, given that both end-devices and edge servers have finite computation resources, inefficient offloading policies may lead to overload, thereby increasing the computation delays of tasks. In this paper, we investigate a multi-task partial computation offloading problem combined with a queue model. Based on achieving load-balancing across the MEC system, our objective is to minimize the long-standing average task-processing cost of the end-devices while ensuring the delay thresholds of tasks. For this purpose, a distributed offloading algorithm utilizing the multi-agent deep reinforcement learning (MADRL) method is proposed. Specifically, through interacting with the MEC environment and accumulating experience data, the device agents can collaborate to optimize their local offloading decisions over continuous time-slots, which includes adjusting the transmission power and determining the tasks' offloading ratios under the dynamic wireless channel conditions. Exhaustive experimental results demonstrate that in contrast with the baseline algorithms, the proposed offloading algorithm can not only better balance the computation loads between the end-devices and the MEC server, but also more effectively reduce the task-processing cost of the end-devices, as well as the percentage of timeout tasks.

关键词： Digital twin   Mobile edge computing   Optimization   Resource management   Robust priority   Unmanned aerial vehicles  

摘要： Unmanned Aerial Vehicles (UAVs) play a critical role in replenishing the energy of power-constrained Internet of Things (IoT) devices, particularly in public safety operations, thereby maintaining continuous system functionality. Integrating Mobile Edge Computing (MEC) into UAV platforms enables offloading computational tasks to aerial nodes, optimizing resource utilization. Efficient orchestration of communication, computation, caching, and energy resources is imperative to maximize the benefits of UAV-assisted MEC networks. Additionally, ensuring high situational awareness is essential for supporting priority-based latency-sensitive applications. Digital twin technology can be instrumental in minimizing latency by generating a real-time digital representation of the physical infrastructure, enabling enhanced system monitoring and optimization. Accordingly, we have formulated an optimization problem to maximize the number of IoT devices UAVs can support while adhering to predefined constraints. The formulated problem is a mixed integer non-linear programming model. Additionally, the dynamic management of tasks with varying priorities and computational demands introduces a significant resource allocation and scheduling challenge. Our proposed approach entails an efficient task offloading and priority-based scheduling strategy that prioritizes tasks, allocating computational resources to those with higher priority. The approach encompasses a multistage offloading strategy combining an interior-point method with a learning algorithm to address the inherent complexity and provide a viable solution. Simulation results validate the effectiveness of the proposed approach, outperforming conventional methods. Specifically, the Penalty Function Method Heuristic combined with the Interior Point Method achieves superior user connectivity compared to the Simple Relaxation Heuristic strategy.

关键词： Mobile edge computing   Truthfulness   Resource allocation   Algorithm design  

摘要： In this paper, we address the multi-mapping multi-task allocation and pricing problem in mobile edge computing (MEC). Based on different allocation scenarios, we propose partial and full allocation models. The partial model enables an MD's tasks to be partially offloaded with the MD bidding for a single task to be performed on the server (one bid per task). The full model offloads all or none of the MD's tasks with the MD bidding for all of them together. We consider a multi-mapping multi-tasking model with platform constraints to enable mobile devices to offload tasks to multiple servers that meet the platform and version requirements. This paper aims to solve the social welfare maximization problem which is the sum of MDs' valuations. We propose the greedy mechanisms for the partial and full allocation models, respectively. Selfish MDs have an incentive to manipulate the bidding system by making untrue bids to obtain a greater allocation. Therefore, we demonstrate that our proposed mechanisms achieve truthfulness, which drives the system into an equilibrium in which no MD has incentives to maximize utility by untruthfully declaring valuations. Moreover, we show the proposed two mechanisms achieve the desired properties, including computation efficiency and individual rationality. Experiment results show that the proposed mechanisms obtain near-optimal solutions in a reasonable amount of time.

关键词： Mobile edge computing   Unmanned aerial vehicles   Energy consumption   Digital twin  

摘要： The future sixth generation (6G) mobile network will be a highly heterogeneous system that integrates diverse technologies and communication paradigms, encompassing diverse consumer electronics devices, including various internet of things (IoT) devices utilizing different protocols. The combination of unmanned aerial vehicles (UAVs) and mobile edge computing (MEC) with 6G has created ground-breaking prospects for effective data processing and computation services in the IoT. Despite this improvement, the presence of an eavesdropper introduces significant security risks into the computing process of mobile devices (MDs), permitting the collection of sensitive data from MDs and perhaps affecting the correctness of offloaded computations. This study proposes an efficient technique for reducing energy usage in a secure digital twin (DT)-enabled UAVassisted MEC metaverse network that faces the threat of a UAV eavesdropper. By ensuring the secure processing of all MDs data, the network meets its energy consumption objectives by optimizing trajectories and resources, taking into account parameters like as time, local computation, and offloading computation dispersion. Because of the complicated arrangement of the interplay of various variables and non-linear constraints, solving the problem directly is extremely difficult. To solve this complexity, an auxiliary variable is used to rearrange the problem into a more understandable format. We employ DT-enabled DRL to address this issue because traditional methods are inadequate. The empirical findings indicate that the proposed methodology achieved a reduction in energy consumption of approximately 83.32% and a decrease in time delay of roughly 11.97%, in comparison to the prevailing baseline methodologies.