Mobile edge computing technology has facilitated the deployment of high computational algorithms on Marine Internet of things (MIoT) devices that equipped with limited computing resources. However, the dynamic changes in the network environment, the strict requirements of system latency and energy consumption of mobile devices, restrict the task execution in MIoT. This paper proposes an offloading framework in a marine mobile edge computing scenario, which is assisted by sea buoys and unmanned aerial vehicles (UAVs). Aiming to achieve long-term system optimization goals under resource-constrained conditions, with task partitioning, user scheduling, and resource allocation joint optimization under the constraints of UAV residual battery life, system latency, and energy consumption. This paper specifically introduces a Network Partition Point Reservation Algorithm (NPPR) that reduces the solution space of the problem through preprocessing. Subsequently, the paper presents a multiagent deep deterministic policy gradient (MADDPG) algorithm, enhanced with standardization and adaptive learning rate decay (SA-MADDPG), to address task heterogeneity and environmental dynamics. Simulation results demonstrate that, compared to existing algorithms, the SA-MADDPG algorithm proposed in this paper reduces system latency and energy consumption by 34.7% and 61.2%, respectively.

With the development of in-orbit satellite hardware technology, distributed model training and updating for in-orbit satellites has become a promising trend. However, performance significantly degrades due to heterogeneous data distribution and the lack of high-quality labeled data across satellites. To address these issues, we propose a distributed unsupervised representation learning method called Progressive Uniformity (DURL-PU) from a deep clustering perspective. DURL-PU explores instance-level, cluster-level, and structure-level representations to progressively enhance uniformity, thereby mitigating dimensional collapse caused by data heterogeneity. Specifically, we first perform instance-level contrastive learning to distinguish between positive and negative samples. Second, we introduce a cluster-level loss to improve within-cluster compactness and between-cluster dispersion while incorporating high-level semantic information. Finally, we propose a deep clustering-based structural loss to enhance the consistency of parameter distribution among different satellites. Extensive experiments under various heterogeneous data scenarios demonstrate the effectiveness of DURL-PU.

This article investigates the issue of cooperative real-time trajectory design for multiple unmanned aerial vehicles (UAVs) to support emergency communication in disaster areas. To restore communication links rapidly between mobile users (MUs) and the ground base stations, UAVs equipped with both radio frequency (RF) modules and free space optics (FSO) modules are utilized as relay nodes. Given the challenges of setting up a central controller for the UAVs and the urgency of emergency communication, the trajectory design problem for these UAVs is formulated as a distributed cooperative optimization problem. Based on the enhanced${K}$-mean algorithm and multiagent PPO (MAPPO) algorithm, a cooperative trajectory design method, abbreviated as KMAPPO, is proposed for the UAVs to minimize interaction overhead and optimize deployment efficiency. Compared to the state-of-the-art deep reinforcement learning (DRL) methods, simulations reveal KMAPPO’s superior performance. It converges 32% faster, boosts RF allocation efficiency, and augments FSO communication backhaul capacity.

Marine Internet of Things (MIoT) has emerged as a prominent technology for the future development of marine applications, in which edge equipment provides a valuable method for information collection and processing on smart mobile devices (SMDs). However, the deployment of edge equipment may result in high latency due to inefficient computing offloading schemes. In this paper, we propose an optimal offloading scheme based on a dynamic unmanned air vehicle (UAV) assisted marine mobile edge computing (MMEC) environment in which latency-sensitive computing tasks can be partially offloaded autonomously. Specifically, we consider a time-varying scenario where the UAV hovers over multiple maritime mobile unmanned surface vessels (USVs) and provides MIoT services over communication periods. Our objective is to minimize the overall task execution time through joint optimization of user scheduling variables, UAV motion trajectory, and resource allocation while considering energy consumption and spatial constraints, thereby achieving enhanced quality of service. Considering the non-convexity of this optimization problem, we propose an advanced Twin Delayed Deep Deterministic policy gradient (ATD3) algorithm and examine the convergence and optimality of different parameter factors. Simulation results demonstrate that the proposed algorithm is superior to the baseline scheme regarding convergence speed, adaptability, and task execution time.

Implementing either Federated learning (FL) or split learning (SL) over clients with limited computation/communication resources faces challenges on achieving delay-efficient model training. To overcome such challenges, we investigate a novel distributedClustering-basedHybrid fEdEratedSplit lEarning (CHEESE) framework, consolidating distributed resources among clients by device-to-device (D2D) communications, working in an intra-serial inter-parallel manner. InCHEESE, each learning client can form a cluster with its neighboring helping clients via D2D communications to train an FL model collaboratively. Inside each cluster, the model is split into multiple segments via a model splitting and allocation (MSA) strategy, while each cluster member trains one segment. After completing intra-cluster training, a transmission client (TC) is determined from each cluster to upload a complete model to the base station for global model aggregation under allocated bandwidth. Accordingly, an overall training delay cost minimization problem is formulated, involving the following subproblems: client clustering, MSA, TC selection, and bandwidth allocation. Due to its NP-Hardness, the problem is decoupled and solved iteratively. The client clustering problem is first transformed into a distributed clustering game based on potential game theory, where each cluster further investigates the remaining three subproblems to evaluate the utility of each clustering strategy. Specifically, a heuristic algorithm is proposed to solve the MSA problem under a given clustering strategy, while a greedy-based convex optimization approach is introduced to solve the joint TC selection and bandwidth allocation problem. Extensive experiments on practical models and datasets demonstrate thatCHEESEcan significantly reduce training delay costs.

Wireless caching improves the efficiency of content delivery in vehicular networks. However, the cached content needs to be frequently updated to adapt to dynamic service requirements, which is challenging due to high vehicle mobility and scarce radio resources. This paper investigates the online joint scheduling strategy of content files (CF) and server vehicles (SV) for cache content updating in a vehicle collaborative transmission scenario, where SVs act as content servers to deliver cached files to customer vehicles. The non-orthogonal multiple access (NOMA) technology is applied to push more CFs to SVs during a limited time interval, compared with the conventional orthogonal multiple access principle. Particularly, we define a utility function to reflect file value from local content popularity, user vehicle distribution, and file updating demand. By solving an "inversed" 0-1 knapsack problem, a suboptimal SV-CF scheduling strategy with the purpose of maximizing system utility is obtained, where the "inversed" means that the knapsack capacity represents current reserved power value. Simulation results illustrate the superiority of our proposed caching strategy.

Resource scheduling directly affects the resource utilization and network capacity in Internet of Vehicles (IoV). However, it’s challenging while vehicles are moving at high speed. We propose a reservation based resource blocks (RBs) allocation scheme for IoV networks. In the proposed scheme, Road Side Units (RSUs) dynamically allocates resources to users considering number of available RBs and arrival rate of service requests. We use inventory theory to formulate an optimization problem to decrease the rejection rate of service requests and improve RBs utilization. Simulation results show that the proposed scheme can achieve lower rejection rate of service requests and higher RBs utilization than the existing schemes.

The high mobility of vehicles in mobile edge computing (MEC) enabled vehicular networks causes the channel estimation error which will influence the quality of service (QoS) of users. In this paper, we explore the computation offloading and resource allocation in orthogonal frequency-division multiple access (OFDMA) based vehicular networks considering the imperfect channel state information (CSI). A mixed integer non-linear programming (MINLP) is formulated to minimize the offloading latency. To tackle this NP-hard problem, we divide the offloading and resource allocation into two subproblems which are offloading decision subproblem and resource allocation sub-problem. Specifically, given the offloading decision, we design a coalition game based algorithm to solve the subcarrier assignment problem and a convex optimization method to solve the power allocation problem. Meanwhile, given the resource allocation, we finally get the offloading decision by solving the linear programme (LP) problem. Numerical results show that the proposed scheme can significantly reduce the offloading latency.

Online video becomes a significant service in daily life, and it usually adopts a caching and playing mechanism. Due to high mobility and changeable topology, challenges of video downloading still exist in vehicular networks, especially in areas where the roadside units (RSUs) are not fully covered. The flexible deployment of the unmanned aerial vehicle (UAVs) compensate for the lack of RSU coverage, and thus this paper considers that a cyclic flight UAV to assist RSUs in providing video download services for vehicles. With the help of UAV, seamless communication coverage and stable transmission links ensure better service quality for vehicles. In addition, we propose a model-free algorithm based on a deep Q network to find the optimal UAV decision policy to achieve the minimized stalling time. Finally, the simulation results are given to demonstrate that the proposed solution can effectively maintain a high-quality user experience.

While home users can easily retrieve all kinds of contents onto their laptops or smartphones, vehicular users are constrained by intermittent connectivity to roadside units (RSUs). In this paper, we propose a cooperative downloading mechanism in homogeneous vehicular networks. In this mechanism, RSUs act as traffic managers to fetch proper data from the Internet and then distribute to vehicles in an approximately optimal manner. Specifically, based on vehicular mobility prediction and inter-node throughput estimation, a storage time aggregated graph (STAG) is constructed for planning transmission scheme, then an iterative greedy-driven algorithm is designed for deriving a suboptimal solution. Simulation results show that our approach can reduce 5% ~ 20% downloading time in an uniform distributed deployment scenario where the spacing between RSUs is not greater than 1500m.