Deploying Unmanned Aerial Vehicles (UAVs) as aerial base stations enhances the coverage and performance of communication networks in Vehicular Edge Computing (VEC) scenarios. However, due to the limited communication range and energy capacity of UAVs, they cannot continuously cover entire areas or sustain long flights. Therefore, achieving full communication coverage of a target area with a minimal number of UAVs and efficient task offloading remains a significant challenge. To address this problem, the UAV-assisted Two-stage Intelligent Collaboration (UTIC) method is proposed in this paper to tackle the joint position optimization and task scheduling issue. Firstly, a UAV-assisted Two-stage Task Scheduling (UTTS) system model is designed to optimize the allocation process. Secondly, an Enhanced Particle Swarm Optimization (E-PSO) algorithm is designed to determine the optimal positions of UAVs, ensuring complete coverage of all mobile vehicles (MVs) with the minimum number of UAVs. Thirdly, Deep Deterministic Policy Gradient (DDPG) method is employed to find the optimal scheduling decisions for MVs, considering energy consumption, delay, and task priorities. Simulation results demonstrate that the proposed UTIC method can achieve nearly 20% reduction in UAV deployment and outperform three other classical reinforcement learning (RL) algorithms in terms of reducing system cost.

In a wireless ad hoc network, the size of the virtual backbone (VB) is an important factor for measuring the quality of the VB. The smaller the VB is, the less the overhead caused by the VB. Since ball graphs (BGs) have been used to model 3-dimensional wireless ad hoc networks and since a connected dominating set can be used to represent a VB undertaking routing-related tasks, the problem of finding the smallest VB is transformed into the problem of finding a minimum connected dominating set (MCDS). Many research results on the MCDS problem have been obtained for unit disk graphs and unit ball graphs, in which the transmission ranges of all nodes are identical. In some situations, the node powers can vary. One can model such a network as a graph with different transmission ranges for different nodes. In this paper, we focus on the problem of minimum strongly connected dominating and absorbing sets (MSCDASs) in a strongly connected directed ball graph with different transmission ranges, which is also NP-hard. We design an algorithm considering the construction of a strongly connected dominating and absorbing set (SCDAS), whose size does not exceed 319/15k3+116/5k2+29/5kopt+29/3k3+116/5k2+87/5k+13/15, where opt is the size of an MCDAS and k denotes the ratio of rmax to rmin in the ad hoc network with transmission range rmin,rmax. Our simulations show the feasibility of the algorithm proposed in this paper.

Accurate assessment of cardiac function is vital for preventing and managing cardiovascular diseases. Recent advancements in machine learning, especially convolutional neural networks, and multi-task learning, have improved the precision of echocardiogram evaluations. However, existing methods often overlook the intrinsic relationships among ejection fraction, end-diastolic volume, and end-systolic volume, which are essential for accurate assessments. We propose a Non-Invasive Framework for Heart Function Assessment using Multi-Task Learning that utilizes a 3D convolutional neural network to extract key spatiotemporal features from echocardiogram videos. By employing a multi-task learning strategy and weight distribution mechanism, this framework enhances the accuracy of ejection fraction predictions and provides a comprehensive assessment of cardiac structure and function. This approach ensures that the model effectively integrates auxiliary task information while focusing on the primary task, resulting in a more precise analysis of cardiac function. The experimental results on the Echonet- Dynamic dataset demonstrate that our method achieves an average absolute error of 3.89, a root mean square error of 5.13, and an R² value of 0.82, outperforming existing methods. Future work will focus on automatic weight optimization, model compression, and improving computational efficiency for broader clinical applications.

In intent-based vehicular edge computing (VEC) networking, escalating demands for computational offloading and resource management in dynamic urban environments necessitate innovative solutions. This article proposes a novel reconfigurable intelligent surface (RIS)-assisted non-orthogonal multiple access (NOMA)-VEC framework that empowers vehicle users (VUs) to offload arbitrary task portions to multiple edge servers via any available subcarrier. This approach overcomes limitations posed by heterogeneous local computing capabilities and stringent latency constraints. By leveraging RISs to enhance channel conditions through both direct and reflected links, our framework significantly improves communication reliability and offloading efficiency. To minimize the average weighted energy consumption of VUs under time-varying channels and traffic dynamics, we formulate a joint optimization problem integrating offloading decisions, power allocation, and transmission time scheduling. Addressing the problem’s inherent complexity, characterized by multivariable coupling and nonconvex constraints, we develop a two-stage decomposition strategy: Offloading decisions are dynamically adapted to environmental fluctuations using a proximal policy optimization (PPO)-based algorithm, while resource allocation is resolved through a hybrid genetic algorithm (GA) and sequential least squares programming (SLSQP) approach, efficiently navigating combinatorial and nonconvex landscapes. Extensive simulations demonstrate that our framework reduces VU energy consumption by 11.12% compared with baseline methods, validating its superior efficiency in RIS-enhanced VEC systems.

The demands from the Internet of Everything (IoE) ecosystem led to the emergence of the sixth-generation (6G) networks. Multi-access Edge Computing (MEC) provides an efficient solution for enhancing the performance of integrated sensing and communications within the 6G IoE through offloading optimization. Nevertheless, existing offloading decision methods encounter difficulties in balancing the computational load within the system and obtaining a globally optimal strategy. In this paper, a Deep Reinforcement Learning based Collaborative Offloading Optimization (DRL-COO) method is proposed to solve the offloading optimization problem in the 6G-enabled Multi-user Collaborative Task Offloading MEC (MU-CTOMEC) system, which aims to minimize delay, energy consumption, and task failure rate. Specifically, this method establishes a Markov Decision Process (MDP) model for collaborative task offloading and trains it through interaction with the environment using a Deep Reinforcement Learning (DRL) learning algorithm based on an asynchronous priority experience replay mechanism. The experimental results demonstrate the superiority of our proposed optimization algorithm over state-of-the-art methods in terms of average computation time, energy consumption, and task failure rate.