Vehicle edge computing (VEC) brings abundant computing resources close to vehicles by deploying them at roadside units (RSUs) or base stations, thereby enabling diverse computation-intensive and delay-sensitive applications. Existing task offloading strategies are often computationally expensive to execute or generate suboptimal solutions. In this paper, we propose a novel learning-based approach, Knowledge-guided Attention-inspired Task Offloading (KATO), designed to efficiently offload tasks from moving vehicles to nearby RSUs. KATO integrates an attention-inspired encoder-decoder model for selecting a subset of RSUs that can reduce overall task processing time, along with an efficient iterative algorithm for computing optimal task allocation among the selected RSUs. Simulation results demonstrate that KATO achieves optimal or near-optimal performance with significantly lower computational overhead and generalizes well across networks of varying sizes and configurations.

Financial large language models (FinLLMs) offer immense potential for financial applications. While excessive deployment expenditures and considerable inference latency constitute major obstacles, as a prominent compression methodology, knowledge distillation (KD) offers an effective solution to these difficulties. A comprehensive survey is conducted in this work on how KD interacts with FinLLMs, covering three core aspects: strategy, application, and evaluation. At the strategy level, this review introduces a structured taxonomy to comparatively analyze existing distillation pathways. At the application level, this review puts forward a logical upstream–midstream–downstream framework to systematically explain the practical value of distilled models in the financial field. At the evaluation level, to tackle the absence of standards in the financial field, this review constructs a comprehensive evaluation framework that proceeds from multiple dimensions such as financial accuracy, reasoning fidelity, and robustness. In summary, this research aims to provide a clear roadmap for this interdisciplinary field, to accelerate the development of distilled FinLLMs.

The rise of edge computing has shifted computing resources closer to end-users, benefiting numerous delay-sensitive, computation-intensive applications. To speed up computation, distributed computing is a promising technique that allows parallel execution of computation tasks across multiple compute nodes. However, current research predominantly revolves around the master-worker paradigm, limiting resource sharing within one-hop neighborhoods. This limitation can render distributed computing ineffective in scenarios with limited nearby resources or constrained/dynamic connectivity. In this paper, we address this limitation by introducing a new distributed computing strategy that extends resource sharing beyond one-hop neighborhoods through exploring layered network structures and multi-hop routing. Our approach involves transforming the network graph into a sink tree and solving a joint optimization problem formulated based on the layered tree structure for task allocation and scheduling. Simulation results demonstrate a significant improvement over the traditional distributed computing and computation offloading strategies.

This abstract describes the project funded by the IEEE SMCS on Transforming Educational Assets and Materials (TEAM) in Systems, Man, and Cybernetics. The project develops an open course on Cyber Physical Systems (CPS) Foundation and Design for Unmanned Aerial Vehicles (UAVs). The course will be available to the public and serve the need of researchers, students and professionals who are interested in conducting UAVs related research. The open course contains integrated modules on control, communication and networking, computing, and artificial intelligence (AI) to provide trainees a comprehensive knowledge needed for UAVs. The course is self-paced and contains quizzes in each module for help students assess the quality of learning and also allow course designers to evaluate the effectiveness of the course materials for continuous improvement. The open course promotes CPS which is a SMCS technical field. It will also attract students and professionals to the SMC community.

Unmanned Aerial Vehicles (UAVs) have gained widespread use across various fields due to their flexibility and multifunctionality. However, their limited onboard computing capacity is often criticized for hindering their ability to execute complex tasks in real-time. To address this challenge, Networked Airborne Computing (NAC) has emerged, which leverages the collective computing power of multiple UAVs to enable efficient handling of large-scale data processing, real-time analytics, and complex mission coordination. Despite its potential, research in this area is still in its infancy. In this paper, we consider a typical NAC scenario where multiple UAVs with collision avoidance capabilities share resources while moving randomly within an area. Without prior knowledge of the system models, we aim to optimize task allocation among UAVs with uncertain mobility. To achieve this, we propose a Deep Reinforcement Learning algorithm based on the Twin Delayed Deep Deterministic Policy Gradient (TD3). Simulation results demonstrate that our approach significantly speeds up task execution compared to existing methods.

This paper considers the single source shortest path (SSSP) problem, which is the key for many applications such as navigation, mapping, routing, and social networking. Existing SSSP algorithms are designed mostly for shared-memory systems. Nevertheless, with the prevalence of diverse smart devices like drones, there is a growing interest in deploying SSSP algorithms over distributed computing systems so that they can run efficiently onboard of smart devices via Mobile Ad Hoc Computing or at the network edges via Mobile Edge Computing. In this paper, we introduce a communication-efficient ∆-stepping algorithm for distributed computing systems. The proposed algorithm is featured by 1) a message coordination architecture for reducing message exchanges between workers, 2) a pruning technique for reducing redundant computations, and 3) an aggregation technique for further reducing message exchanges when communication delay is significant. Theoretical analyses and experimental studies on real-world graph datasets demonstrate the promising performance of proposed algorithm.

Unmanned aerial vehicles (UAVs) based mobile edge computing (MEC) systems have attracted increasing research attention recently. They can provide on-demand computing services for ground users (GUs) without relying on any communication infrastructures and have the potential to provide better computing services with lower latency, compared with the conventional ground-based MEC or cloud-based systems. Considering the limited battery capacity of the UAVs, existing studies on UAV-based MEC have focused on using UAVs to serve GUs over small areas so that all tasks can be completed during a single flight. In this paper, we aim to remove this restriction and expand the range of users the UAV-based MEC system can serve, by integrating charge stations into the system. A joint task scheduling, routing, and charging problem is then formulated with the objective to minimize the total energy consumption, total service time, and total energy charged simultaneously. To solve this problem, we develop a mixed-integer programming (MIP) model and an equivalent mixed-integer linear programming (MILP) model. Comparative numerical studies demonstrate the optimal solutions found by the proposed approaches.

Multi-agent reinforcement learning (MARL) meth-ods face a curse of dimensionality in the policy and value function representations as the number of agents increases. The development of distributed or parallel training techniques is also hindered by the global coupling among the agent dynamics, requiring simultaneous state transitions. This paper introduces Distributed multi-Agent Reinforcement Learning with One-hop Neighbors (DARLIN). DARLIN is an off-policy actor-critic MARL method that breaks the curse of dimensionality and achieves distributed training by restricting the agent interactions to one-hop neighborhoods. Each agent optimizes its value and policy functions over a one-hop neighborhood, reducing the representation complexity, yet maintaining expressiveness by training with varying numbers and states of neighbors. This structure enables the key contribution of DARLIN: a distributed training procedure in which each compute node simulates the state transitions of only a small subset of the agents, greatly accelerating the training of large-scale MARL policies. Comparisons with state-of-the-art MARL methods show that DARLIN significantly reduces training time without sacrificing policy quality as the number of agents increases.

In recent years, we have witnessed a growing use of unmanned aerial vehicles (UAVs) in a variety of civil, commercial and military applications. Among these applications, many require the UAVs to scan or survey one or more regions, such as land monitoring, disaster assessment, search and rescue. To realize such applications, path planning is a key step. Although the coverage path planning (CPP) problem for a single region has been extensively studied in the literature, CPP for multiple regions has gained much less attention. This multi-regional CPP problem can be considered as a variant of the (multiple) traveling salesman problem (TSP) enhanced with CPP. Previously, we have studied the case of a single UAV. In this paper, we extend our previous studies to further consider multiple UAVs with energy constraints. To solve this new path planning problem, we develop two approaches: 1) a branch-and-bound (BnB) based approach that can find (near) optimal tours and 2) a genetic algorithm (GA) based approach that can solve large-scale problems efficiently under different objectives. Comprehensive theoretical analyses and computational experiments demonstrate the promising performance of the proposed approaches in terms of optimality and efficiency.

Federated learning (FL) is a promising machine learning paradigm because it allows distributed edge devices to collaboratively train a model without sharing their raw data. In practice, a major challenge to FL is that edge devices are heterogeneous, so slow devices may compromise the convergence of model training. To address such a challenge, several recent studies have suggested different solutions, in which a promising scheme is to utilize coded computing to facilitate the training of linear models. Nevertheless, the existing coded FL (CFL) scheme is limited by a fixed coding redundancy parameter, and a weight matrix used in the existing design may introduce unnecessary errors. In this paper, we tackle these issues and propose a novel framework, namely CFL-HC, to facilitate CFL in heterogeneous computing scenarios. In our framework, we consider a computing system consisting of a central server and multiple computing devices with original or coded datasets. Then we specify an expected number of input-output pairs that are used in one round. Within such a framework, we formulate an optimization problem to find the best deadline of each training round and the optimal size of the computing task allocated to each computing device. We then design a two-step optimization scheme to obtain the optimal solution. To evaluate the proposed framework, we develop a real CFL system using the message passing interface platform. Based on this system, we conduct numerical experiments, which demonstrate the advantages of the proposed framework, in terms of both accuracy and convergence speed.

Mobile ad hoc computing (MAHC), which allows mobile devices to directly share their computing resources, is a promising solution to address the growing demands for computing resources required by mobile devices. However, offloading a computation task from a mobile device to other mobile devices is a challenging task due to frequent topology changes and link failures because of node mobility, unstable and unknown communication environments, and the heterogeneous nature of these devices. To address these challenges, in this paper, we introduce a novel coded computation scheme based on multi-agent reinforcement learning (MARL), which has many promising features such as adaptability to network changes, high efficiency and robustness to uncertain system disturbances, consideration of node heterogeneity, and decentralized load allocation. Comprehensive simulation studies demonstrate that the proposed approach can outperform state-of-the-art distributed computing schemes.

This paper aims to mitigate straggler effects in synchronous distributed learning for multi-agent reinforcement learning (MARL) problems. Stragglers arise frequently in a distributed learning system, due to the existence of various system disturbances such as slow-downs or failures of compute nodes and communication bottlenecks. To resolve this issue, we propose a coded distributed learning framework, which speeds up the training of MARL algorithms in the presence of stragglers, while maintaining the same accuracy as the centralized approach. As an illustration, a coded distributed version of the multi-agent deep deterministic policy gradient (MADDPG) algorithm is developed and evaluated. Different coding schemes, including maximum distance separable (MDS) code, random sparse code, replication-based code, and regular low density parity check (LDPC) code are also investigated. Simulations in several multi-robot problems demonstrate the promising performance of the proposed framework.

In recent years, we have witnessed fast‐growing unmanned aerial systems (UAS) based applications. To better facilitate these applications, many efforts have been made to enhance the capability of UAS from various aspects, including communications, control and networking, and so on. Nevertheless, most of these studies neglect the computation aspect. Recently, the UAS‐enabled mobile edge computing (MEC) has attracted increasing research attention, which utilises UAS with onboard computing capability to provide on‐demand computing services for mobile users. However, existing research on UAS‐enabled MEC remains at the theory stage and how to design a UAS platform with advanced onboard computing capability has not been addressed. In this study, the authors aim to fill this research gap and design an open UAS‐based airborne computing platform with advanced onboard computing capability. This platform was designed from three aspects: hardware, software, and applications. In particular, feasible computing hardware to perform UAS onboard computing is first considered and a prototype is then designed. To enhance the flexibility and programmability of the platform, two key virtualisation techniques are then investigated. Finally, they test the performance of their prototype by executing real UAS onboard computing tasks, the results of which verify the feasibility and potentials of the proposed airborne computing platform.

In this paper, we study properties of distributed consensus in layered sensor networks of the multi-layer multi-group (MLMG) structure. We show that properly designed MLMG networks maintain decentralized communication, whereas show the advantage of centralized structures. In particular, they require less number of transmissions required to reach consensus. This feature is critical for efficient distributed computing in large-scale sensor network applications. For typical classes of MLMG networks, we mathematically characterize the reduced number of transmissions compared to equivalent egalitarian decentralized structures of the same consensus dynamics. This explicit characterization based on simple graphical characteristics of MLMG structures permits an efficient design of large-scale network structures to meet desired performance requirements. In addition, we characterize the asymptotic and transient properties of consensus in MLMG networks of limited channel rates, using the probabilistic quantization schemes.

Modern large-scale infrastructure systems have typical complicated structure and dynamics, and extensive simulations are required to evaluate their performance. The probabilistic collocation method (PCM) has been developed to effectively simulate a system's performance under parametric uncertainty. In particular, it allows reduced-order representation of the mapping between uncertain parameters and system performance measures/outputs, using only a limited number of simulations; the resultant representation of the original system is provably accurate over the likely range of parameter values. In this paper, we extend the formal analysis of single-variable PCM to the multivariate case, where multiple uncertain parameters may or may not be independent. Specifically, we provide conditions that permit multivariate PCM to precisely predict the mean of original system output. We also explore additional capabilities of the multivariate PCM, in terms of cross-statistics prediction, relation to the minimum mean-square estimator, computational feasibility for large dimensional parameter sets, and sample-based approximation of the solution. At the end of the paper, we demonstrate the application of multivariate PCM in evaluating air traffic system performance under weather uncertainties.