Satellite communication environments, characterized by dynamic links, exacerbate packet reordering, which leads to fundamental limitations in QUIC’s loss detection mechanism. A fixed packet threshold may misinterpret reordered packets as lost, leading to spurious retransmissions and throughput degradation. To address the challenge, we propose GRU-QUIC, a machine learning-enhanced QUIC protocol with packet reordering resilience for satellite networks. First, we model the relationship between the probability of spurious packet loss due to reordering and the throughput of QUIC, providing a theoretical basis for performance optimization. Second, we design a gated recurrent unit (GRU)-based prediction module on the receiver side to predict the next slot packet reordering displacement by learning the temporal correlation of packet reordering sequences. The sender then uses the reordering displacement indicated by the GRU to dynamically adjust the packet threshold of loss detection, effectively alleviating the problem of packet loss misjudgment due to packet reordering. Experiments demonstrate that GRU-QUIC significantly enhances the performance of QUIC transmission while maintaining protocol compatibility. In a simulated satellite network, GRU-QUIC effectively reduces unnecessary retransmissions and shortens file download completion time compared to standard QUIC.

This paper aims to develop flexible and scalable scene understanding capabilities usable on resource-constrained devices. We propose an edge computing framework to enable large-scale object detection within complex visual scenes. The framework attempts to emulate human intuition - leveraging commonsense reasoning and fast concept learning from small data. Structured knowledge about objects and their relationships guides efficient deduction of scene contents. Specifically, we develop an ontology-based scene deduction model to refine an initial scene parsing, which incorporates declarative constraints on objects, their attributes, and interactions. It represents such domain knowledge explicitly and reasons over ontological relationships. By representing this domain knowledge explicitly and reasoning over ontological relationships, the model refines an initial scene parsing to better reflect commonsense consistency. The evaluation compared the proposed method to other large-scale object detection methods. The results show that our ontology-driven deduction model scale-ups the data-driven scene parser’s capability while reducing the data and compute requirements.

With the advent of Web3.0 and Metaverse technologies, there is a growing interest in integrating web and metaverse environments to create immersive, cross-platform "Webverse" experiences. Despite active research in this area, the lack of standardized testing approaches has hindered the development of robust Webverse applications. This paper addresses this gap by proposing a comprehensive framework for testing Web-Metaverse integrations. We analyze existing Webverse platforms, characterize different stages and scenes through network data analysis, and design a suite of standardized tests. Furthermore, we implement a one-stop benchmarking tool tailored for web-based metaverse applications. The fidelity of our tool is validated through comparisons with human testing results. To demonstrate its applicability, we utilize the proposed testing process to evaluate and compare Webverse implementations on popular platforms like Galacean and Mozilla Hubs, deployed on their respective servers. The outcomes highlight platform-specific differences and the efficacy of our standardized approach in enabling seamless Web-Metaverse convergence.

The design of a virtual exhibition layout is a meticulous process. It usually starts with floor plan design, adhering to guidelines such as visitor traffic flow, exhibit prominence, and aesthetic considerations. Based on floor plans, 3D models are constructed, followed by the creation of final scenes. However, there is a notable lack of an efficient method to streamline this complex design process, which often requires profound professional expertise. To address this gap, we propose a novel approach leveraging generative AI and a floor plan recognition algorithm to reconstruct 3D exhibition layouts, thereby enhancing designers’ efficiency. We have achieved the design of exhibition hall floor plans using Stable Diffusion and completed the automatic conversion from floor plans to 3D models, as well as the web-based operation and display platform. We undertake a comparative analysis between our solution and a CGAN-based method, evaluating the results through both subjective and objective measures. Furthermore, we present a comprehensive flowchart outlining the process of generating a museum exhibition hall on a web platform. Additionally, we demonstrate the design impact of our methodology in creating a virtual exhibition hall utilizing Web3D technology.

In recent years, the emergence of deep learning has brought Physics-Informed Neural Networks (PINNs) into the spotlight as a promising method for solving partial differential equations (PDEs). Despite the attention received by PINNs, their accuracy and convergence face significant challenges, particularly when dealing with complex equations containing multiple components. In PDEs, due to their inherent rigidity, distinct components may converge at varying rates, resulting in an unbalanced convergence process. This imbalance during optimization has the potential to yield suboptimal solutions. We introduce an importance-guided sequential training method to regulate the competition of different components in PDEs. The importance can be quantitatively defined through correlation analysis, enabling the formulation of an algorithm that systematically instructs neural networks to acquire insights from individual terms in accordance with their respective importance levels. To evaluate the performance of our proposed approach, we applied it to solve the Burgers equation and the Klein-Gordon equation. The results clearly demonstrate that our proposed approaches outperform the original model, showcasing the effectiveness of our internal weighting method in improving the accuracy and convergence of PINNs. Furthermore, our research serves as a stepping stone for exploring and harnessing the power of correlation analysis in the realm of PINNs.

Edge computing places the computational services and resources closer to the user proximity, to reduce latency, and ensure the quality of service and experience. Low latency, context awareness and mobility support are the major contributors to edge-enabled smart systems. Such systems require handling new situations and change on the fly and ensuring the quality of service while only having access to constrained computation and communication resources and operating in mobile, dynamic and ever-changing environments. Hence, adaptation and self-organisation are crucial for such systems to maintain their performance, and operability while accommodating new changes in their environment. This article reviews the current literature in the field of adaptive edge computing systems. We use a widely accepted taxonomy, which describes the important aspects of adaptive behaviour implementation in computing systems. This taxonomy discusses aspects such as adaptation reasons, the various levels an adaptation strategy can be implemented, the time of reaction to a change, categories of adaptation technique and control of the adaptive behaviour. In this article, we discuss how these aspects are addressed in the literature and identify the open research challenges and future direction in adaptive edge computing systems. The results of our analysis show that most of the identified approaches target adaptation at the application level, and only a few focus on middleware, communication infrastructure and context. Adaptations that are required to address the changes in the context, changes caused by users or in the system itself are also less explored. Furthermore, most of the literature has opted for reactive adaptation, although proactive adaptation is essential to maintain the edge computing systems’ performance and interoperability by anticipating the required adaptations on the fly. Additionally, most approaches apply a centralised adaptation control, which does not perfectly fit the mostly decentralised/distributed edge computing settings.

Creating memory representations is essential for developing viable navigation strategies for intelligent agents. Although neural radiance fields (NeRFs) have shown great promise as a novel method for spatial representation, their potential for integration into learning-based navigation as a memory structure has been largely overlooked in the existing literature. In this article, we introduce a navigation pipeline that incorporates NeRF into visuomotor navigation. Initially, we present a derivative radiance field that facilitates one-shot pose and depth estimation from a single query image. By assuming equivalence between density and space occupancy, we generate a geometric accessibility map based on an offline-constructed NeRF prior. Utilizing the above information, we design a global planner that decomposes long-term tasks by performing waypoint estimation and rendering. Finally, we employ an imitation-learned local controller to achieve a reliable navigation policy. Our pipeline effectively utilizes NeRF's compact spatial representation for task decomposition and action generation, enabling efficient navigation. Experimental results highlight its advantages over recent implicit and explicit memory approaches in image-goal navigation tasks. Moreover, we conduct interpretability studies and apply our algorithm in real-world scenarios to further attest to its practicality and effectiveness.

Fog computing has emerged as a promising solution for service provisioning. To support the increasing number of smart end devices and their requirement for intelligent data processing, numerous learning-based applications will be deployed at the edge network to provide services with an acceptable level of latency. However, such applications assume that the training and the actual data are independent and identically distributed, which is impractical in the dynamic and heterogeneous environment of the real world. This makes service matchmaking a challenge because even if the service functionality matches user requirements, the mismatched actual data will inevitably result in the degradation of service quality. This paper targets the mismatching issue of learning-based applications and the high requirement of intelligent data processing at the network edge and proposes a novel service provisioning model. This model introduces a novel service description model to resolve data mismatch, a clustering algorithm that pre-processes requests to deal with high concurrency requirements, and a heuristic joint service searching method to reduce traffic costs. This work was evaluated through a case study and simulations. Evaluation results show that the proposed model can achieve a good quality of services and reduce response time as well as traffic costs.

To cope with the high communication overhead caused by frequent aggregation of Federated Learning (FL) in Multi-access Edge Computing (MEC) scenarios, Hierarchical Federated Edge Learning (HFEL) is proposed as an evolving framework. HFEL offloads tasks to edge servers for partial model aggregation to reduce network traffic. However, most of the existing research focuses on resource optimization for HFEL without considering the impact of data characteristics and cannot guarantee the quality of FL training. To this end, we propose a task offloading approach based on data and resource heterogeneity under HFEL to improve training performance and reduce system cost. Specifically, we leverage information entropy to incorporate data statistical features into the cost function to reshape edge datasets. In addition, we applied Multi-Agent Deep Deterministic Policy Gradient (MADDPG) with a resource allocation module to generate distributed offloading policy more efficiently. Our algorithm not only adopts local observations to obtain the optimal action but also takes into account device heterogeneity, which can adapt to the unstable edge environment. Extensive experiments under multiple datasets and baselines are carried out, which demonstrate that our algorithm can effectively improve the accuracy of aggregated models while reducing system cost.

Through offloading the computing tasks of the task nodes (TNs) to the fog nodes (FNs) located at the network edge, the fog network is expected to address the unacceptable processing delay and heavy link burden existed in current cloud-based networks. Unlike most existing researches based on the command-mode offloading and full capability report, this paper develops a general analytical model of the task scheduling among voluntary nodes (VNs) in fog networks, wherein the VNs voluntarily contribute their capabilities for serving their neighboring TNs. A novel delay-optimal task scheduling (DOTS) algorithm is proposed to obtain the delay-optimal offloading solution according to the reported capabilities of the VNs. Extensive simulations are carried out in a fog network, and the numerical results indicate that the proposed DOTS algorithm can effectively provide the optimal set of the helper nodes, subtask sizes, and the TN transmission power to minimize the overall task processing delay. Moreover, compared with the command-mode offloading, the voluntary-mode achieves more balanced offloading and a higher fairness level among the FNs.

Mobile, pervasive computing environments respond to users’ requirements by providing access to and composition of various services over networked devices. In such an environment, service composition needs to satisfy a request’s goal, and be mobile-aware even throughout service discovery and service execution. A composite service also needs to be adaptable to cope with the environment’s dynamic network topology. Existing composition solutions employ goal-oriented planning to provide flexible composition, and assign service providers at runtime, to avoid composition failure. However, these solutions have limited support for complex service flows and composite service adaptation. This paper proposes a self-organizing, goal-driven service model for task resolution and execution in mobile pervasive environments. In particular, it proposes a decentralized heuristic planning algorithm based on backward-chaining to support flexible service discovery. Further, we introduce an adaptation architecture that allows execution paths to dynamically adapt, which reduces failures, and lessens re-execution effort for failure recovery. Simulation results show the suitability of the proposed mechanism in pervasive computing environments where providers are mobile, and it is uncertain what services are available. Our evaluation additionally reveals the model’s limits with regard to network dynamism and resource constraints.