arXiv subjects

Hailiang Zhao

Publications and source records attributed to Hailiang Zhao.

At least 37 records · Page 2Linked to original sources

arXiv2026

Adaptive Dual-Weighting Framework for Federated Learning via Out-of-Distribution Detection

Federated Learning (FL) enables collaborative model training across large-scale distributed service nodes while preserving data privacy, making it a cornerstone of intelligent service systems in edge-cloud environments. However, in real-world service-oriented deployments, data generated by heterogeneous users, devices, and application scenarios are inherently non-IID. This severe data heterogeneity critically undermines the convergence stability, generalization ability, and ultimately the quality of service delivered by the global model. To address this challenge, we propose FLood, a novel FL framework inspired by out-of-distribution (OOD) detection. FLood dynamically counteracts the adverse effects of heterogeneity through a dual-weighting mechanism that jointly governs local training and global aggregation. At the client level, it adaptively reweights the supervised loss by upweighting pseudo-OOD samples, thereby encouraging more robust learning from distributionally misaligned or challenging data. At the server level, it refines model aggregation by weighting client contributions according to their OOD confidence scores, prioritizing updates from clients with higher in-distribution consistency and enhancing the global model's robustness and convergence stability. Extensive experiments across multiple benchmarks under diverse non-IID settings demonstrate that FLood consistently outperforms state-of-the-art FL methods in both accuracy and generalization. Furthermore, FLood functions as an orthogonal plug-in module: it seamlessly integrates with existing FL algorithms to boost their performance under heterogeneity without modifying their core optimization logic. These properties make FLood a practical and scalable solution for deploying reliable intelligent services in real-world federated environments.

cs.LG↗

arXiv2025

TrackTeller: Temporal Multimodal 3D Grounding for Behavior-Dependent Object References

Understanding natural-language references to objects in dynamic 3D driving scenes is essential for interactive autonomous systems. In practice, many referring expressions describe targets through recent motion or short-term interactions, which cannot be resolved from static appearance or geometry alone. We study temporal language-based 3D grounding, where the objective is to identify the referred object in the current frame by leveraging multi-frame observations. We propose TrackTeller, a temporal multimodal grounding framework that integrates LiDAR-image fusion, language-conditioned decoding, and temporal reasoning in a unified architecture. TrackTeller constructs a shared UniScene representation aligned with textual semantics, generates language-aware 3D proposals, and refines grounding decisions using motion history and short-term dynamics. Experiments on the NuPrompt benchmark demonstrate that TrackTeller consistently improves language-grounded tracking performance, outperforming strong baselines with a 70% relative improvement in Average Multi-Object Tracking Accuracy and a 3.15-3.4 times reduction in False Alarm Frequency.

cs.CV↗

arXiv2025

PeerSync: Accelerating Containerized Service Delivery at the Network Edge

Efficient container image distribution is crucial for enabling machine learning inference at the network edge, where resource limitations and dynamic network conditions create significant challenges. In this paper, we present PeerSync, a decentralized P2P-based system designed to optimize image distribution in edge environments. PeerSync employs a popularity- and network-aware download engine that dynamically adapts to content popularity and real-time network conditions. PeerSync further integrates automated tracker election for rapid peer discovery and dynamic cache management for efficient storage utilization. We implement PeerSync with 8000+ lines of Rust code and test its performance extensively on both large-scale Docker-based emulations and physical edge devices. Experimental results show that PeerSync delivers a remarkable speed increase of 2.72$\times$, 1.79$\times$, and 1.28$\times$ compared to the Baseline solution, Dragonfly, and Kraken, respectively, while significantly reducing cross-network traffic by 90.72% under congested and varying network conditions.

cs.NI↗

arXiv2025

Robustifying Learning-Augmented Caching Efficiently without Compromising 1-Consistency

The online caching problem aims to minimize cache misses when serving a sequence of requests under a limited cache size. While naive learning-augmented caching algorithms achieve ideal $1$-consistency, they lack robustness guarantees. Existing robustification methods either sacrifice $1$-consistency or introduce excessive computational overhead. In this paper, we introduce Guard, a lightweight robustification framework that enhances the robustness of a broad class of learning-augmented caching algorithms to $2H_{k-1} + 2$, while preserving their $1$-consistency. Guard achieves the current best-known trade-off between consistency and robustness, with only O(1) additional per-request overhead, thereby maintaining the original time complexity of the base algorithm. Extensive experiments across multiple real-world datasets and prediction models validate the effectiveness of Guard in practice.

cs.DS↗

arXiv2025

OmniFuser: Adaptive Multimodal Fusion for Service-Oriented Predictive Maintenance

Accurate and timely prediction of tool conditions is critical for intelligent manufacturing systems, where unplanned tool failures can lead to quality degradation and production downtime. In modern industrial environments, predictive maintenance is increasingly implemented as an intelligent service that integrates sensing, analysis, and decision support across production processes. To meet the demand for reliable and service-oriented operation, we present OmniFuser, a multimodal learning framework for predictive maintenance of milling tools that leverages both visual and sensor data. It performs parallel feature extraction from high-resolution tool images and cutting-force signals, capturing complementary spatiotemporal patterns across modalities. To effectively integrate heterogeneous features, OmniFuser employs a contamination-free cross-modal fusion mechanism that disentangles shared and modality-specific components, allowing for efficient cross-modal interaction. Furthermore, a recursive refinement pathway functions as an anchor mechanism, consistently retaining residual information to stabilize fusion dynamics. The learned representations can be encapsulated as reusable maintenance service modules, supporting both tool-state classification (e.g., Sharp, Used, Dulled) and multi-step force signal forecasting. Experiments on real-world milling datasets demonstrate that OmniFuser consistently outperforms state-of-the-art baselines, providing a dependable foundation for building intelligent industrial maintenance services.

cs.AI↗

arXiv2025

Data-Locality-Aware Task Assignment and Scheduling for Distributed Job Executions

This paper addresses the data-locality-aware task assignment and scheduling problem for distributed job executions. Our goal is to minimize job completion times without prior knowledge of future job arrivals. We propose an Optimal Balanced Task Assignment algorithm (OBTA), which achieves minimal job completion times while significantly reducing computational overhead through efficient narrowing of the solution search space. To balance performance and efficiency, we extend the approximate Water-Filling (WF) algorithm, providing a rigorous proof that its approximation factor equals the number of task groups in a job. We also introduce a novel heuristic, Replica-Deletion (RD), which outperforms WF by leveraging global optimization techniques. To further enhance scheduling efficiency, we incorporate job ordering strategies based on a shortest-estimated-time-first policy, reducing average job completion times across workloads. Extensive trace-driven evaluations validate the effectiveness and scalability of the proposed algorithms.

cs.DC↗

arXiv2025

CADRef: Robust Out-of-Distribution Detection via Class-Aware Decoupled Relative Feature Leveraging

Deep neural networks (DNNs) have been widely criticized for their overconfidence when dealing with out-of-distribution (OOD) samples, highlighting the critical need for effective OOD detection to ensure the safe deployment of DNNs in real-world settings. Existing post-hoc OOD detection methods primarily enhance the discriminative power of logit-based approaches by reshaping sample features, yet they often neglect critical information inherent in the features themselves. In this paper, we propose the Class-Aware Relative Feature-based method (CARef), which utilizes the error between a sample's feature and its class-aware average feature as a discriminative criterion. To further refine this approach, we introduce the Class-Aware Decoupled Relative Feature-based method (CADRef), which decouples sample features based on the alignment of signs between the relative feature and corresponding model weights, enhancing the discriminative capabilities of CARef. Extensive experimental results across multiple datasets and models demonstrate that both proposed methods exhibit effectiveness and robustness in OOD detection compared to state-of-the-art methods. Specifically, our two methods outperform the best baseline by 2.82% and 3.27% in AUROC, with improvements of 4.03% and 6.32% in FPR95, respectively.

cs.CV↗

arXiv2025

SeMi: When Imbalanced Semi-Supervised Learning Meets Mining Hard Examples

Semi-Supervised Learning (SSL) can leverage abundant unlabeled data to boost model performance. However, the class-imbalanced data distribution in real-world scenarios poses great challenges to SSL, resulting in performance degradation. Existing class-imbalanced semi-supervised learning (CISSL) methods mainly focus on rebalancing datasets but ignore the potential of using hard examples to enhance performance, making it difficult to fully harness the power of unlabeled data even with sophisticated algorithms. To address this issue, we propose a method that enhances the performance of Imbalanced Semi-Supervised Learning by Mining Hard Examples (SeMi). This method distinguishes the entropy differences among logits of hard and easy examples, thereby identifying hard examples and increasing the utility of unlabeled data, better addressing the imbalance problem in CISSL. In addition, we maintain a class-balanced memory bank with confidence decay for storing high-confidence embeddings to enhance the pseudo-labels' reliability. Although our method is simple, it is effective and seamlessly integrates with existing approaches. We perform comprehensive experiments on standard CISSL benchmarks and experimentally demonstrate that our proposed SeMi outperforms existing state-of-the-art methods on multiple benchmarks, especially in reversed scenarios, where our best result shows approximately a 54.8\% improvement over the baseline methods.

cs.CV↗

arXiv2024

Decentralized Proactive Model Offloading and Resource Allocation for Split and Federated Learning

In the resource-constrained IoT-edge computing environment, Split Federated (SplitFed) learning is implemented to enhance training efficiency. This method involves each terminal device dividing its full DNN model at a designated layer into a device-side model and a server-side model, then offloading the latter to the edge server. However, existing research overlooks four critical issues as follows: (1) the heterogeneity of end devices' resource capacities and the sizes of their local data samples impact training efficiency; (2) the influence of the edge server's computation and network resource allocation on training efficiency; (3) the data leakage risk associated with the offloaded server-side sub-model; (4) the privacy drawbacks of current centralized algorithms. Consequently, proactively identifying the optimal cut layer and server resource requirements for each end device to minimize training latency while adhering to data leakage risk rate constraint remains a challenging issue. To address these problems, this paper first formulates the latency and data leakage risk of training DNN models using Split Federated learning. Next, we frame the Split Federated learning problem as a mixed-integer nonlinear programming challenge. To tackle this, we propose a decentralized Proactive Model Offloading and Resource Allocation (DP-MORA) scheme, empowering each end device to determine its cut layer and resource requirements based on its local multidimensional training configuration, without knowledge of other devices' configurations. Extensive experiments on two real-world datasets demonstrate that the DP-MORA scheme effectively reduces DNN model training latency, enhances training efficiency, and complies with data leakage risk constraints compared to several baseline algorithms across various experimental settings.

cs.DC↗

arXiv2024

Learning-Augmented Algorithms for the Bahncard Problem

In this paper, we study learning-augmented algorithms for the Bahncard problem. The Bahncard problem is a generalization of the ski-rental problem, where a traveler needs to irrevocably and repeatedly decide between a cheap short-term solution and an expensive long-term one with an unknown future. Even though the problem is canonical, only a primal-dual-based learning-augmented algorithm was explicitly designed for it. We develop a new learning-augmented algorithm, named PFSUM, that incorporates both history and short-term future to improve online decision making. We derive the competitive ratio of PFSUM as a function of the prediction error and conduct extensive experiments to show that PFSUM outperforms the primal-dual-based algorithm.

cs.LG↗

arXiv2023

Tail-Learning: Adaptive Learning Method for Mitigating Tail Latency in Autonomous Edge Systems

In the realm of edge computing, the increasing demand for high Quality of Service (QoS), particularly in dynamic multimedia streaming applications (e.g., Augmented Reality/Virtual Reality and online gaming), has prompted the need for effective solutions. Nevertheless, adopting an edge paradigm grounded in distributed computing has exacerbated the issue of tail latency. Given a limited variety of multimedia services supported by edge servers and the dynamic nature of user requests, employing traditional queuing methods to model tail latency in distributed edge computing is challenging, substantially exacerbating head-of-line (HoL) blocking. In response to this challenge, we have developed a learning-based scheduling method to mitigate the overall tail latency, which adaptively selects appropriate edge servers for execution as incoming distributed tasks vary with unknown size. To optimize the utilization of the edge computing paradigm, we leverage Laplace transform techniques to theoretically derive an upper bound for the response time of edge servers. Subsequently, we integrate this upper bound into reinforcement learning to facilitate tail learning and enable informed decisions for autonomous distributed scheduling. The experiment results demonstrate the efficiency in reducing tail latency compared to existing methods.

cs.DC↗

arXiv2023

Scheduling Multi-Server Jobs with Sublinear Regrets via Online Learning

Multi-server jobs that request multiple computing resources and hold onto them during their execution dominate modern computing clusters. When allocating the multi-type resources to several co-located multi-server jobs simultaneously in online settings, it is difficult to make the tradeoff between the parallel computation gain and the internal communication overhead, apart from the resource contention between jobs. To study the computation-communication tradeoff, we model the computation gain as the speedup on the job completion time when it is executed in parallelism on multiple computing instances, and fit it with utilities of different concavities. Meanwhile, we take the dominant communication overhead as the penalty to be subtracted. To achieve a better gain-overhead tradeoff, we formulate an cumulative reward maximization program and design an online algorithm, named OGASched, to schedule multi-server jobs. OGASched allocates the multi-type resources to each arrived job in the ascending direction of the reward gradients. It has several parallel sub-procedures to accelerate its computation, which greatly reduces the complexity. We proved that it has a sublinear regret with general concave rewards. We also conduct extensive trace-driven simulations to validate the performance of OGASched. The results demonstrate that OGASched outperforms widely used heuristics by $11.33\%$, $7.75\%$, $13.89\%$, and $13.44\%$, respectively.

cs.DC↗

arXiv2023

Cloud-Native Computing: A Survey from the Perspective of Services

The development of cloud computing delivery models inspires the emergence of cloud-native computing. Cloud-native computing, as the most influential development principle for web applications, has already attracted increasingly more attention in both industry and academia. Despite the momentum in the cloud-native industrial community, a clear research roadmap on this topic is still missing. As a contribution to this knowledge, this paper surveys key issues during the life-cycle of cloud-native applications, from the perspective of services. Specifically, we elaborate the research domains by decoupling the life-cycle of cloud-native applications into four states: building, orchestration, operate, and maintenance. We also discuss the fundamental necessities and summarize the key performance metrics that play critical roles during the development and management of cloud-native applications. We highlight the key implications and limitations of existing works in each state. The challenges, future directions, and research opportunities are also discussed.

cs.DC↗

arXiv2022

Learning to Schedule Multi-Server Jobs with Fluctuated Processing Speeds

Multi-server jobs are imperative in modern cloud computing systems. A noteworthy feature of multi-server jobs is that, they usually request multiple computing devices simultaneously for their execution. How to schedule multi-server jobs online with a high system efficiency is a topic of great concern. Firstly, the scheduling decisions have to satisfy the service locality constraints. Secondly, the scheduling decisions needs to be made online without the knowledge of future job arrivals. Thirdly, and most importantly, the actual service rate experienced by a job is usually in fluctuation because of the dynamic voltage and frequency scaling (DVFS) and power oversubscription techniques when multiple types of jobs co-locate. A majority of online algorithms with theoretical performance guarantees are proposed. However, most of them require the processing speeds to be knowable, thereby the job completion times can be exactly calculated. To present a theoretically guaranteed online scheduling algorithm for multi-server jobs without knowing actual processing speeds apriori, in this paper, we propose ESDP (Efficient Sampling-based Dynamic Programming), which learns the distribution of the fluctuated processing speeds over time and simultaneously seeks to maximize the cumulative overall utility. The cumulative overall utility is formulated as the sum of the utilities of successfully serving each multi-server job minus the penalty on the operating, maintaining, and energy cost. ESDP is proved to have a polynomial complexity and a logarithmic regret, which is a State-of-the-Art result. We also validate it with extensive simulations and the results show that the proposed algorithm outperforms several benchmark policies with improvements by up to 73%, 36%, and 28%, respectively.

cs.DC↗

arXiv2022

Theoretically Guaranteed Online Workload Dispatching for Deadline-Aware Multi-Server Jobs

Multi-server jobs are imperative in modern computing clusters. A multi-server job has multiple task components and each of the task components is responsible for processing a specific size of workloads. Efficient online workload dispatching is crucial but challenging to co-located heterogeneous multi-server jobs. The dispatching policy should decide $(i)$ where to launch each task component instance of the arrived jobs and $(ii)$ the size of workloads that each task component processes. Existing policies are explicit and effective when facing service locality and resource contention in both offline and online settings. However, when adding the deadline-aware constraint, the theoretical superiority of these policies could not be guaranteed. To fill the theoretical gap, in this paper, we design an $α$-competitive online workload dispatching policy for deadline-aware multi-server jobs based on the spatio-temporal resource mesh model. We formulate the problem as a social welfare maximization program and solve it online with several well designed pseudo functions. The social welfare is formulated as the sum of the utilities of jobs and the utility of the computing cluster. The proposed policy is rigorously proved to be $α$-competitive for some $α\geq 2$. We also validate the theoretical superiority of it with simulations and the results show that it distinctly outperforms two handcrafted baseline policies on the social welfare.

cs.DC↗

arXiv2021

DPoS: Decentralized, Privacy-Preserving, and Low-Complexity Online Slicing for Multi-Tenant Networks

Network slicing is the key to enable virtualized resource sharing among vertical industries in the era of 5G communication. Efficient resource allocation is of vital importance to realize network slicing in real-world business scenarios. To deal with the high algorithm complexity, privacy leakage, and unrealistic offline setting of current network slicing algorithms, in this paper we propose a fully decentralized and low-complexity online algorithm, DPoS, for multi-resource slicing. We first formulate the problem as a global social welfare maximization problem. Next, we design the online algorithm DPoS based on the primal-dual approach and posted price mechanism. In DPoS, each tenant is incentivized to make its own decision based on its true preferences without disclosing any private information to the mobile virtual network operator and other tenants. We provide a rigorous theoretical analysis to show that DPoS has the optimal competitive ratio when the cost function of each resource is linear. Extensive simulation experiments are conducted to evaluate the performance of DPoS. The results show that DPoS can not only achieve close-to-offline-optimal performance, but also have low algorithmic overheads.

cs.NI↗

arXiv2020

Placement is not Enough: Embedding with Proactive Stream Mapping on the Heterogenous Edge

Edge computing is naturally suited to the applications generated by Internet of Things (IoT) nodes. The IoT applications generally take the form of directed acyclic graphs (DAGs), where vertices represent interdependent functions and edges represent data streams. The status quo of minimizing the makespan of the DAG motivates the study on optimal function placement. However, current approaches lose sight of proactively mapping the data streams to the physical links between the heterogenous edge servers, which could affect the makespan of DAGs significantly. To solve this problem, we study both function placement and stream mapping with data splitting simultaneously, and propose the algorithm DPE (Dynamic Programming-based Embedding). DPE is theoretically verified to achieve the global optimality of the embedding problem. The complexity analysis is also provided. Extensive experiments on Alibaba cluster trace dataset show that DPE significantly outperforms two state-of-the-art joint function placement and task scheduling algorithms in makespan by 43.19% and 40.71%, respectively.

cs.NI↗

arXiv2020

Distributed Redundant Placement for Microservice-based Applications at the Edge

Multi-access Edge Computing (MEC) is booming as a promising paradigm to push the computation and communication resources from cloud to the network edge to provide services and to perform computations. With container technologies, mobile devices with small memory footprint can run composite microservice-based applications without time-consuming backbone. Service placement at the edge is of importance to put MEC from theory into practice. However, current state-of-the-art research does not sufficiently take the composite property of services into consideration. Besides, although Kubernetes has certain abilities to heal container failures, high availability cannot be ensured due to heterogeneity and variability of edge sites. To deal with these problems, we propose a distributed redundant placement framework SAA-RP and a GA-based Server Selection (GASS) algorithm for microservice-based applications with sequential combinatorial structure. We formulate a stochastic optimization problem with the uncertainty of microservice request considered, and then decide for each microservice, how it should be deployed and with how many instances as well as on which edge sites to place them. Benchmark policies are implemented in two scenarios, where redundancy is allowed and not, respectively. Numerical results based on a real-world dataset verify that GASS significantly outperforms all the benchmark policies.

cs.DC↗