With the rise of cloud-edge collaboration, recommendation services are increasingly trained in distributed environments. Federated Recommendation (FR) enables such multi-end collaborative training while preserving privacy by sharing model parameters instead of raw data. However, the large number of parameters, primarily due to the massive item embeddings, significantly hampers communication efficiency. While existing studies mainly focus on improving the efficiency of FR models, they largely overlook the issue of embedding parameter overhead. To address this gap, we propose a FR training framework with Parameter-Efficient Fine-Tuning (PEFT) based embedding designed to reduce the volume of embedding parameters that need to be transmitted. Our approach offers a lightweight, plugin-style solution that can be seamlessly integrated into existing FR methods. In addition to incorporating common PEFT techniques such as LoRA and Hash-based encoding, we explore the use of Residual Quantized Variational Autoencoders (RQ-VAE) as a novel PEFT strategy within our framework. Extensive experiments across various FR model backbones and datasets demonstrate that our framework significantly reduces communication overhead while improving accuracy.

The metaverse has evolved from a theoretical concept into a digital infrastructure with multiple simultaneous users, real-time interaction, and high-quality immersive experiences. Traditional architecture inherited from multiplayer gaming cannot adequately meet these requirements due to rendering pressure, battery life, device weight, and service cost. While Cloud and Edge XR architectures address some constraints by offloading computation, they introduce latency issues that com promise immersion. Edge-end collaborative rendering XR offers a promising direction but struggles with task decomposition, edge resource utilization, and practical infrastructure. This paper presents Crossport, a cloud-edge-end architecture that enables microservice-based collaborative rendering for metaverseser vices. Crossport introduces a scene-based microservice decomposition methodology, a composition toolchain supporting multiple consumers, and containerization middleware for integration with existing infrastructure. Experimental evaluation shows Crossport reduces edge resource consumption by up to 29.5% in single-user scenarios and 55.3% in multi-user contexts while maintaining high-quality immersive experiences.

While microservices architectures have significantly enhanced software modularity and scalability, they introduce unique security challenges that traditional monolithic approaches do not face. The proliferation of microservices often leads to homogeneity within systems, where multiple instances share iden-tical container images and software versions, creating a systemic security vulnerability that attackers can exploit. In homogeneous microservice environments, a single vulnerability can be exploited to compromise multiple services simultaneously, facilitating rapid lateral movement attacks and significantly reducing system re-silience against cyber threats. While existing security research has focused primarily on preventive measures like authentication protocols, the potential of strategic deployment patterns as a reactive defense mechanism remains largely unexplored. In this paper, we base our analysis on information theory to demonstrate that microservice deployment schemes affect system heterogeneity, which in turn influences overall security. Additionally, we proposed the Heterogeneous Microservices Placement Problem HMPP and developed a heuristic-based deployment algorithm, H- HMPP, to effectively address it. Extensive experimental evaluations demonstrate that our approach yields high-quality solutions while significantly reducing execution times.

Personal Knowledge Graphs (PKGs) organize an individual user's information into a structured format comprising entities, attributes, and relationships. By leveraging this structured and semantically rich data, PKGs have become essential for securing personal data management and delivering personalized services. To unlock their potential in personalized recommendations, prior research has explored the construction of PKGs and recommendation methods built upon them. However, these studies often overlook challenges associated with distributed PKGs across different users, such as joint training and privacy protection. To address these challenges, we propose PKGRec, a federated graph recommendation method specifically designed for PKGs, which utilizes a federated learning framework to ensure user privacy and data security during joint learning. Furthermore, to accommodate the user-centric graph structure of PKGs, our approach categorizes entities into three types: users, items, and other entities. It then applies a novel staged graph convolution method to model various entities based on these entity categories during local training. To enable efficient graph information sharing among distributed PKGs without requiring additional data transfer or aggregation, PKGRec performs graph expansion on the trained gradients by federated aggregation. Extensive experiments conducted on four publicly available datasets demonstrate that our method consistently outperforms the existing federated recommendation approaches.

The rapid adoption of microservice architectures has spurred research into addressing various challenges in microservice system evolution, such as service deployment and task offloading. While numerous solutions have been proposed by researchers, validating these approaches requires extensive experimentation, including both algorithmic testing and simulations across different Microservice-Based Applications (MBAs). However, current research lacks sufficient physical experimentation, which is essential for evaluating how these solutions perform when deployed on actual computing infrastructure with running microservice instances. This gap exists mainly because conducting physical experiments is complex and costly, primarily due to two barriers: the lack of standardized problem modeling approaches and the absence of suitable experimental platforms. To address these challenges, the Universal Evolution Model (UEM) was proposed that unifies different microservice research problems within a common evolution framework. This model frames research problems as approaches to evolve microservice systems under specific constraints to achieve desired objectives. Based on UEM and the widely-adopted Monitor-Analyze-Plan-Execute over a shared Knowledge (MAPE-K) model, a general Microservice System Evolution Architecture was designed that standardizes experimental processes across diverse research scenarios. Building on these foundations, MicroForge was presented, an open-source experimental platform equipped with tools for accelerating experiments. The platform incorporates RescueService, a carefully curated MBA. Through comprehensive evaluation across various MBAs, the functional effectiveness of MicroForge was validated. Additionally, a practical user guide was provided to facilitate the platform's adoption and utilization.

Edge servers cache applications as microservices to provide low-latency services to end-device users. Due to heterogeneous resource constraints, microservices supporting the same application may be distributed across multiple edge servers. When an end-device makes a request, it generates a task execution path comprising interconnected dependent microservices across the edge network. However, task execution time can vary significantly across these servers due to resource heterogeneity, unpredictable workloads, cyberattacks, and software exceptions. The interdependent nature of microservice execution makes single-path models highly vulnerable, as any failure or delay compromises the entire task chain. This paper formulates the Dependent Task Execution Path Planning (DTEP) problem-a novel NP-hard optimization that transforms single-path dependency models into robust multi-path redundancy frameworks. With a redundancy budget, it plans multiple redundant, shortest node-disjoint execution paths to enhance reliability and latency performance. To solve this complex problem efficiently, we propose KPath, a novel two-phase approximation algorithm consisting of base path generation followed by target path acquisition, offering a guaranteed approximation ratio of$O(m)$. Experimental results demonstrate that KPath efficiently generates path sets with average path lengths up to 17.87% shorter than modified state-of-the-art methods across various problem scales. Furthermore, when node issues occur, KPath achieves up to 26% reduction in communication latency through redundancy.

Data ecosystems (DE) are the future directions of data management and play a vital role in unlocking the value of data. Service Recommender Systems (RS) are typical applications in DEs. For example, deep learning-based RS on the basis of extensive data from various fields can help organizations obtain valuable insights of data. As organizations from different fields share data for better recommendation services, the risk of privacy leakage which is harmful to DEs increases. Due to privacy concerns, Vertical Federated Learning (VFL), a privacy-preserving computing technology for joint learning and privacy recommendation models among organizations in various fields, has garnered significant attention. Existing VFL methods are training models with static data from specific fields when there are new recommendation scenarios or fields. In addition, the models are fixed after one training session. Therefore, these models can only be applied to several specific recommendation fields and they can’t utilize continuously generated data that corresponds to various fields, posing challenges for long-term and extensive cooperation. To tackle these challenges, we introduce Vertical Federated Continual Learning (VFCL), which extends Continual Learning (CL) into the VFL framework to enable VFL models to sustainably adapt to new scenarios. We discuss feasible solutions based on existing CL methods. Furthermore, we propose GCPN, a method based on a dynamic architecture in VFCL. GCPN introduces fewer parameters for each new field by utilizing group connected layers and scale layers, eliminating the need for storing or using past data. It effectively alleviates the problem of catastrophic forgetting, a major issue in CL, while preserving privacy in joint recommendations. To evaluate GCPN, we construct the VFCL scenario using Amazon's public recommendation datasets. Experiments demonstrate that our method enhances the effectiveness of most tasks in CL and VFCL scenarios.

Microservices along with cloud-edge computing technologies are widely adopted to take advantage of the abundant computing resources of the cloud and the low latency, high bandwidth capabilities of the edge. However, factors such as frequent user requirement changes have made the current deployment scheme not fully adaptable to new requirements, resulting in an increase in average response time. Therefore, the microservice system needs to adjust the microservice deployment scheme online in response to continuously changing user requirements to reduce the average response time, which is known as the cloud-edge collaborative microservice deployment problem. However, existing methods are not able to meet the efficiency requirements and do not fully consider the complex dependencies of microservices and budget constraints in the cloud. To address this problem, this paper proposes a solution to the cloud-edge collaborative microservice deployment problem by modeling it as an NP-hard integer nonlinear programming problem in a cloud-edge environment consisting of private edge clouds and public cloud. An efficient Two-stage Iterated Greedy Optimization (TIGO) algorithm is also proposed and its convergence is proven. Extensive experimental results show that this approach achieves better average response times in less time compared to existing methods.

Recently, microservices architecture has become the mainstream software development and deployment architecture for most enterprises, with the advantages of continuous delivery/deployment. However, the inability of microservices systems to meet changing user requirements and other external factors can lead to a degradation of quality of service (QoS), necessitating microservice evolution to ensure QoS stability by adjusting the deployment structure and configuration of microservices through evolutionary means such as redeployment. In order to study different evolutionary problems, simulation of microservice evolution is critical because the flexibility and diversity of the simulation environment can provide a more prosperous experimental environment for the researchers involved compared single physical experiments and numerical experiments. However, existing simulators have limitations when it comes to simulating microservice evolution. They either lack the capability to support the simulation of microservice evolution or only provide simulation for specific evolutionary means, which limits their extensibility. To address this issue, this paper designs and implements a simulation toolkit EvolutionSim for MSS evolution based on discrete event. It aimes to simulate the running and evolution process of MSSs with the support of several evolutionary means. In addition, Experiments were carried out to compare the simulated results with the actual results, and the distribution was found to be similar by ADF test, which indicates that EvolutionSim can provide valid results. Moreover, three experimental scenarios were conducted to simulate three mainstream evolutionary means, and the results indicate that EvolutionSim can accurately simulate the effects of different evolutionary means on MSS.

Cloud and edge computing have been widely adopted in many application scenarios. With the increasing demand of fast iteration and complexity of business logic, it is challenging to achieve rapid development and continuous delivery in such highly distributed cloud and edge computing environment. At present, the microservice-based architecture has been the dominant deployment style, and a microservice system has to evolve agilely to offer stable Quality of Service (QoS) in the situation where user requirement changes frequently. A lot of research have been conducted to optimally re-deploy microservices to adapt to changing requirements. Nevertheless, complex dependencies between microservices and the existence of multiple instances of one single microservice in a microservice system together have not been fully considered in existing work. This article defines SPPMS, the Service Placement Problem in Microservice Systems that featurecomplex dependenciesandmultiple instances, as a Fractional Polynomial Problem (FPP). Considering the high computation complexity of FPP, it is then transformed into a Quadratic Sum-of-Ratios Fractional Problem (QSRFP) which is further solved by the our proposed greedy-based algorithms. Experiments demonstrate that our models and algorithms outperform existing approaches in both qualities of the generated solutions and computation speed.

Although many principles have been put forward to guide microservice design, such as Domain-Driven Design, Architectural Bad Smells (ABS) would be inevitably imported during microservice system design and development. There has been a consensus that the existence of ABSs would bring negative effects to microservice systems. Although some approaches use static analysis to detect ABSs in the design phase, we conjecture that the design-phase ABSs would result in some performance deficiencies at runtime. This paper conducts an empirical study on underlying correlations between runtime performance deficiencies and ABSs in microservice system design. An automated experimental stress testing framework called MRSTF is developed for automatic deployment of MSS and collecting/analyzing runtime performance data. A microservice system TrainTicket is used in this empirical study. We manually inject several typical ABSs into the system and use MRSTF to compare the runtime performances before and after the injection and check if the existence of ABSs has significant effects on the runtime performance. Experiment results show that ABSs have significant negative effects on some performance metrics of runtime MSS while doing not on others. This study provides solid evidence on the feasibility of optimizing MSS design by eliminating ABSs based on runtime performance data.