Federated learning (FL) is an emerging promising paradigm of privacy-preserving machine learning (ML). An important type of FL is cross-silo FL, which enables a moderate number of organizations to cooperatively train a shared model by keeping confidential data locally and aggregating gradients on a central parameter server. However, the central server may be vulnerable to malicious attacks or software failures in practice. To address this issue, in this paper, we propose$\mathtt{DegaFL} $, a novel decentralized gradient aggregation approach for cross-silo FL.$\mathtt{DegaFL} $eliminates the central server by aggregating gradients on each participant, and maintains and synchronizes gradients of only the current training round. Besides, we propose$\mathtt{AdaAgg} $to adaptively aggregate correct gradients from honest nodes and use HotStuff to ensure the consistency of the training round number and gradients among all nodes. Experimental results show that$\mathtt{DegaFL} $defends against common threat models with minimal accuracy loss, and achieves up to$50\times$reduction in storage overhead and up to$13\times$reduction in network overhead, compared to state-of-the-art decentralized FL approaches.

Federated learning (FL) is a privacy-preserving deep learning paradigm. An important type of FL is cross-silo FL, which enables a moderate number of organizations to cooperatively train a shared model while keeping private data locally and aggregating parameters on a central parameter server. However, the central server may be vulnerable to malicious attacks or software failures. To address this problem, Swarm Learning (SL) has emerged to perform FL in a decentralized manner by introducing a blockchain to securely onboard members and dynamically elect the leader for parameter aggregation. Despite tremendous attention to SL recently, few measurement studies provide comprehensive knowledge of best practices and precautions for deploying SL in real-world scenarios. To this end, we conduct the first empirical study of SL, to fill the knowledge gap between SL research and real-world deployment. We conduct various experiments on 3 public datasets for 4 research questions, present interesting findings, quantitatively analyze the reasons behind these findings, and provide developers and researchers with practical suggestions.

Data infrastructure plays a vital role as the cornerstone supporting the operations of modern society. Nevertheless, data characterization methods within this infrastructure encounter challenges related to inefficiency, high costs, and the management of fuzzy and complex retrieval requirements. To tackle these obstacles, this paper introduces a highly efficient approach for characterizing digital objects called DOR-RAF. This methodology integrates Large Language Models (LLMs) and Retrieval-Augmented Generation (RAG) to enhance the intelligent retrieval experience for users. It presents an interactive retrieval mechanism that accurately identifies users’ true needs through iterative dialogues, effectively handling user queries despite their initial vagueness. The experimental results show that DOR-RAF outperforms the traditional metadata-based retrieval methods in terms of F1 Score and other metrics on the task of retrieving digital objects in a data infrastructure scenario. Meanwhile, the experiments show that the development of data characterisation methods using RAG technology has a broad prospect, and confirm that DOR-RAF outperforms traditional RAG methods in terms of Context Precision, Answer Correctness, etc. in the task of data characterisation.

Mobile extended reality (XR) has developed rapidly in recent years. Compared with the app-based XR, XR in web browsers has the advantages of being lightweight and cross-platform, providing users with a pervasive experience. Therefore, many frameworks are emerging to support the development of XR in web browsers. However, little has been known about how well these frameworks perform and how complex XR apps modern web browsers can support on mobile devices. To fill the knowledge gap, in this paper, we conduct an empirical study of mobile XR in web browsers. We select seven most popular web-based XR frameworks and investigate their runtime performance, including 3D rendering, camera capturing, and real-world understanding. We find that current frameworks have the potential to further enhance their performance by increasing GPU utilization or improving computing parallelism. Besides, for 3D scenes with good rendering performance, developers can feel free to add camera capturing with little influence on performance to support augmented reality (AR) and mixed reality (MR) applications. Based on our findings, we draw several practical implications to provide better XR support in web browsers.