English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Federated edge learning (FEEL) has attracted much attention as aprivacy-preserving paradigm to effectively incorporate the distributed data atthe network edge for training deep learning models. Nevertheless, the limitedcoverage of a single edge server results in an insufficient number ofparticipated client nodes, which may impair the learning performance. In thispaper, we investigate a novel framework of FEEL, namely semi-decentralizedfederated edge learning (SD-FEEL), where multiple edge servers are employed tocollectively coordinate a large number of client nodes. By exploiting thelow-latency communication among edge servers for efficient model sharing,SD-FEEL can incorporate more training data, while enjoying much lower latencycompared with conventional federated learning. We detail the training algorithmfor SD-FEEL with three main steps, including local model update, intra-cluster,and inter-cluster model aggregations. The convergence of this algorithm isproved on non-independent and identically distributed (non-IID) data, whichalso helps to reveal the effects of key parameters on the training efficiencyand provides practical design guidelines. Meanwhile, the heterogeneity of edgedevices may cause the straggler effect and deteriorate the convergence speed ofSD-FEEL. To resolve this issue, we propose an asynchronous training algorithmwith a staleness-aware aggregation scheme for SD-FEEL, of which, theconvergence performance is also analyzed. The simulation results demonstratethe effectiveness and efficiency of the proposed algorithms for SD-FEEL andcorroborate our analysis.

Semi-Decentralized Federated Edge Learning with Data and Device Heterogeneity