1 May 2021 in IEEE Transactions on Computers
Federated learning (FL) has attracted increasing attention as a promising approach to driving a vast number of end devices with artificial intelligence. However, it is very challenging to guarantee the efficiency of FL considering the unreliable nature of end devices while the cost of device-server communication cannot be neglected. In this paper, we propose SAFA, a semi-asynchronous FL protocol, to address the problems in federated learning such as low round efficiency and poor convergence rate in extreme conditions (e.g., clients dropping offline frequently). We introduce novel designs in the steps of model distribution, client selection and global aggregation to mitigate the impacts of stragglers, crashes and model staleness in order to boost efficiency and improve the quality of the global model. We have conducted extensive experiments with typical machine learning tasks. The results demonstrate that the proposed protocol is effective in terms of shortening federated round duration, reducing local resource wastage, and improving the accuracy of the global model at an acceptable communication cost.
26 November 2020 in IEEE Transactions on Parallel and Distributed Systems
Mobile Edge Computing (MEC), which incorporates the Cloud, edge nodes, and end devices, has shown great potential in bringing data processing closer to the data sources. Meanwhile, Federated learning (FL) has emerged as a promising privacy-preserving approach to facilitating AI applications. However, it remains a big challenge to optimize the efficiency and effectiveness of FL when it is integrated with the MEC architecture. Moreover, the unreliable nature (e.g., stragglers and intermittent drop-out) of end devices significantly slows down the FL process and affects the global model's quality in such circumstances. In this article, a multi-layer federated learning protocol called HybridFL is designed for the MEC architecture. HybridFL adopts two levels (the edge level and the cloud level) of model aggregation enacting different aggregation strategies. Moreover, in order to mitigate stragglers and end device drop-out, we introduce regional slack factors into the stage of client selection performed at the edge nodes using a probabilistic approach without identifying or probing the state of end devices (whose reliability is agnostic). We demonstrate the effectiveness of our method in modulating the proportion of clients selected and present the convergence analysis for our protocol. We have conducted extensive experiments with machine learning tasks in different scales of MEC system. The results show that HybridFL improves the FL training process significantly in terms of shortening the federated round length, speeding up the global model's convergence (by up to 12×) and reducing end device energy consumption (by up to 58 percent).
30 May 2017 in The VLDB Journal
28 October 2004 in The Astrophysical Journal
We present the first interferometric CO(J=3-2) observations (beam size of 3.9"x1.6" or 160pc x 65pc) with the Submillimeter Array (SMA) toward the center of the Seyfert 2 galaxy M51. The image shows a strong concentration at the nucleus and weak emission from the spiral arm to the northwest. The integrated intensity of the central component in CO(J=3-2) is almost twice as high as that in CO(J=1-0), indicating that the molecular gas within an ~80 pc radius of the nucleus is warm (>~100 K) and dense (~10^4 cm^-3). Similar intensity ratios are seen in shocked regions in our Galaxy, suggesting that these gas properties may be related to AGN or starburst activity. The central component shows a linear velocity gradient (~1.4 km/s/pc) perpendicular to the radio continuum jet, similar to that seen in previous observations and interpreted as a circumnuclear molecular disk/torus around the Seyfert 2 nucleus. In addition, we identify a linear velocity gradient (~0.7 km/s/pc) along the jet. Judging from the energetics, the velocity gradient can be explained by supernova explosions or energy and momentum transfer from the jet to the molecular gas via interaction, which is consistent with the high intensity ratio.Comment: 12 pages, 3 figures, accepted for publication in ApJ
1 February 2019 in IEEE Transactions on Parallel and Distributed Systems
1 May 2015 in Computers in Industry