Deep Q Networks with Centralized Learning over LEO Satellite Networks in a 6G Cloud Environment

Rodrigues, Tiago Koketsu; Kato, Nei

doi:10.1109/globecom48099.2022.10000709

Cited by 4 publications

(3 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, because it does not need to send unprocessed information from the system edge to the system heart, distributed learning is characterized by its scalability in terms of communication [ 49 ]. This tradeoff between the two strategies has been the focus of existing research, with most studies primarily emphasizing either centralized or distributed learning [ 50 , 51 ]. Rarely do these studies acknowledge the drawbacks of both approaches and attempt to address them using techniques from the alternative paradigm [ 52 ].…”

Section: Related Workmentioning

confidence: 99%

Enhancing Earth data analysis in 5G satellite networks: A novel lightweight approach integrating improved deep learning

Yang,

Ren,

Song

2024

Heliyon

View full text Add to dashboard Cite

Section: Related Workmentioning

confidence: 99%

Enhancing Earth data analysis in 5G satellite networks: A novel lightweight approach integrating improved deep learning

Yang,

Ren,

Song

2024

Heliyon

View full text Add to dashboard Cite

“…The parameters in the experiment are set in Table 1. Such values were obtained after careful tests to find the best hyperparameters for our desired application [40]. Other hyperparameters are the same as the default values in MMDetection.…”

Section: Parameters and Metricsmentioning

confidence: 99%

An Effective Multi-Layer Attention Network for SAR Ship Detection

Suo

Zhao

2023

JMSE

View full text Add to dashboard Cite

The use of deep learning-based techniques has improved the performance of synthetic aperture radar (SAR) image-based applications, such as ship detection. However, all existing methods have limited object detection performance under the conditions of varying ship sizes and complex background noise, to the best of our knowledge. In this paper, to solve both the multi-scale problem and the noisy background issues, we propose a multi-layer attention approach based on the thorough analysis of both location and semantic information. The solution works by exploring the richness of spatial information of the low-level feature maps generated by a backbone and the richness of semantic information of the high-level feature maps created by the same method. Additionally, we integrate an attention mechanism into the network to exclusively extract useful features from the input maps. Tests involving multiple SAR datasets show that our proposed solution enables significant improvements to the accuracy of ship detection regardless of vessel size and background complexity. Particularly for the widely-adopted High-Resolution SAR Images Dataset (HRSID), the new method provides a 1.3% improvement in the average precision for detection. The proposed new method can be potentially used in other feature-extraction-based classification, detection, and segmentation.

show abstract

“…Because atomization in LJIC is a dynamic process, frequent adjustment of the mesh size to adapt to the environment is preferred. 46 The Level 3 Adaptive Mesh Refinement (AMR) method based on liquid volume fraction gradient is used in our simulation. The refinement threshold of AMR is set to 0.75, while the coarsening threshold is put at 0.25.…”

Section: E Grid Independence Verificationmentioning

confidence: 99%

Atomization of liquid pulsed jet in subsonic crossflow

Chang

Chen

et al. 2023

AIP Advances

View full text Add to dashboard Cite

Pulsed jet is an effective solution to improve fuel jet penetration depth and consequently increase the mixing efficiency of gas–liquid in conventional combustion chambers. This has the benefits of reducing pollutant emissions and diminishing the instability of fuel combustion. However, the atomization process of pulsed jets with small amplitude has still not been properly investigated. This paper studies such a process through Large Eddy Simulation and a Coupled Level Set and Volume of Fluid method. We investigate the atomization process in a liquid pulsed jet with a subsonic crossflow and the impact of the Strouhal number on atomization morphology and the behavior of the pulsed jet in general. Results show that, with a constant mass flow rate, the role of Rayleigh–Taylor instability is replaced by the periodic fluctuation of the jet velocity, which ends up dominating the primary process of atomization of the liquid transverse pulsed jet. This also improves atomization, in general, and the fragmentation of the jet. We also show that the Strouhal number significantly impacts the penetration depth of the jet, with high values increasing penetration by up to 12%.

show abstract

Deep Q Networks with Centralized Learning over LEO Satellite Networks in a 6G Cloud Environment

Cited by 4 publications

References 25 publications

Enhancing Earth data analysis in 5G satellite networks: A novel lightweight approach integrating improved deep learning

Enhancing Earth data analysis in 5G satellite networks: A novel lightweight approach integrating improved deep learning

An Effective Multi-Layer Attention Network for SAR Ship Detection

Atomization of liquid pulsed jet in subsonic crossflow

Contact Info

Product

Resources

About