.png){kind=icon}
.png){kind=icon}
Return
Device to Device Collaboration for Mobile Clouds in Drop Computing
.png){kind=icon}
PRE
AI
DOI：10.1109/percomw.2019.8730788.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The large number of mobile devices existing nowadays has led to the evolution of mobile cloud computing towards bringing data and computations closer to the nodes. This has manifested first in the shape of fog and edge computing, where an additional communication and processing layer is added at the edge of the network. However, the fast adoption of the Internet of Things has shown the limitations of even this model, so the focus now is moving towards another layer that is one level below: the ad hoc network composed of the mobile devices themselves. One paradigm based on this model is Drop Computing, where nodes that need to do some computations first attempt to process them through the help of neighbor devices using close-range communication (such as Wi-Fi Direct or Bluetooth), and only then do they attempt to contact the fog/edge nodes or the cloud itself. In this paper, we propose an Android implementation of the device-to-device layer of Drop Computing. On top of this implementation, we present an application that creates a video collage from multiple photos using ffmpeg with the help of neighboring nodes through close-range communication using the HYCCUPS and Google Nearby frameworks. Through experiments on four Android devices, we show that our implementation can drastically decrease CPU usage per device, which in turn increases the overall quality of experience for Android users. Furthermore, the total battery consumption is lowered, since nodes have less computations to perform and the CPU cores spend less time in higher frequencies.
Keywords:
mobile
cloud
edge
opportunistic
Android
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
I
IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops)
IF:
0
Papers:
10
Citations:
0