search

hongzimao / pensieve

Neural Adaptive Video Streaming with Pensieve (SIGCOMM '17)
http://web.mit.edu/pensieve/
MIT License
517 stars 280 forks source link

FCC Data Set #96

Open yechunli opened 4 years ago

yechunli commented 4 years ago

As I know from the paper, the RL algorithmn is used to promote the QoS of mobile phone users. I find you use 3G/HSDPA mobile dataset collected inNorway. Why do you choose the Fixed Broadband Information rather than Mobile Broadband Information in FCC .

yechunli commented 4 years ago

http://data.fcc.gov/download/measuring-broadband-america/2016/data-raw-2016-jun.tar.gz. It is the FCC data set link in README.md under traces. I can't access it.

hongzimao commented 4 years ago

RL algorithm is agnostic to the underlying network it is optimized over. It will also improve the performance if we train it with broadband network, although the improvement might be minor compared to wireless network because the existing algorithms might be close to optimal already.

It looks like FCC removed the original data from their website. We stored a portion of the data here (see data format in traces/): https://www.dropbox.com/sh/ss0zs1lc4cklu3u/AAAOEcmVUpOrOJ0mizaqzEFEa/cooked_traces?dl=0&subfolder_nav_tracking=1

yechunli commented 4 years ago

RL algorithm is agnostic to the underlying network it is optimized over. It will also improve the performance if we train it with broadband network, although the improvement might be minor compared to wireless network because the existing algorithms might be close to optimal already.

I don't mean the broadband network. I mean the RL algorithm aims to protmote QoS of MOBILE PHONE USERS. And you use the FIXED Broadband Information in this project. Why do you choose the Fixed Broadband Information. I find there are MOBILE Broadband Information in FCC and why don't you use it?

hongzimao commented 4 years ago

Chunli, it's good that you are careful looking at the dataset and found mobile network traces. We did make use of all data and evaluated the model over different networks in our experiment. For example, Figure 7b of the paper is evaluated on the mobile network traces.

joybi531 commented 3 years ago

@yechunli Hi, I was trying to do this experiment but unfortunately, I am not being able to get access to any of the raw data (zip file) and cooked_trace data. Do you have any of these two or both of them? I would really appreciate any support or information on where I can find the dataset.

Note: I have checked on the fcc.gov website but couldn't download the data due to some error.

hongzimao commented 3 years ago

I uploaded the trace files in my local machine in #134. Hope this can help!

yourexpress commented 2 years ago

Here is the latest FCC broadband-measuring raw data:https://www.fcc.gov/reports-research/reports/measuring-broadband-america/raw-data-measuring-broadband-america-2016

But I'm not sure the data of "data-raw-2016-jun.tar.gz" on this webpage is exactly the same as the data used by Hongzi. Hope this could be useful.