AStream: A rate adaptation model for DASH

AStream is a Python based emulated video player to evalutae the perfomance of the DASH bitrate adaptaion algorithms.

Rate Adaptatation Algorithms Supported

1. Basic Adaptation
2. Segment Aware Rate Adaptation (SARA): Our algorithm
3. Buffer-Based Rate Adaptation (Netflix): This is based on the algorithm presented in the paper. Te-Yuan Huang, Ramesh Johari, Nick McKeown, Matthew Trunnell, and Mark Watson. 2014. A buffer-based approach to rate adaptation: evidence from a large video streaming service. In Proceedings of the 2014 ACM conference on SIGCOMM (SIGCOMM '14). ACM, New York, NY, USA, 187-198. DOI=10.1145/2619239.2626296 http://doi.acm.org/10.1145/2619239.2626296

Logs

Buffer Logs:

1. Epoch time
2. Current Playback Time
3. Current Buffer Size (in segments)
4. Current Playback State

Playback Logs:

1. Epoch time
2. Playback time
3. Segment Number
4. Segment Size
5. Playback Bitrate
6. Segment Size
7. Segment Duration
8. Weighted harmonic mean average download rate

Requirements

The AStream DASH video client is compatible with both Python 2 and Python 3. For Python 2, you must also install the six library.

Sample Run

python dist/client/dash_client.py -m <URL TO MPD> -p <PlaybackType> 

Command Line options

dash_client.py [-h] [-m MPD] [-l] [-p PLAYBACK] [-n SEGMENT_LIMIT] [-d]

Process Client parameters

optional arguments:
  -h, --help            show this help message and exit
  -m MPD, --MPD MPD     Url to the MPD File
  -l, --LIST            List all the representations and quit
  -p PLAYBACK, --PLAYBACK PLAYBACK
                        Playback type ('basic', 'sara', 'netflix', or 'all')
  -n SEGMENT_LIMIT, --SEGMENT_LIMIT SEGMENT_LIMIT
                        The Segment number limit
  -d, --DOWNLOAD        Keep the video files after playback