layout: true background-image: url(SlideTemplate/img/Background.png) --- class: center, middle ### A Brief Task on Networks and # Communications:
From Fire Signals to NetFlix ### Forschungstage Informatik - 2018 ##### *Prof. Dr.-Ing. Thorsten Herfet, .hon-speaker[Andreas Schmidt], Pablo Gil Pereira, Kelvin Chelli* ##### Telecommunications Lab, Saarland Informatics Campus, 14th and 15th June 2018 ??? Welcome students. TH --- # DASH Emulator **Practical** analysis of video streaming **concepts**. Implementation in **Python 3**. Emulated environment: 3 minutes video session in milliseconds. **Metrics**: * Downloaded video quality. * Buffer level. * Total duration of re-bufferings. * Wasted download oportunities. --- # Adaptive Bit Rate (ABR) Algorithms Implemented in the video client to select the segment quality. They consider metrics such as **average throughput** and **buffer level**. **SimpleABR**: next quality is the immediately inferior to the average throughput of the last download. **TC_TeamABR**: Can you beat us? .center.dash-abr[ .small[[Source](https://www.researchgate.net/profile/Tobias_Hossfeld/publication/271524253_A_Survey_on_Quality_of_Experience_of_HTTP_Adaptive_Streaming/links/55f4224e08ae63926cf2688f/A-Survey-on-Quality-of-Experience-of-HTTP-Adaptive-Streaming.pdf)] ] --- # Jupyter .center.jupyter[  ] --- # How does the emulator look like? .center.jupyter[  ] --- # Tasks * Play with player configuration parameters to get better performance. * Buffer size, reference buffer, N... * RANDOM_SEED controls the network scenario. * Build up your own ABR based on `SimpleABR`. * Modify `SimpleABR` inside the notebook and execute the cells. * Beat us and get a bounty... --  .small[[Source](https://upload.wikimedia.org/wikipedia/en/1/19/Bounty-Wrapper-Small.jpg)]