Home

Jump to Current week!

Week 1: Project overview and course expectations

Sep 27 - Oct 1

Discussion topics:

• Introductions
• TCP/IP Networking Overview
• Tools introduction: Wireshark, Docker, and DANE
  

Assignment:

• Install Wireshark, capture and investigate packet data
• Create a brief (no more than 5 slides) presentation highlighting some of your key Wireshark analysis & takeaways. We recommend using screenshot snippets and annotating them with your notes. Be prepared to share your findings in Section next week.

Week 2:

Oct 4 - Oct 8

Discussion topics:

• Discuss Wireshark analysis
• Going deeper with Wireshark packet capture data

Assignment:

• Generate packet capture data in Wireshark and export to CSV
• Analyze and visualize CSV data
• Create a brief (no more than 5 slides) presentation showcasing your DANE-data visualizations and key takeaways. Be prepared to share your findings in Section next week.

NOTE: See Homework 2 Detailed Instructions

Week 3:

Oct 11 - Oct 15

Discussion topics:

• Introduction to DANE and Docker

Assignment:

• Install Docker and DANE
• Capture network traffic data with DANE and visualize it in your preferred environment (Jupyter, R Studio, etc)
• Create a brief (no more than 5 slides) presentation showcasing your visualizations and key takeaways. Be prepared to share your findings in Section next week.

NOTE: See Homework 3 Detailed Instructions

Week 4:

Oct 18 - 22

Discussion topics:

• Network degradation scenarios - what are they? Why do we care?
• How to configure DANE to emulate degradation scenarios.
• Looking ahead - what’s to expect for the next few weeks and rest of Fall Quarter.

Assignment:

• Capture network traffic data with DANE for various packet loss scenarios.
• Analyze and visualize your data in your preferred environment.
• Identify candidate features for a model to predict packet loss.
• Create a brief presentation showcasing your visualizations and key takeaways. Be prepared to share your findings in Section next week.

Week 5:

Oct 25 - Oct 29

Discussion topics:

• DANE, CPU utilization, and other concurrent internet usage
• Work in pairs? Checkpoint 1 status?
• Building a model to predict packet loss ratio

Assignment:

• Build and train a model to predict packet loss ratio.
• Use DANE to generate as much data as you need to sufficiently train and test your model.
  • Remember it may be better to run DANE scenarios one or two at a time, vs. try to run too many concurrently.
• Share your progress on your model. Summarize results in a brief presentation (as typical) but also share your notebooks, scripts, etc.

Week 6:

Nov 1 - Nov 5

Discussion topics:

• Checkpoint 1 review
• Packet Loss model review
• Adding variability in latency

Assignment:

• Finish up your packet loss models
• Start exploring how modifying latency affects your features & models
• Propose (and define!) candidate features for predicting latency.
  • You may want to start by keeping packet loss fixed and generating data sets for various latencies.
  • FYI cable/fiber internet services typically have latencies of 20ms or less (what’s yours?). Geostationary satellites like ours typically have a round trip (to the satellite and back) latency of 260ms or so.

Week 7:

Nov 8 - Nov 12

Discussion topics:

• Review Packet Loss models
• Latency prediction models - review candidate features
• Summarize timeline for the rest of the quarter
  • Begin discussion about Q2 project proposals and teams
  • Final Replication Report and Final Presentations

Assignment:

• Build & train a model to detect latency, given a fixed packet loss ratio (you choose the ratio for your initial model)

Week 8: Q2 Project Proposals

Nov 15 - 19

Discussion topics:

• Review latency prediction model results
• Discuss candidate Q2 project ideas

Assignment:

• Extend your models to predict both latency AND packet loss
• Use DANE to generate as much data as you need to build, train, and test your models
• Summarize model results, including a list (and definitions!) of your features.
• Include links to your models (notebooks, scripts, etc)
  
  

Week 9

Thanksgiving Week

Nov 22 - Nov 26

Discussion topics:

We will be holding discussion this week! Make sure you attend.

• Finishing up your latency & packet loss classification models - your results will be documented in your Final Replication Paper.
• Continue discussion about Q2 - project proposals and teams

Assignment:

• Finish up your latency & packet loss models
• Finish up your Q2 project proposals

Week 10

Nov 29 - Dec 3

1. Teams Present to Viasat
   • Each team delivers their elevator pitch and at least the first couple weeks of their detailed task schedule.
2. Viasat shares our expectations for next quarter.