CS 2520nonono
WeChat:yj4399_
Sina Visitor System
High-level●3 Congestion Control Algorithms:○TCP Reno:■additive increase, multiplicative decrease function to adjust window size for every RTTuntil a packet loss is detected○TCP CUBIC:■based on a cubic function of time since the last congestion event○TCP BBR:■Model-based●Implement Reno and CUBIC●Comparison of the three
-
Experiments●Set up○Nodes placement in the network graph/Topology etc●Variables being tested for:○Throughput○Latency○Packet Loss (vary)○Congestion Window●Repeat experiments to get averages and standard deviations.
-
Report●Hypotheses○For example, we expect to see Reno do X if we change the variable Y (because of such and suchreasons), but CUBIC instead should do Z for the same experiment.●Experimental Setup:○For example, two nodes with an average latency of X. Locations if that is relevant.●Experimental Results:○Data and Graphs●Explanation of results:○Do the results make sense or not.○Might have results that do not make sense. Should attempt to explain what happened there.●Validation of Hypotheses of not:○Whether your results validate your hypotheses or not.●Also write about the implementation of your algorithms:○Brief overview, Data Structures, Packet Headers, any challenges, improvements that you could make,etc.
-
Bonus●Completely Optional.●Implement and compare a 4th Algorithm.●A list of some algorithms
-
Final Submission●Code●Data●Report
-
Resources●Samples of how such experiments can be performed:○Link 1○Link 2●BBR:○Github Repository