The following ExFiles run the experiments described in: G. Bartlett and J. Mirkovic, "Expressing Different Traffic Models Using The LegoTG Framework", Proceedings of the Workshop on Computer and Networking Experimental Research using Testbeds (CNERT), 2015.
Each ExFile has a configuration to run the different background traffic models used in the paper along with TGBlocks for two bandwidth estimation tools: pathload and pathchirp. These ExFiles are configured to run on 10 nodes in a dumbell configuration (5 nodes on each side). This number can be increased or decreased via modification to the "nodes" attribute in the ExFile.
In this paper, we compared two bandwidth estimation tools, pathchirp and pathload and demonstrated that with different models of background traffic, these two tools performed differently and acheived differing levels of accuracy. Ultimately the point is background traffic matters when testing a tool or system. What traffic model is used can drastically ulter results and conclusions.
In these ExFiles, machine names and specific environment configurations (such as your user name and log in credentials to reach your nodes) should all be changed to match your testbed environment and topology before running these experiment via LegoTG's orchestrator. The models used in the three experiments are