Hello Zhe,
Thanks for your email!
The feature I mentioned in the talk has been implemented in this merged pull request:
https://github.com/StanfordSNR/pantheon/pull/39, except the analysis scripts might have to be adapted to analyze the output results. However, this feature is meant to evaluate multiple TCP flows sharing the same bottleneck link, where each flow can use different congestion control (CC) schemes.
It sounds to me that your question is not about different CC algorithms though. If you simply would like to run multiple TCP flows, each executing the same CC algorithm, then "experiments/test.py -f" is what you are looking for. If these two multiflow scenarios are not what you are asking, I am afraid you would have to write a script by yourself, e.g., leveraging our existing scripts and tools like Pantheon-tunnel.