I'm working on a task on comparing TCP Tahoe and Reno.

Tahoe drops the window size to 0 every time it detects packet loss, that is clear. However, when I plot my window size for Reno, it does fast recovery sometimes ( as expected, so the window size drops to the current threshold) but other times it drops the size to 0 and does slow start.

Why is this? I thought Reno always did fast recovery?


TCP Reno is not great at handling multiple packet losses and in such cases will need to RTO, and I'm guessing that this is what you see.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.