I am making a paper about the different congestion avoidance protocols of TCP and one of them is the TCP Vegas. I have searched a lot on internet about how it's work but I have some questions. If I am not wrong, the algorithm should be like follows:
When ???? change to congestion avoidance phase
if there are no duplicated ACKS
if it is the first or second one after retransmission then check the timeout and retransmit
if Diff < alpha then congestion window size++
if Diff > beta then congestion window size--
if duplicated ACK is received
if time-out for this segment
// any reduction of the window???
if ACK time-out
congestion window size = 2
I have 3 questions:
- The change of the modified slow-start is that the window is increased every other RTT (like I found in a lot of papers), but what is the meaning of other RTT?
- When the algorithm change to the lineal growth phase (or congestion avoidance phase)?
- And, there is any reduction of the window when a duplicated ACK is received?