Super User is a question and answer site for computer enthusiasts and power users. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

How does the packet loss rate affect the TCP throughput rate: throughput rate reduce exponentially with respect to loss increase or throughput rate reduces by the same amount as loss increase?

share|improve this question

migrated from stackoverflow.com Apr 24 '13 at 18:48

This question came from our site for professional and enthusiast programmers.

I believe the answer you are looking for, is that the exponential is the worst case, and the linear the best. In general when one or more packets are lost/damaged, they are resent, leading to a drop in throughput equal to the size of the packet payload/resend time. that said, if the resend itself fails and the packet must be resent again, the trend can approach exponential.

Keep in mind, there is only a general correlation between packets and volume of data, so a relationship between packet count and data volume only works for perfectly spherical chickens in a vaccuum (eg: only under artificial circumstances like a lab).

share|improve this answer
    
Thank you Thomas! – arun Apr 25 '13 at 16:14

I simulated this once, with random packet drops. One the simulated error rate exceeded about 2% there was a definite upwards curve (latency vs error rate), and beyond about 9% throughput basically ceased altogether. I wouldn't care to speculate what law the curve was.

share|improve this answer

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .