I have designed and developed a protocol for a progressive file transfer. Now I want to calculate some metrics for the same. The most important one is speed. Currently, I ran both the client and the server on a single machine, hence the speed was say,x. But if the protocol is to be run in a distributed environment with routers, there will be some delays introduced, like processing delay, queue delay, propagation delay, etc. Could some one tell me what are the points I should consider to calculate a rough speed of the protocol. Should it be a function of these delays?
2
-
Given that the variations between distributed endpoints are effectively infinite, then it doesn't mean much to describe a speed of a protocol in any specific environment, as it will always be different to real world. It is probably best to have a metric that is a speed comparison to a similar protocol in an ideal environment.– PaulJul 12, 2016 at 6:16
-
Thanks a lot @Paul. I also would like to know how are the links taken into account. For instance, is it generally a standard practice to say that speed of the protocol is x in a link with Rbps or something?– Ankur BhatiaJul 12, 2016 at 6:20
Add a comment
|