RFC 2488, "Enhancing TCP Over Satellite Channels using Standard Mechanisms", January 1999Source of RFC: tcpsat (tsv)
See Also: RFC 2488w/ inline errata
Errata ID: 5250
Publication Format(s) : TEXT
Reported By: Lloyd Wood
Date Reported: 2018-02-01
Verifier Name: Magnus Westerlund
Date Verified: 2019-03-27
Section 2 says:
The propagation delay to a LEO orbit ranges from several milliseconds when communicating with a satellite directly overhead, to as much as 80 ms when the satellite is on the horizon.
It should say:
The propagation delay to a LEO orbit ranges from several milliseconds when communicating with a satellite directly overhead, to as much as 20 ms when the satellite is on the horizon.
80 ms * speed of light gives 24,000 km,
which is a distance larger than MEO GPS altitude.
Given that the diameter of the Earth is just over 12,000 km,
a LEO satellite on the horizon will clearly not need that
propagation delay or light travel time to reach it.
(As it's defined as propagation delay,
we're not considering MAC retransmits.)
30ms gives up to 6000km slant path, which is barely
possible for a very high definition of LEO orbit that is
almost MEO. Tighten the value further, if you like, after
drawing a couple of circles and doing a bit of trigonometry.
This 80ms is being cited uncritically in papers
(found it in a recent PhD thesis),
so needs correction.