[Starlink] It’s the Latency, FCC

[David Lang]

On Tue, 30 Apr 2024, Alexandre Petrescu via Starlink wrote:

> I agree with you: two distinct parameters, bandwidth and latency.  But 
> they evolve simultenously, relatively bound by a constant relationship.  
> For any particular link  technology (satcom is one) the bandwidth and 
> latency are in a constant relationship.  One grows, the other 
> diminishes.  There are exceptions too, in some details.

No, As a general rule you do not have to accept poor latency to get good speed.

at least not until the link is very close to full utilization. But below some 
figure (depending on how you are managing latency, it may be 80% utilization, 
90% utilization, or even higher), there is no bandwidth gain from lowering 
latency.

Even for media that batches traffic (wifi), the bandwidth tradeoff between 'send 
everything you have now, let anything that misses this window go out in the next 
transmission' vs 'delay sending anything just in case there is more that will 
arrive and can be sent in the same transmission slow instead of it having to 
wait' only affects the bandwidth used in the first transmission op, after that 
the bandwidth usage will be the same, but you have better latency. Now when you 
try to add fairness for something like this you may end up costing latency 
and/or bandwidth, but that's a different discussion than there being a 
latency/bandwith constant relationship like you claim.

So I very much disagree that you have to trade off latency for bandwidth.

There may be some special data link technologies were there is a 
bandwidth/latency tradeoff, but I am not aware of them (please educate me if 
there are ones that I am not aware of)

David Lang