cancel
Showing results for 
Search instead for 
Did you mean: 

To deepdiver

Tutor

To deepdiver

deepdiver

There is statistically a very, very little chance of everyone downloading data at the same time. Communication systems are designed to provide a certain Grade of Service at peak times based on the number of communication channels and users as predicted by Erlang statistical formulas. If some people are data hogs, then the rest of the users are degraded and aren't able to use what they paid for.

As I'm always attempting to understand the how's and why's of the satellite internet, I'd like to ask you to elaborate a bit on this recent post you made. Knowing your background in communication technology, I find it fascinating.

I understand the concept of load balancing but I don't understand why in actual usage, the system is so badly broken. Why can some of us regularly, reliably and predictably see up to and over 3000% differences in benchmark performance with the only differences being time of day, day of week and peak vs. off-peak?

If there's such a low chance of everyone on a certain beam/gateway downloading at the same time, does it mean the Hughes network adjusts the overall system capacity to lower the level of service to less important areas to enable higher capacity for more important areas?

Since roughly half the footprint of my beam serves the open water of the Gulf of Mexico, I often wonder if Hughes allocates proper resources to service my area.
8 REPLIES 8
Tutor

Re: To deepdiver

Bumping since this is many pages back already.

If anyone else has thoughts on the grade of service/load balancing feature, don't be shy. I directed  this to deepdiver since he had things to say that I've not seen before.
Honorary Alumnus

Re: To deepdiver

Demographics certainly has to enter into it.

Look at a beam map of ES17:


Four beams per (presumed) automated Gateway overseen by two NOC's doing load balancing on the fly.

I too wonder how resources are allocated. I'm on beam 13 with a good share of it centered over Lake Michigan:

   

Possible choke points are many:

saturation of the satellite transponder itself

Gateway capacity

Gateway server capacity

Many of the answers are going to be proprietary in nature so supposition will have to suffice.

  

Distinguished Professor IV

Re: To deepdiver

I wonder if, even though at a given time the satellite itself could handle more traffic, the Gateway has it's own limit on throughput to keep the satellite balanced, so to speak.  Perhaps even the beam has a limit on throughput.  I imagine one or the other must have a limit, or the satellite itself has a limit for one or both, as it would make sense why one place can be bogged down while someone on another Gateway, or possibly even another beam within the same Gateway (I don't know), is having no slowdown to speak of. 

AMD Ryzen 5 3400G | XPG SX8200 Pro 500GB M.2 NVMe SSD | Western Digital Blue 500GB HDD | 16GB Corsair DDR4-3000 | Windows 10 Pro 64-bit
Tutor

Re: To deepdiver

Gwalk, you've done a lot of work on that map since the last time I saw it.

I wonder what's the difference between a gateway that can handle four beams and a gateway that can't even keep up with one?

Anybody have any gateway pictures or GPS coordinates of facilities?
Honorary Alumnus

Re: To deepdiver

All I have is a picture of the Germantown headquarters and that thing is just studded with dishes.
The above map is for ES17 beams. We also have to consider the 120 Spaceway3 beams plus wherever 7000s traffic is routed.
Tutor

Re: To deepdiver

And then all the beams from the new satellite have to be aimed somewhere. A little scary when considering a gateway itself and the infrastructure serving it may be the major choke point
Assistant Professor

Re: To deepdiver

Yea, I've been more concerned with the ground side of things once the new satellite is online. Wondering if my recent problem has to do with switching over to new equipment in order to do testing for the added capacity for the new satellite.
Distinguished Professor IV

Re: To deepdiver

A little scary when considering a gateway itself and the infrastructure serving it may be the major choke point
That's what I'm wondering.  I'm sure they are taking this into consideration, though, and upgrading them if needs be. 

AMD Ryzen 5 3400G | XPG SX8200 Pro 500GB M.2 NVMe SSD | Western Digital Blue 500GB HDD | 16GB Corsair DDR4-3000 | Windows 10 Pro 64-bit