[Starlink] Starlink for Tonga?

Ulrich Speidel ulrich at cs.auckland.ac.nz
Mon Feb 7 18:36:19 EST 2022 

Always interesting to hear how dishy actually behaves - which may depend 
a bit on where you are. In the conventional service areas, there are 
generally multiple gateways, so there's probably no benefit in probing 
east or west, you'll just need to find north or south and the right 
elevation. But that wouldn't be the case where all satellites with 
gateway connectivity are relatively far to the west of you.

FAIK FINTEL's Suva facility hasn't been affected (only outlying islands 
in Fiji have been hit by tsunami with damage resulting, and it's been 
largely minor).

However, the FINTEL teleport you see on Google maps isn't Starlink 
capable, and it's not a given that it would be available as a site 
either, although there seems to be land around it that could be used I 
suppose.

Starlink teleports look a bit like an egg carton from the air, with 
typically 6 to 9 radome "eggs" in them, and these would have to be 
installed there first. This is the closest known one to Tonga at present:

https://goo.gl/maps/5Risyqk9tC2QDJwB8

(Yes, as irony would have it, on the back paddock of an egg farm ;-))

As I said in a previous post: As attractive as the geometry with a lower 
elevation might look, it carries with it the price of serious link 
deterioration, especially in an area like this where rain fade is 
common. Satellite link budgets tend to be designed with small fade 
margins because every extra dB adds to cost in a significant way, and by 
having both a longer path and more of it running the gauntlet through 
the atmosphere, unusual setups like this one are usually a bit hard to 
do in practice.

Moreover, news just in minutes ago from Tonga Cable:

> Reliance still out there during this bad weather (as they’re behind 
> schedule) trying to recover the part of the cable from Suva….
>
>
The Tongan end of the cable has been found and secured. Cable is 
presumed to be broken in multiple places, though, so they'll probably 
have to lay a stretch of new cable. I'd expect them to find the Suva end 
within the next day or so, weather permitting, and then it'll be mostly 
a matter of whether they can put the cable back where it was. If not, it 
could take a little longer, but I'm optimistic that connectivity will be 
restored within the week.


On 8/02/2022 11:29 am, Mike Puchol wrote:
> You are right in your comments, however - Dishy does not currently 
> shift in azimuth, only in elevation in the North/South axis, in 
> essence, to focus its boresight towards the area of the sky where it 
> will see most viable satellites. If you play around with Tonga (Fiji 
> emergency gateway is live on the site, plus a couple of fixes that 
> returned negative azimuths!), you will see that North/South tilt has 
> no real effect, but if you shift azimuth towards Fiji (from Tonga) 
> your candidate satellites increase. You can also see the effect of 
> 15º minimum elevation, which suddenly make satellites on NZ gateways 
> available to a Tonga terminal too.
>
> As far as placing a gateway in Fiji, it already has a teleport 
> facility, which will have power and fibre (unless that one has been 
> taken out too?). Check https://goo.gl/maps/6BYXf4R17yys7zNe9 
> <https://goo.gl/maps/6BYXf4R17yys7zNe9>
>
> In terms of plain service, it’s true the GSO operators have stepped 
> in, and for actual emergencies, it’s sufficient. Starlink can start 
> supplementing that with more fibre-like backhaul which would enable 
> those additional services you mention.
>
> Best,
>
> Mike
> On Feb 7, 2022, 23:23 +0100, Ulrich Speidel 
> <ulrich at cs.auckland.ac.nz>, wrote:
>>
>> Nice tracker, Mike!
>>
>> That said: It's not just a matter of geometry though. Presuming that 
>> a dishy aims itself at whichever portion of the sky it seems most 
>> satellites (with gateway service), there's another aspect to the 
>> elevation angle. The lower your elevation, the longer the path that 
>> the signal has to take through the atmosphere, and the higher the 
>> degree of service disruption as a result of rain fade, or in Tonga's 
>> case, potential ash cloud fading. In Fiji's case, Starlink also has 
>> to find a site with low elevation take-off to Tonga AND fibre 
>> connectivity. Plus they have to get past the regulator in a country 
>> that's know to regulate heavily and that currently experiences a lot 
>> of Covid-related disruption with people in badly connected home office.
>>
>> Note that Tonga has had emergency service for a while now: Intelsat, 
>> SES and Kacific are all at present providing service (Kacific at 1 
>> Gbps even). That's plenty for emergency use.
>>
>> The big sticky point is that mobile Internet isn't working, people 
>> have to go to Tonga Telecom offices for access or wifi. That's an 
>> issue because 4 out of 5 households in Tonga rely on remittances from 
>> family overseas, which are usually sent via Western Union. Moneygram 
>> etc., whose apps are out of action right now.
>>
>> And yes, the date line is fun. Pacific countries are known for having 
>> shifted it back and forth for spurious reasons, which is why it's 
>> badly bent, and these days I'm still grateful if people have maps 
>> online that don't end at the date line ;-)
>>
>> On 8/02/2022 10:36 am, Mike Puchol wrote:
>>> Placing a gateway at the Fiji teleport results in satellites that 
>>> cover Tonga being serviced. However, satellite density, but most 
>>> importantly, GSO protection, take out a significant portion of the 
>>> coverage from the satellites.
>>>
>>> They would need to lower minimum elevation (now 25° everywhere) to 
>>> improve the situation. With things as-is, Tonga would be covered by 
>>> 1-2 satellites which should be enough for emergency service restoration.
>>>
>>> You can play around with this on my tracker at https://starink.sx 
>>> <https://starink.sx>, 
>>> I will be adding the temporary gateway in a few minutes.
>>>
>>> Please not that the date line causes some weird issues with the map 
>>> and the algorithms, which I’m still trying to fix. You can thank 
>>> whoever thought moving from -180° to +180° across a line was a good 
>>> idea.
>>>
>>> Best,
>>>
>>> Mike
>>> On Feb 7, 2022, 22:24 +0100, David Lang <david at lang.hm>, wrote:
>>>> the dish aimes at where it sees the most satellites, not 
>>>> necessarily 53. I took
>>>> my dish to a campground where there were trees to the north and it 
>>>> ended up
>>>> pointing straight up (Los Angeles area) performance was fine.
>>>>
>>>> when you power it on, the dish tilts and swivels to point straight 
>>>> up (no idea
>>>> what the angle logic is), and after a few min of watching the sky 
>>>> will re-aim
>>>> itself if/as needed.
>>>>
>>>> David Lang
>>>>
>>>> On Mon, 7 Feb 2022, Ben Greear wrote:
>>>>
>>>>> Date: Mon, 7 Feb 2022 13:18:49 -0800
>>>>> From: Ben Greear <greearb at candelatech.com>
>>>>> To: starlink at lists.bufferbloat.net
>>>>> Subject: Re: [Starlink] Starlink for Tonga?
>>>>>
>>>>> 90 or even much smaller percentage is a lot better than zero.
>>>>>
>>>>> I wonder if they can point the dish towards the horizon to pick up 
>>>>> the sat
>>>>> where
>>>>> it can best see the functional downlink. I realize the dish 
>>>>> auto-points
>>>>> itself
>>>>> now, but surely engineers that can design that can also design an 
>>>>> 'off'
>>>>> switch
>>>>> for that and let their on-the-ground folks do some hacking....
>>>>>
>>>>> Thanks,
>>>>> Ben
>>>>>
>>>>> On 2/7/22 11:05 AM, Christian von der Ropp wrote:
>>>>>> But in practice the satellites won't be sitting and waiting at 
>>>>>> the edge of
>>>>> this 940km radius. They are moving in and out the radius and the 
>>>>> question is
>>>>> if
>>>>>> satellite density is high enough so that once the serving 
>>>>>> satellite loses
>>>>> its gateway link there's another satellite in the 940km radius 
>>>>> which also
>>>>> covers
>>>>>> Tonga. And then this new satellite cannot be within certain elevation
>>>>> angles (~60-80° at 0° azimuth) where the geostationary arc crosses 
>>>>> Fijian
>>>>> skies and the
>>>>>> gateway antennas have to seize emission. My gut feeling is that
>>>>> availability in Tonga would be <90% simply because it's too far 
>>>>> out at the
>>>>> edge of a Fijian
>>>>>> gateway's range where there will be frequent service interruptions.
>>>>>>
>>>>>>
>>>>>> Am 07.02.2022 um 19:51 schrieb Nathan Owens:
>>>>>>> The current coverage radius of a gateway/ground station with a 
>>>>>>> 25 degree
>>>>> minimum elevation is ~940km, so nothing in theory.
>>>>>>>
>>>>>>> On Mon, Feb 7, 2022 at 10:50 AM Daniel AJ Sokolov 
>>>>>>> <daniel at sokolov.eu.org>
>>>>> wrote:
>>>>>>>
>>>>>>> Hello,
>>>>>>>
>>>>>>> I hope this is not offtopic: Starlink wants to build a ground 
>>>>>>> station
>>>>> on
>>>>>>> Fiji to supply Tonga with internet.
>>>>>>>
>>>>>>> The distance between Tonga and Fiji is about 750 km minimum. That's
>>>>>>> quite the distance.
>>>>>>>
>>>>>>> What does Starlink have to do to make this work?
>>>>>>>
>>>>>>> Cheers
>>>>>>> Daniel
>>>>>>> _______________________________________________
>>>>>>> Starlink mailing list
>>>>>>> Starlink at lists.bufferbloat.net
>>>>>>> https://lists.bufferbloat.net/listinfo/starlink 
>>>>>>> <https://lists.bufferbloat.net/listinfo/starlink>
>>>>>>>
>>>>>>>
>>>>>>> _______________________________________________
>>>>>>> Starlink mailing list
>>>>>>> Starlink at lists.bufferbloat.net
>>>>>>> https://lists.bufferbloat.net/listinfo/starlink 
>>>>>>> <https://lists.bufferbloat.net/listinfo/starlink>
>>>>>>
>>>>>>
>>>>>> _______________________________________________
>>>>>> Starlink mailing list
>>>>>> Starlink at lists.bufferbloat.net
>>>>>> https://lists.bufferbloat.net/listinfo/starlink 
>>>>>> <https://lists.bufferbloat.net/listinfo/starlink>
>>>>>>
>>>>>
>>>>>
>>>>>
>>>> _______________________________________________
>>>> Starlink mailing list
>>>> Starlink at lists.bufferbloat.net
>>>> https://lists.bufferbloat.net/listinfo/starlink 
>>>> <https://lists.bufferbloat.net/listinfo/starlink>
>>>
>>> _______________________________________________
>>> Starlink mailing list
>>> Starlink at lists.bufferbloat.net
>>> https://lists.bufferbloat.net/listinfo/starlink
>>>
>> --
>> ****************************************************************
>> Dr. Ulrich Speidel
>>
>> School of Computer Science
>>
>> Room 303S.594 (City Campus)
>> Ph: (+64-9)-373-7599 ext. 85282
>>
>> The University of Auckland
>> ulrich at cs.auckland.ac.nz   
>> http://www.cs.auckland.ac.nz/~ulrich/
>> ****************************************************************
>>
>>
>>
>> _______________________________________________
>> Starlink mailing list
>> Starlink at lists.bufferbloat.net
>> https://lists.bufferbloat.net/listinfo/starlink 
>> <https://lists.bufferbloat.net/listinfo/starlink>

-- 
****************************************************************
Dr. Ulrich Speidel

School of Computer Science

Room 303S.594 (City Campus)
Ph: (+64-9)-373-7599 ext. 85282

The University of Auckland
ulrich at cs.auckland.ac.nz  
http://www.cs.auckland.ac.nz/~ulrich/
****************************************************************


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.bufferbloat.net/pipermail/starlink/attachments/20220208/b769bdae/attachment-0001.html>

More information about the Starlink mailing list