From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from vsmx001.dclux.xion.oxcs.net (vsmx001.dclux.xion.oxcs.net [185.74.65.81]) (using TLSv1.2 with cipher ADH-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by lists.bufferbloat.net (Postfix) with ESMTPS id 6F8FE3B29E for ; Tue, 22 Feb 2022 02:42:44 -0500 (EST) Received: from proxy-2.proxy.oxio.ns.xion.oxcs.net (proxy-2.proxy.oxio.ns.xion.oxcs.net [197.248.130.130]) by mx-out.dclux.xion.oxcs.net (Postfix) with ESMTPA id C8F238C03F3; Tue, 22 Feb 2022 07:42:37 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=simple/simple; d=dclux.xion.oxcs.net; s=mail1; t=1645515763; bh=yoPQ4PtUJkc1dTlog0bAQ015AGyEPhN6Qw4VqY9yBbk=; h=Date:From:To:Cc:In-Reply-To:References:Subject:From; b=uv7+PNHQ221+g3hFxbDx7JRojZlkaPWimgwAdanSx6FSAOACZGWTRPqIROZzTF+f3 hg9heqL9us5s6pYGadF3RNeyKOYYvupBvavZGjgjCLSB7ECwhn9Hfn4qnYjZ8CL16q obrDaM1bRMNDNSskY+NwV75vv+CwUg1i3u1y/dfoEtZEb8PIug4LTZrsK62wEHlD5V TL3aKyYDRPu/yQSzoW8oKf/zLvHZA0A/LRQWHMGwD0fPqYraGOMLUX8+IhDzIp0Dcp TPUxgpMCO4I5oyXic1xMbSV/vPUQlObzZYxrEoC7wie+j6/zpLjvNt7SQ9C4L4SGa1 sZQ1B34D0N5sA== Date: Tue, 22 Feb 2022 10:42:28 +0300 From: Mike Puchol To: 'Daniel AJ Sokolov' , 'David Lang' , dickroy@alum.mit.edu Cc: starlink@lists.bufferbloat.net Message-ID: <80753e77-f7ba-466f-8222-66c16059f600@Spark> In-Reply-To: References: <1p492142-q944-r494-6s6r-p6q37s57qnq4@ynat.uz> <1F1EB112F8CB446FAB4BF308A76955FA@SRA6> <0ac195f5-3668-4c96-8dec-8a2d59a0bd52@Spark> <866405-s043-n12n-6pqs-46o38r189218@ynat.uz> <38pr9p5s-3ro4-49p9-9535-7o92oqrq62r1@ynat.uz> X-Readdle-Message-ID: 80753e77-f7ba-466f-8222-66c16059f600@Spark MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="621493ea_62bbd95a_2ed6" X-VadeSecure-Status: LEGIT X-VADE-STATUS: LEGIT Subject: Re: [Starlink] Starlink Roaming X-BeenThere: starlink@lists.bufferbloat.net X-Mailman-Version: 2.1.20 Precedence: list List-Id: "Starlink has bufferbloat. Bad." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 22 Feb 2022 07:42:44 -0000 --621493ea_62bbd95a_2ed6 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Content-Disposition: inline I did over-simplify so the point was better understood. On the optical ga= teways, these exist already:=C2=A0https://mynaric.com/products/ground-cap= abilities/ Once you have an optical mesh in orbit, the only practical way to provide= it with massive capacity is optical links - there isn=E2=80=99t enough r= adio spectrum that would do it (without a massive ground gateway network = with enough physical separation). You can create a network of optical gat= eways that guarantees a number of them will not be impared by cloud cover= at any given time. Optical has the advantage of being license-free, too.= Best, Mike On =46eb 22, 2022, 10:20 +0300, Dick Roy , wrote:= > > > =46rom: Starlink =5Bmailto:starlink-bounces=40lists.bufferbloat.net=5D = On Behalf Of Mike Puchol > Sent: Monday, =46ebruary 21, 2022 9:35 PM > To: Daniel AJ Sokolov; David Lang > Cc: starlink=40lists.bufferbloat.net > Subject: Re: =5BStarlink=5D Starlink Roaming > > Actually, laser links would make gateway connectivity *worse*. If we ta= ke the scenario attached, one gateway is suddenly having to serve traffic= from all UTs that were not previously under coverage. > > A satellite under full load can saturate two gateway links by itself. I= f you load, say, 20 satellites in an orbital plane, onto a single gateway= , over ISL, you effectively have 5% of each satellite=E2=80=99s capacity = available (given an equal distribution of demand, of course there will be= satellites with no UTs to cover etc.). > =5BRR=5D I think to do this analysis correctly; one needs to consider t= he larger system and the time-varying loads on the components thereof. Wh= at you say is true; just a bit over-simplified to be maximally useful. Ro= uting through complex congested networks is well-studied problem and hnts= at possible solutions can probably be found thereJ) > > > Eventually they will go for optical gateways, it=E2=80=99s the only way= to get enough capacity to the constellation, specially the 30k satellite= version. > =5BRR=5D What do you mean by =E2=80=9C=E2=80=9Doptical gateway=E2=80=9D= =3F An optical link from the satellite to the ground station=3F That woul= d be real expensive at least power-wise and unreliable. > > Best, > > Mike > On =46eb 22, 2022, 05:17 +0300, David Lang , wrote: > > On Mon, 21 =46eb 2022, Daniel AJ Sokolov wrote: > > On 2022-02-21 at 13:52, David Lang wrote: > > > They told me that I could try it, and it may work, may be degraded a > bit, or may not work at all. They do plan to add roaming capabilities i= n > the future (my guess is that the laser satellites will enable a lot mor= e > flexibility) > > Isn't that a very optimistic assessment=3F :-) > > Laser links are great for remote locations with very few users, but how= > could they relieve overbooking of Starlink in areas with too many users= =3F > > The laser links can reduce the required density of ground stations, but= > they don't add capacity to the network. Any ground station not built > thanks to laser links adds load to other ground stations - and, maybe > more importantly, adds load to the satellite that does eventually > connect to a ground station. > > Can laser links really help on a large scale, or are they just a small > help here and there=3F > > My thinking is that the laser links will make it possible to route the = traffic > from wherever I am to the appropriate ground station that I'm registere= d with as > opposed to the current bent-pipe approach where, if I move to far from = my > registered location, I need to talk to a different ground station. > > Currently there are two limits in any area for coverage: > > 1. satellite bandwidth > 2. ground station bandwidth > > laser links will significantly reduce the effect of the second one. > > We know that they can do mobile dishes (they are testing it currently o= n Elon's > gulfstream, =46AR more mobile that I will ever be :-) ) > > David Lang > =5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F= =5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F > Starlink mailing list > Starlink=40lists.bufferbloat.net > https://lists.bufferbloat.net/listinfo/starlink --621493ea_62bbd95a_2ed6 Content-Type: text/html; charset="utf-8" Content-Transfer-Encoding: quoted-printable Content-Disposition: inline
I did over-simplify so the point was better underst= ood. On the optical gateways, these exist already:&=23160;https://mynaric.com/products/ground-capabilities/

Once you have an optical mesh in orbit, the only practical way to provide= it with massive capacity is optical links - there isn=E2=80=99t enough r= adio spectrum that would do it (without a massive ground gateway network = with enough physical separation). You can create a network of optical gat= eways that guarantees a number of them will not be impared by cloud cover= at any given time. Optical has the advantage of being license-free, too.=

Best,

Mike
On =46eb 22, 2022, 10:20 +0300, Dic= k Roy <dickroy=40alum.mit.edu>, wrote:


=46ro= m: Starlink =5Bmailto:starlink-boun= ces=40lists.bufferbloat.net=5D On Be= half Of Mike Puchol
Sent: Monday, =46ebruary 2= 1, 2022 9:35 PM
To: Daniel AJ Sokolov; Dav= id Lang
Cc: starlink=40lists.buffe= rbloat.net
Subject: Re: =5BStarlink=5D= Starlink Roaming

&=23160;

Actual= ly, laser links would make gateway connectivity *worse*. If we take the s= cenario attached, one gateway is suddenly having to serve traffic from al= l UTs that were not previously under coverage.&=23160;

A satellite under full load can saturate two gateway links by itself. If = you load, say, 20 satellites in an orbital plane, onto a single gateway, = over ISL, you effectively have 5% of each satellite=E2=80=99s capacity av= ailable (given an equal distribution of demand, of course there will be s= atellites with no UTs to cover etc.).

J)<= /font>


=
Eventually they will go for optical gateways, it=E2=80=99s the only way t= o get enough capacity to the constellation, specially the 30k satellite v= ersion.


Best,

Mike

On =46eb 22, 2022, 05:17 +0300, Dav= id Lang <david=40lang.hm>, wrote:

On Mon, 21 =46eb 2022, Daniel AJ So= kolov wrote:

On 2022-02-21 at 13:52, David Lang = wrote:


They told me that I could try it, and it may work, may be degraded a
bit, or may not work at all. They do plan to add roaming capabilities in<= br /> the future (my guess is that the laser satellites will enable a lot more<= br /> flexibility)


Isn't that a very optimistic assessment=3F :-)

Laser links are great for remote locations with very few users, but how could they relieve overbooking of Starlink in areas with too many users=3F=

The laser links can reduce the required density of ground stations, but they don't add capacity to the network. Any ground station not built
thanks to laser links adds load to other ground stations - and, maybe
more importantly, adds load to the satellite that does eventually
connect to a ground station.

Can laser links really help on a large scale, or are they just a small help here and there=3F


My thinking is that the laser links will make it possible to route the tr= affic
from wherever I am to the appropriate ground station that I'm registered = with as
opposed to the current bent-pipe approach where, if I move to far from my=
registered location, I need to talk to a different ground station.

Currently there are two limits in any area for coverage:

1. satellite bandwidth
2. ground station bandwidth

laser links will significantly reduce the effect of the second one.
=
We know that they can do mobile dishes (they are testing it currently on = Elon's
gulfstream, =46AR more mobile that I will ever be :-) )

David Lang
=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F= =5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F
Starlink mailing list
Starlink=40lists.bufferbloat.net
https://lists.bufferbloat.net/listinfo/starlink

--621493ea_62bbd95a_2ed6--