Starlink has bufferbloat. Bad.
 help / color / mirror / Atom feed
From: Sauli Kiviranta <smksauli@gmail.com>
To: David Lang <david@lang.hm>
Cc: Ulrich Speidel <u.speidel@auckland.ac.nz>,
	"starlink@lists.bufferbloat.net" <starlink@lists.bufferbloat.net>
Subject: [Starlink] Re: Starlink and Iran
Date: Thu, 15 Jan 2026 13:59:06 +0200	[thread overview]
Message-ID: <CAJbqEYDAkEfWr_+sUtFHd9gTuQJuAQF1eEyvxNOJom0a7vbiuQ@mail.gmail.com> (raw)
In-Reply-To: <q459rp19-7ppq-2o1p-5128-n4088op69o0n@ynat.uz>

"I have seen reports that SpaceX is in all-hands-on-deck mode to work to
defeat
the jammers. I expect that this will include things like pointing the beams
at
different dishy cells than they normally would use, etc."

Do we have any technical characteristics about the jamming how it is
visible to the application layer? I saw some X posts talk about 20% packet
loss, some said 80%.

Anyone from SpaceX can comment on what is the challenge "requirements" for
contributions?

Best regards,
Sauli

On Thu, Jan 15, 2026 at 1:17 PM David Lang via Starlink <
starlink@lists.bufferbloat.net> wrote:

> Authorized or not, starlink service has been turned on in Iran, I would
> assume
> that D2C is enabled as well.
>
> D2C is not going to be good for livestreaming video, it will support text
> and
> voice communications
>
> Jamming the D2C signal is far easier than jamming the dishys, all you need
> is
> to have the towers running with a stronger signal. (the phones will
> connect to
> the tower with the strongest signal). It's also easier to jam the D2C
> satellites
> (wider beamwidth and fewer that need to be jammed)
>
> I don't think they would need to get new sims, esims, etc in place, they
> just
> need to instruct the satellites to accept any sim rather than only
> authorized
> sims.
>
> creative use of aluminum foil should be able to shield your phone from the
> ground based towers, leaving them nothing other than the satellites for
> the
> phone to connect to (this will also block bluetooth, but wired connections
> to
> the phones will work)
>
>
> In terms of tracking/jamming the dishys normal signal, I think it would be
> easier to track/jam their wifi signal, those default to SSID STARLINK and
> will
> be in a known MAC range. disabling wifi and only working with a wired
> connection
> is going to be much safer (admittedly, less convienient, but when they are
> threatening to kill you if you use a dishy, you should batch upload via a
> wired
> connection, not try to livestream the protests, at least, not unless you
> are
> truely mobile)
>
> At the beginning of the invasion of Ukraine, Elon said that people needed
> to be
> careful, have a hill or other obstruction between the dishy and the bad
> guys,
> don't leave it running all the time, etc (basic signal security)
>
> I have seen reports that SpaceX is in all-hands-on-deck mode to work to
> defeat
> the jammers. I expect that this will include things like pointing the
> beams at
> different dishy cells than they normally would use, etc.
>
> I could see GPS jamming being a problem, if the dishy thinks it's
> somewhere
> other than where it really is, how successful will it be in connecting to
> the
> right satellites?? there are ways to work around this, and that may also
> be what
> starlink is working on.
>
> David Lang
>
>
>
> Ulrich Speidel wrote:
>
> > Jamming: Over every location in Iran, there are several dozen Starlink
> > satellites visible at any one time that Dishys on the ground can in
> principle
> > communicate with (read: 25 deg plus above the horizon and clear of the
> > geostationary arc). There are purportedly tens of thousands of Dishys in
> the
> > country. Each of those Dishys (when working) communicates with one of
> the
> > satellites, and does so by pointing a beam at the satellite - which
> points a
> > beam back. Even two Dishys in close vicinity of each other generally
> talk to
> > different satellites.
> >
> > To jam communication between a Dishy and a satellite you have to insert
> the
> > jammer into the transmission path - either by pointing it at the
> satellite's
> > receiver, or by pointing it at the Dishy. In either case, you want to do
> so
> > ideally from the direction of the respective transmitter that the
> receiver is
> > listening to, because there isn't all that much sensitivity if you're
> jamming
> > off beam. Basically, because signal power drops of with the square of
> the
> > distance, you need to be fairly close to a Dishy in order to out-shout
> the
> > transmitter at the far end of the beam if you're trying to jam from
> outside
> > the beam.
> >
> > Jamming the main traffic channels to Dishy is an uphill task - for a
> total
> > blackout, you'd have to cover a good part of the 2 GHz total in Ku with
> > sufficient power in terms of spectral density to cause mischief. Not
> easy.
> >
> > Likewise, pointing your jammer at the satellite(s) is mission impossible
> > because there's every chance that the satellites will be listening to
> Dishys
> > that are in a different direction from you.
> >
> > There would I guess be some opportunity in terms of jamming management
> > channels (e.g., access grant channel) but even this is complex with that
> > number of satellites around.
> >
> > Plus those babies move, so you need jammers that can track. And 15
> seconds
> > after you're worked out what you need to point where, Starlink changes
> the
> > game on you.
> >
> > The Independent quotes "a specialist in digital repression and associate
> > director of the Technology Threats and Opportunities Program at Witness"
> >
> https://www.independent.co.uk/news/world/middle-east/iran-internet-blackout-protest-starlink-musk-b2900101.html
> > saying that they think it's GPS jamming. For all I know Starlink doesn't
> need
> > GPS - while Dishys have GPS receivers, Starlink's got its own
> positioning
> > system, too. Also, GPS jamming is fairly common globally and it's not
> known
> > to be impairing Starlink all that much.
> >
> > D2C: Maybe someone should have asked the NZ Commerce Commission for
> advice
> > first. They figured out a long time ago that D2C isn't quite there yet
> (and
> > may never get fully there). It's only capable of supporting a
> comparatively
> > small number of devices per unit area on the ground, and apart from a
> small
> > number of premium phones, with text and perhaps RCS/MMS only. And that's
> with
> > a telco on the ground that's actually cooperating and making frequencies
> > available. One NZ, the New Zealand telco who partners with SpaceX on the
> D2C
> > here, had its marketing department shouting the virtues from the
> rooftops
> > until the Commerce Commission filed criminal charges. They're still in
> the
> > game but when I bought a new mobile phone the other day, it took me
> several
> > minutes to find the page that listed compatible phones. The service is
> now a
> > little less prominent on their home page - you have to scroll a little
> to
> > find it. Also, word doesn't seem to have gotten around that your mobile
> phone
> > - even if satellite-capable - will connect to terrestrial networks with
> > priority. So Iranians would have to go pretty far out into the desert
> just to
> > TXT. Oh, and cellphones are a lot easier to jam than Dishys...
> >
> > Of course, that's not the only consideration here. Using Starlink is
> illegal
> > in Iran, and I guess getting caught with a Dishy is a bit risky there at
> the
> > moment. But RF direction finding 50k+ Dishys that change transmit
> frequency a
> > couple of times a minute isn't trivial: It requires specialised
> equipment and
> > skill, both of which are likely to be in short supply at the moment. So
> I
> > suppose visual identification of Dishys (from the air or high rise
> buildings)
> > might be a more promising tactic. But of course they can be camouflaged
> to an
> > extent as well as moved.
> >
> > Also, Starlink tends to be more of a technology for underserved rural
> areas
> > rather than cities in countries with high Internet penetration rates -
> which
> > Iran is one of. So it's likely that many of the tens of thousands of
> Dishys
> > are in rural locations where there haven't been any large protests.
> >
> >
> >
> _______________________________________________
> Starlink mailing list -- starlink@lists.bufferbloat.net
> To unsubscribe send an email to starlink-leave@lists.bufferbloat.net
>

  reply	other threads:[~2026-01-15 11:59 UTC|newest]

Thread overview: 52+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-01-15  9:51 [Starlink] Starlink and Iran Ulrich Speidel
2026-01-15 10:06 ` [Starlink] " Inemesit Affia
2026-01-15 10:30   ` Ulrich Speidel
2026-01-15 10:44     ` Inemesit Affia
2026-01-15 11:16       ` Ulrich Speidel
2026-01-15 10:32 ` Inemesit Affia
2026-01-15 10:51   ` Ulrich Speidel
2026-01-15 11:17 ` David Lang
2026-01-15 11:59   ` Sauli Kiviranta [this message]
2026-01-15 14:08     ` David Lang
2026-01-15 15:29       ` Sauli Kiviranta
     [not found]   ` <3af2ac06-e098-4c79-869d-9c389959ca07@gmail.com>
     [not found]     ` <q9304244-661o-3qsr-o6rp-9q1nqq09r419@ynat.uz>
     [not found]       ` <4ba64a41-bbbf-4fb5-adb0-c77c15e4ca0f@gmail.com>
2026-01-15 16:20         ` Inemesit Affia
2026-01-15 20:12   ` Ulrich Speidel
2026-01-15 17:10 ` J Pan
2026-01-15 20:07   ` Ulrich Speidel
2026-01-15 21:47     ` Oleg Kutkov
2026-01-16  4:18       ` Ulrich Speidel
2026-01-16  8:12         ` Frantisek Borsik
2026-01-16  8:24           ` Inemesit Affia
  -- strict thread matches above, loose matches on Subject: below --
2026-01-15 14:50 David Fernández
2026-01-15 16:11 ` Oleg Kutkov
2026-01-15 17:13   ` J Pan
     [not found] <176849731431.1249.14387618908540773471@gauss>
2026-01-15 17:42 ` Colin_Higbie
2026-01-15 18:56   ` Jim Forster
2026-01-15 20:15     ` Ulrich Speidel
2026-01-15 20:27   ` Ulrich Speidel
2026-01-15 20:30     ` Hayden Simon
2026-01-15 21:06       ` Ulrich Speidel
2026-01-15 21:09         ` Hayden Simon
2026-01-15 21:20           ` Ulrich Speidel
2026-01-15 21:23             ` Hayden Simon
     [not found] <176851123059.1249.8585659892308012167@gauss>
2026-01-15 21:49 ` Colin_Higbie
2026-01-15 23:15   ` Frantisek Borsik
2026-01-16  0:13   ` Ulrich Speidel
2026-01-16  1:29     ` David Lang
2026-01-16 22:55       ` Frantisek Borsik
2026-01-16 23:06         ` J Pan
     [not found]     ` <13187.1768590201@obiwan.sandelman.ca>
2026-01-16 23:30       ` Ulrich Speidel
2026-01-17  0:07         ` David Lang
2026-01-17 21:56           ` Ulrich Speidel
2026-01-19 20:39             ` David Lang
2026-01-28  3:09               ` Ulrich Speidel
2026-01-28  3:30                 ` David Lang
2026-01-28  4:02                   ` Mike Puchol
2026-01-28  9:05                     ` Ulrich Speidel
2026-01-28  9:53                       ` David Lang
2026-01-28 20:43                         ` Ulrich Speidel
2026-01-28 20:55                           ` David Lang
2026-01-17 18:32         ` Michael Richardson
2026-01-17 18:38           ` Inemesit Affia
2026-01-17 19:25             ` Michael Richardson
2026-01-17 22:12           ` Ulrich Speidel

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: https://lists.bufferbloat.net/postorius/lists/starlink.lists.bufferbloat.net/

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=CAJbqEYDAkEfWr_+sUtFHd9gTuQJuAQF1eEyvxNOJom0a7vbiuQ@mail.gmail.com \
    --to=smksauli@gmail.com \
    --cc=david@lang.hm \
    --cc=starlink@lists.bufferbloat.net \
    --cc=u.speidel@auckland.ac.nz \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox