[ih] Internet at Sea

[vinton cerf]

see below re AALPS

On Fri, Oct 3, 2025 at 7:26 PM Barbara Denny via Internet-history <
internet-history at elists.isoc.org> wrote:

>  I am not going to try to answer Jack's message   but I am glad to see he
> brought up MATNET.  I was going to ask about that.  SRI   had a contract
> with the Navy to implement a gateway in Ada; and design, implement, and
> test a way to support ships, subs, etc when they were under EMCON (emission
> control) - an early application of delay tolerant networking.  I was
> working on the gateway.  Unfortunately something changed within the navy so
> that our contract was cancelled early. However I was tagged slightly later
> to support installing a couple Sun? workstations on board the Carl Vinson
> to support them looking at an application I think another group at SRI had
> developed.  I assume the application used MATNET in some way (  I think the
> workstations used Ethernet to get to the rest of the onboard  network).
> One guy in our group did remain on board to support the equipment when the
> ship left Alameda for Hawaii (He got off the ship in Hawaii). At that point
> in time,  women weren't even allowed on board but they made an exception
> for me and one other person.  They also gave us a great tour of the ship.
> This was probably in 1984 or slightly later
>

SRI developed an Advanced Airborne Load Planning System to demonstrate the
utility of having access to computing power on the Internet while literally
on the tarmac at the 18th Airborne Corps. The system could load plan in
seconds the entire 18th Airborne using AALPS. Eventually a flyoff was done
with the regular load masters and they concluded the system did better than
they did. It could also include low altittude air extraction (eg, using a
parachute to pull a tank out of the loading bay) and was smart enough to
organize the loading to offload the equipment before the pax (passengers)
so we didn't drop equipment on the troops.

vint

> Another somewhat  related question is what other impact having these
> networks had on application development, especially for the military. I am
> trying to see if I can untangle the history of an air load planning system
> (ALPS)  that SRI did. I think it may be related to the testbed at Fort
> Bragg that included packet radio. It may have also been an early use of AI
> (or whatever was thought of as AI in those days).
> barbara
>
>     On Friday, October 3, 2025 at 02:16:18 PM PDT, Jack Haverty via
> Internet-history <internet-history at elists.isoc.org> wrote:
>
>  Michael's suggestion of near-port traffic increasing with people
> planning their shore time sounds reasonable to me.  There's a lot of
> possible "social" influences on traffic patterns.   I remember there was
> one net we installed (for a credit card processor IIRC) where it was
> possible to see such effects - e.g., bursts of traffic during ad breaks
> in the Super Bowl.
>
> I'm curious about the history preceding today's world.  Back in the
> early 1980s, the Internet reached a "cruise ship", possibly the first
> such maritime connection on the Internet.  The ship involved was the USS
> Carl Vinson - the huge aircraft carrier that the US Navy used as a
> testbed for new technology "in the field."   The project was called
> MATNET - Mobile Access Terminal Network, with nodes on shore and on the
> ship.  The technology was a derivative of SATNET - the satellite network
> with linked the US and Europe over Intelsat-IV.   IIRC, MATNET used
> FLTSATCOM satellite resources.
>
> MATNET was the next step in the "pipeline" from research to operational
> use within the US (and probably others such as UK) Navy.  Various Packet
> Radio projects filled a similar pipeline for terrestrial use in the
> Army.   I'm not sure but I think another set of projects involved the
> Air Force; I recall one meeting at SAC HQ in Nebraska to work out some
> logistics for a testbed involving air assets, including the Looking
> Glass.   The vision extended to some future combination of Packet Radio
> and Satellite technologies, i.e., "Packet Radio in the Sky", with
> terrestrial nodes talking to an overhead satellite, and satellites
> talking to each other to get datagrams to their destinations.
>
> Some information about MATNET is here:
> https://apps.dtic.mil/sti/tr/pdf/ADA122596.pdf
> https://apps.dtic.mil/sti/tr/pdf/ADA222356.pdf
>
> Now that we have systems like Starlink, it seeme we may now have "Packet
> Radio in the Sky" -- but I don't know the details of how Starlink
> actually works.
>
> So, my question for History is "How did we get from MATNET and Packet
> Radio to the Internet systems that are operational today?"
>
> E.g., did MATNET and/or Packet Radio have any influence on today's
> technology?   How did that happen?  Perhaps the protocols and/or
> algorithms migrated and ended up in today's systems?  Perhaps the people
> involved migrated and took their experiences with them?   Or perhaps
> today's systems were built by a new generation who never even heard of
> those systems from the 1980s?
>
> BBN reorganized (frequently) and I lost contact with the research work
> after mid-1983.   I know how the ARPANET migrated to become the DDN.
> And TCP/IP which annihilated all of its competitors.   But Packet
> Radio?   MATNET?
>
> Anybody know what happened to the other networking technologies that
> ARPA drove into pipelines to field deployment, and what happened after
> those testbeds?   How much of today's Internet technology (such as
> Starlink) contains ARPA DNA?
>
> /Jack Haverty
>
> PS - It occurred to me that Detectives "follow the money" to investigate
> crimes.   Investors "follow the people" to place their bets.  I wonder
> if Internet Historians have to "follow the code".
>
> On 10/3/25 10:26, Karl Auerbach via Internet-history wrote:
> > Some years back (circa 2010?) we also took a cruise and were annoyed
> > by how slow was access to the net.
> >
> > So we made a video about it:
> >
> > https://vimeo.com/815203660
> >
> >         --karl--
> >
> > On 10/3/25 2:05 AM, Michael Grant via Internet-history wrote:
> >> From "Jack Haverty via Internet-history"
> >> <internet-history at elists.isoc.org>
> >>> Recently I heard anecdotal reports that the Internet on cruise ships
> >>> works well - but is reliable only when the ship is far out to sea.
> >>> When it's in port, or even just approaching port, teleconferencing
> >>> is unreliable.   My speculation is that traffic loads when near a
> >>> port include all the land-based users and the network may be
> >>> overwhelmed.  But that's just speculation, I have no data.
> >>>
> >> I have spent quite a bit of time as passenger on cruise ships these
> >> last few years.  I don't know when the Internet started becoming
> >> available on cruise ships but I first used it on-board in 2010 on
> >> Holland America.  Simply put, the service was awful and expensive
> >> regardless if the ship was in port or not.  It used a geostationary
> >> satellite connection using a dish on a rocker under a ball.  When the
> >> ship moved a lot, the internet stopped working. It was charged by the
> >> minute and you had to connect like dial-up through the ship's captive
> >> web portal and manually disconnect when no longer using it.  In 2010,
> >> iirc, the cost was something like 50 US cents/min and you spent most
> >> of your money waiting.
> >>
> >> Since then, the geostationary ship internet has gotten much better.
> >> 3 or 4 years ago, I was on a cruise ship which which no longer
> >> charged a per minute charge but still used the similar geostationary
> >> link and you still had to "log on" via a captive portal.  We were
> >> able to have absolutely acceptable video calls part of the time.  I
> >> was amazed because even with the delay through the geo satellite, the
> >> delay was unnoticeable. I had to prove to myself using ping and
> >> traceroute that we were actually using a geostationary link and not
> >> Starlink!
> >>
> >> In the last year, I have been on cruise ships that now have
> >> Starlink.  They have 12 stationary Starlink pizza box antennae
> >> mounted high up near one of the stacks (6 on each side).  They run 12
> >> separate Starlink connections.  I am going to describe what Cunard
> >> does but I suspect they are all similar.  Cunard runs a VPN which
> >> bonds together the 12 connections such that if any of them goes down
> >> the packet is sent out any antenna.  (I don't know which VPN they use
> >> nor do I know how they bond the links, if it's round-robin or not,
> >> sorry!)  The VPN is terminated in either Southampton UK or Miami
> >> Florida, US depending on where the ship is (they change over midway
> >> across the Atlantic).  They run something like bufferbloat to share
> >> the b/w more fairly.  They also run a firewall where they block quite
> >> a few sites.  I have had to ask them to unblock things which they
> >> have kindly done.
> >>
> >> I understand that Starlink sets the pricing on the ships and that
> >> they may have installed the system for little or nothing in exchange
> >> for being able to capture that market.  I don't know for sure
> >> though.  I do know a lot of cruise ships are installing Starlink
> >> now.  My only gripe is they charge per device and it's 2x if you want
> >> 2 devices, so for my wife and I, if we both just want to be able to
> >> use whatsapp to find one another on-board, it's well overpriced.
> >> They charge about USD $25/day per device and have a "special" price
> >> where for the cost of 2 devices you can have up to 4.
> >>
> >> The ships do not disable the satellite internet near ports. They
> >> continue to use the satellite network even in port.  This is likely
> >> because if they were to somehow connect the ship to the port's wifi
> >> they would quickly saturate it.  It's also likely because several of
> >> the ship's systems use their satellite internet and for security
> >> reasons they want that going over their VPN.  I have not noticed much
> >> difference sitting in port using the satellite internet than at sea,
> >> if anything, it's better since many people are off the ship.  I
> >> suspect what Jack may be referring to is when approaching port people
> >> saturate the ship's internet as they prepare to disembark, like doing
> >> research for things to do or coordinating with friends and family and
> >> such.  I have my doubts it has much to do with interference other
> >> than maybe if it's Starlink that the ship starts to compete with
> >> other Starlink users on land using the same space based resource in
> >> the sky.
> >>
> >> Internet in the cabins has also gotten a lot better over the years.
> >> Cunard has installed Aruba (HP) APs in all the hallways and cabins.
> >> Rather unfortunately one cruise I was on, the wifi in my cabin would
> >> literally seize up.  I spent quite a bit of time tracking it down to
> >> a firmware issue that Aruba had issued a patch for but unfortunately
> >> it was not possible to update the firmware at sea because they would
> >> have had to take the entire ship off-line while they did it so it
> >> would have to wait until next time the ship went in for maintenance.
> >>
> >> Every cruise ship I have been on in the last 10-15 years also had
> >> separate mobile phone and data available via Maritime
> >> Telecommunications Network (MTN) or other similar companies. This is
> >> only turned on when at sea.  At port, they turn off the maritime
> >> telecom network and you are expected to roam on to the local carrier.
> >>
> >> I am fairly certain this uses a separate geostationary satellite and
> >> separate dish-on-rocker-in-ball antenna from the internet
> >> connection.   Receiving text messages (SMS) has always been free for
> >> me and that's good because the number of times that's saved me when
> >> on-board and needing to go through some 2FA to get into something has
> >> been plenty.  However, making and receiving calls, sending texts, and
> >> especially using the mobile data is scarily expensive.  I have talked
> >> with plenty of passengers who had no idea and those who had used it
> >> unwittingly on previous cruises only to gotten home to $1000+ phone
> >> bills!  Unfortunately, unless you are a little bit savvy, it's all to
> >> easy to just use it because it works.
> >>
> >> Ships also seem to have Inmarsat and/or Iridium phones on board but
> >> as far as I am aware this is  not hooked up to some general system
> >> for the public.  It seems like this is backup or ship to shore coms
> >> for the bridge.
> >>
> >> All of my experience above is solely from my experiences as a
> >> passenger. Everything I know here is just from chatting up tech
> >> people on board ships and my own probing around with tools like ping
> >> and traceroute.  Unfortunately, I never worked in the maritime or
> >> satellite networking industry myself.
> >>
> >> I'd be interested in hearing some of the technical details if someone
> >> on this list knows more.  It would also be interesting to see a
> >> thread of the history of internet on air crafts if someone knows
> >> about that.
> >>
> >> Michael Grant
>
> --
> Internet-history mailing list
> Internet-history at elists.isoc.org
> https://elists.isoc.org/mailman/listinfo/internet-history
> -
> Unsubscribe:
> https://app.smartsheet.com/b/form/9b6ef0621638436ab0a9b23cb0668b0b?The%20list%20to%20be%20unsubscribed%20from=Internet-history
>
> --
> Internet-history mailing list
> Internet-history at elists.isoc.org
> https://elists.isoc.org/mailman/listinfo/internet-history
> -
> Unsubscribe:
> https://app.smartsheet.com/b/form/9b6ef0621638436ab0a9b23cb0668b0b?The%20list%20to%20be%20unsubscribed%20from=Internet-history
>