Conversation

Notices

1. Embed this notice
   Strypey (strypey@mastodon.nzoss.nz)'s status on Friday, 07-Mar-2025 12:29:18 JST Strypey

   Anyone out there know about satellite internet tech? How many geostationary satellites would it take to provide a wireless net connection to every net-enabled device in Aotearoa?

   Is a satellite a realistic thing for a forward-looking government to pay for, to guarantee sovereign intra-government communications? Plus emergency comms during natural disasters.

   #satellites #NetworkSovereignty

   • Embed this notice
     翠星石 (suiseiseki@freesoftwareextremist.com)'s status on Friday, 07-Mar-2025 12:29:16 JST 翠星石

     in reply to
     @strypey I know the basics about radio transmission and well wireless transmission simply doesn't have
     the bandwidth for acceptable connection speeds if you have more than ~40 wireless clients/km².
     
     Starlink is one example - it's acceptably fast if there are only a handful of stations in each cell, but as soon as many people start
     actually using it, the connection speeds drop substantially.
     
     LTE and 5G use many techniques to increase the possible bandwidth, but as soon as enough people actually start using it in an
     area, the speeds drop substantially, even to the point where emergency callls start getting dropped.
     
     There is a lot more bandwidth available in currently unused extremely high frequency bands and lab testing has shown promising
     results (although that doesn't necessarily mean practical devices will be possible to manufacture), but due to patents, any working
     technology will be unusable until at least 20 years after practical devices are made available.
     
     The atmosphere also tends to aggressively attenuate extremely high frequencies, so I'm not sure if such bands will be practical to
     use from satellites.
     
     If you want emergency announcements during natural disasters, the most reliable thing would be one big AM tower that can transmit an AM radio signal to the whole country (with maybe text announcements also encoded into the audio stream).
     
     In many natural disasters you're really on your own and should focus on being prepared to handle them, as even if you can make a call for help, that's no good if the emergency services are too overwhelmed to help you.
     
     
     Instead the fibre internet rollout should continue, as that has enough bandwidth and is quite tolerant of natural disasters if properly installed.
   • Embed this notice
     翠星石 (suiseiseki@freesoftwareextremist.com)'s status on Friday, 07-Mar-2025 12:32:13 JST 翠星石

     in reply to

     • Roknrol
     @roknrol GPS is transmit-only and only needs to transmit a couple of bits of information a second and current receivers can extract those bits from below the noise floor, which is why it only needs 31 satellites to cover the Earth.
   • Embed this notice
     Roknrol (roknrol@beige.party)'s status on Friday, 07-Mar-2025 12:32:14 JST Roknrol

     in reply to

     @strypey Sorry...boosted for reach. I'm not a rocketry or satellite guy so I can't give you a better answer...just bits and pieces picked up over the years.

     GPS seems to take 31 satellites, according to BBC.

     https://www.youtube.com/watch?v=IVa-NynMFcI

   • Embed this notice
     Strypey (strypey@mastodon.nzoss.nz)'s status on Friday, 07-Mar-2025 12:32:16 JST Strypey

     in reply to

     • Roknrol

     @roknrol
     > This doesn't exactly answer the question

     I can see two answers, which directly contradict each other, and offer no references. Thanks anyway, but answers.com seems more like noise than signal to me.

   • Embed this notice
     Roknrol (roknrol@beige.party)'s status on Friday, 07-Mar-2025 12:32:17 JST Roknrol

     in reply to

     @strypey This doesn't exactly answer the question, but it comes close and should give you what you need to determine details.

     https://www.answers.com/astronomy/What_is_the_minimum_number_of_satellites_needed_to_cover_the_whole_Earth

   • Embed this notice
     LisPi (lispi314@udongein.xyz)'s status on Friday, 07-Mar-2025 12:38:18 JST LisPi

     in reply to

     • 翠星石
     @Suiseiseki @strypey Another thing that would help is resorting to LiFi.
     
     Short-range high-speed wireless communications backed by fiber and with little to no interference issues unlike LTE, 5G and WiFi.
     
     Obviously this requires a hub architecture and doesn't provide direct-to-device connectivity.
     
     > but due to patents, any working technology will be unusable until at least 20 years after practical devices are made available.
     
     We should get rid of those, they're generally harmful.
     翠星石 likes this.
   • Embed this notice
     翠星石 (suiseiseki@freesoftwareextremist.com)'s status on Friday, 07-Mar-2025 12:44:45 JST 翠星石

     in reply to

     • LisPi
     @lispi314 I'm not sure about Li-Fi, as I reckon many manufacturers couldn't help but to crank up the TX power and blind people.
     
     It could be a device-to-device meshnet, with many base stations, provided all the devices are in visual range (with maybe a backup wireless link), but sensitive enough detectors that can pick up for example a stream of single photos are patented up the wazoo and such sort of bandwidth and connection sharing arrangements are prevented as much as possible.
   • Embed this notice
     翠星石 (suiseiseki@freesoftwareextremist.com)'s status on Friday, 07-Mar-2025 13:00:52 JST 翠星石

     in reply to

     • LisPi
     @lispi314 >I think infrared signaling would be a potentially interesting option there.
     Yes, IR would be safest, as you can't see it, it won't be transmitted as a laser beam and it won't damage your eyesight until you hit thermal heating effects.
     
     >Basically all of those patents should be invalidated for prior work
     Yes, most, if not all patents on radio communications are totally invalid due to all the prior art, but the patent offices grant them anyway, even granting 2 or 3 or 4 or 5 patents on the same thing.
     
     Yes, patents should be annulled, as they no longer serve a purpose and are only harmful.
   • Embed this notice
     LisPi (lispi314@udongein.xyz)'s status on Friday, 07-Mar-2025 13:00:54 JST LisPi

     in reply to

     • 翠星石
     @Suiseiseki > I'm not sure about Li-Fi, as I reckon many manufacturers couldn't help but to crank up the TX power and blind people.
     
     It'd clearly need to be adjustable, because risking the triggering of seizures or blinding (whether temporary or not) would be unacceptable.
     
     I think infrared signaling would be a potentially interesting option there. It doesn't need to be particularly powerful for electronics to receive it and very few user-oriented items emit it in a way that couldn't be fairly easily filtered out/tolerated as noise.
     
     > It could be a device-to-device meshnet, with many base stations, provided all the devices are in visual range (with maybe a backup wireless link), but sensitive enough detectors that can pick up for example a stream of single photos are patented up the wazoo and such sort of bandwidth and connection sharing arrangements are prevented as much as possible.
     
     Basically all of those patents should be invalidated for prior work (of which there is quite a bit online), but mostly I think they should all be invalidated because the entire patent office should be invalidated as harmful to scientific and societal progress. (A lot of energy/resource efficiency improvements that we *need* with the climate crisis are locked away by patent rentseeking.)
   • Embed this notice
     翠星石 (suiseiseki@freesoftwareextremist.com)'s status on Monday, 10-Mar-2025 18:17:44 JST 翠星石

     in reply to
     @strypey >Is the limit the number of connections, or the level of packet throughput?
     It's both.
     
     The greater the number of connections, the more connections the available bandwidth needs to be subdivided between and the more interference mitigations need to be applied (i.e. "delays" between "packets" need to be increased).
     
     
     One example of the issue of too many clients is how the range of UMTS base stations would reduce the more clients that were connected (although later encoding schemes don't face this exact same limitation).
     
     CDMA is analogous to a room full of pairs of people speaking different languages (with each pair rejecting languages that they don't understand as noise) - if too many pairs keep entering the room, every pair needs to move closer and closer to be able to hear the other.
   • Embed this notice
     Strypey (strypey@mastodon.nzoss.nz)'s status on Monday, 10-Mar-2025 18:17:46 JST Strypey

     in reply to

     • 翠星石

     (1/2)

     @Suiseiseki
     > as soon as enough people actually start using it in an
     area, the speeds drop substantially

     Is the limit the number of connections, or the level of packet throughput?

     > emergency announcements during natural disasters, the most reliable thing would be one big AM tower

     We have an AM network, although it's in danger of being defunded 🙄 But that's one way broadcast. I'm interested in 2-way emergency communications, ideally without a SPoF.