Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror
×
Cellphones Wireless Networking Networking

'5G Just Got Weird' (ieee.org) 132

SuperKendall (Slashdot reader #25,149) shared this review of the recent 5G standards codified by the 3rd Generation Partnership Project (3GPP) in Release 16 (finalized on July 3).

"5G just got weird," writes IEEE Spectrum: 4G and other earlier generations of cellular focused on just that: cellular. But when 3GPP members started gathering to hammer out what 5G could be, there was interest in developing a wireless system that could do more than connect phones... One of the flashiest things in Release 16 is V2X, short for "Vehicle to Everything." In other words, using 5G for cars to communicate with each other and everything else around them... The 3GPP standards bring those benchmarks into the realm of gigabytes per second, 99.999 percent reliability, and just a few milliseconds.

Matthew Webb, a 3GPP delegate for Huawei and the other rapporteur for the 3GPP item on V2X, adds that Release 16 also introduces a new technique called sidelinking. Sidelinks will allow 5G-connected vehicles to communicate directly with one another, rather than going through a cell-tower intermediary... Tseng says that sidelinking started as a component of the V2X work, but it can theoretically apply to any two devices that might need to communicate directly rather than go through a base station first. Factory robots are one example, or large-scale Internet of Things installations.

Some other "weird" highlights of the new 5G standards:
  • "5G incorporates millimeter waves, which are higher frequency radio waves (30 to 300 GHz) that don't travel nearly as far as traditional cell signals. Millimeter waves means it will be possible to build a network just for an office building, factory, or stadium. At those scales, 5G could function essentially like Wi-Fi networks."
  • "In past generations of cellular, three cell towers were required to triangulate where a phone was by measuring the round-trip distance of a signal from each tower. But 5G networks will be able to use the round-trip time from a single tower to locate a device."
  • "Release 17 includes a work item on extended reality — the catch-all term for alternate reality and virtual reality technologies."

This discussion has been archived. No new comments can be posted.

'5G Just Got Weird'

Comments Filter:
  • So they were right all along!
    • Yup. If you read the annex to the appendix to the apocrypha of 3GPP Release 16i you'll see all sort of weird shit about modulating radio signals to turn skin cells into Covid19. 16i is a bit of a tricky read though, you have to rotate it 90 degrees to be able to see it.
      • Just as long as Bill Gates and George Soros can work out how to put autism in the vaccine, everything will be back to normal.

      • Make sure you rotate it the right way, though. The other way is a strange treatise on how all the gummy bears in the universe are actually sentient creatures.

  • I give it three years from when this becomes common in automobiles and a worm is developed that spreads across the 5G interconnected cars that does something that no one wants. Will be pure luck if no one gets killed.

    More radiation too.
    /tinfoil and out.
  • But these cars talking to each other. I don't think that's safe.They could be conspiring against us useless carbon units. Really, what do they need us for?

  • "In past generations of cellular, three cell towers were required to triangulate where a phone was by measuring the round-trip distance of a signal from each tower. But 5G networks will be able to use the round-trip time from a single tower to locate a device."

    No, three cell towers are not required to triangulate... only two.

    You can't locate a device in a two-dimensional space with any accuracy with only one tower unless that tower has discrete antennas with VERY VERY narrow degree of coverage. For example if you had 360 antennas one degree apart, you can get pretty good if the signal is close, but if you have 6 antennas (like 1G-4G) then you can narrow it down to 60 degrees... which is a huge swath of area, the further you are.

    Further, with the "low latency" ca

    • I just signed into my account to post the very same thing. I'm not clear how this is possible in a typical installation. If you're using some type of dynamic beam forming such that you know what direction the antenna is aimed,, then I could see using the round-trip time to figure out the distance and thus calculate the location, but that assumes you're not utilizing a reflection instead. (Or, worse, more than one.) That's more likely the higher you go, and you can't really do that type of thing at lower

      • Re:Triangulation (Score:5, Interesting)

        by squidinkcalligraphy ( 558677 ) on Saturday August 08, 2020 @08:54PM (#60381339)

        I used to work on location tech back in the 90s, and one project in our group was getting location of a sat-phone handset (I think this was relating to the newly launched Iridium network). A handset might be only visible to one satellite, but obtaining location was still possible, due to the satellite moving (this method doesn't work with geosynchronous sat-phones). Over a period of time, the satellite moves w.r.t the handset, so with at least two round-trip-time readings, and if you know the location of the satellite, you can get the handset location. Unfortunately that doesn't work for a moving handset in the same way, so this single-tower-location must be using something else (as said, probably beamforming to work out direction + RTT for distance).

        WFIW, triangulation is literally taking three angles...

        • Triangulation is taking three angles

          No, it's using a triangle to determine location. You only need to take two angles if you know the length of a side.

    • 5G uses MIMO phased array antennas. Direction is known by the transmitter/receiver by looking at the phase difference between the multiple antennas. Not sure how it would work with wavelengths shorter than the antenna spacing but I'm sure it would use a similar technique as laser rangefinders.
      Once you know the direction the round trip time of flight is all you need for decent accuracy location.

    • No, three cell towers are not required to triangulate... only two.

      Three towers is required for a unique location, if all you know is distance (with only two distance measurements you get two possible locations because the distance maps out a circle around each cell tower, and circles intersect at either 2 or 0 points on a plane). 5G uses some beamforming and MIMO techniques to measure both distance and direction, so it only needs a single tower. With a phased array, you don't need individual antennas to have high angular resolution, you can use interferometry to make prec

    • "You can't locate a device in a two-dimensional space with any accuracy with only one tower unless"

      With the unless being your definition.

      BLuetooth already does this with a single antenna in a box the size of a wifi access point. Sure, that antenna is not a single radiator. But so what?

  • Huh (Score:3, Funny)

    by Mixmaster Bri ( 819575 ) on Saturday August 08, 2020 @08:43PM (#60381309)
    How does slashdot reader #25,148 feel about this?
  • So is it safe to say these "sidelinks" would be more like UDP than TCP? Just broadcast packets spewed out there to be forwarded?

    • That's always been how WiFi works already. Only the end nodes calculate checksums. It would be WAY too computationally heavy to have each node do a handshake, not to mention how much it would slow the net down due to the increase in aggregate propagation delayfor each packet.
    • So is it safe to say these "sidelinks" would be more like UDP than TCP? Just broadcast packets spewed out there to be forwarded?

      You can rest assured, the security issues will not be sorted out until long after (if ever) the time that the physical and MAC layers are defined.

  • V2V (Score:5, Interesting)

    by sinij ( 911942 ) on Saturday August 08, 2020 @09:02PM (#60381351)
    I spoke at a conference to someone developing a standard to vehicle to vehicle communication protocol. I really didn't like their answers about authentication and privacy. First, since you always broadcast your position and speed, law enforcement only need to have a passive receiver to write speeding tickets. Basically, the car is snitches on you as you drive. Second, latest design does not attempt authentication with proof, you are assigned static ID (think MAC) and everyone else assumes you are accurately reporting it.

    What could possibly go wrong with such setup, right?
    • Re: (Score:2, Flamebait)

      by coastwalker ( 307620 )

      It is curious to relate but the paranoia and FUD about technology extensions does not exist in much of the world outside the USA. Does this mean that flying cars and next generation tech will be developed elsewhere? Interesting.

    • You mean the cops will catch you speeding with their passive receivers. I mean my car won't look like yours but they won't know that.

      On the Inter^H^H^H^H^H freeway, no one knows you're a dog.

    • Sounds like building a jamming device that spams hundreds of random speeds would be fun. But I assume that signed keys or something similar would be used for a protocol like this.

    • by AmiMoJo ( 196126 )

      Good news is that it will probably never take off. The rate at which lidar and vision systems are getting cheaper and smaller means that by the time they figure out how to make a compatible V2V system it will already be redundant.

    • they can probably already get your speeds and locations from data thats already recorded in your car and sent to the manufacturer, more and more cars have SIMs in them already. I'd also wait until more information comes out about the standard being proposed as it is more than likely that it will change fundamentally since you heard about it. Then again, don't break the law and you'll be fine.
    • Re:V2V (Score:4, Insightful)

      by thegarbz ( 1787294 ) on Sunday August 09, 2020 @06:39AM (#60382125)

      law enforcement only need to have a passive receiver to write speeding tickets.

      Perfect. Finally. If you want to drive on a race track then drive on a damn race track. Your ability to get away with breaking the law and putting eveyone else at risk should not depend on the attentiveness of police.

      • by Lordfly ( 590616 )

        Or, as was seen with automatic traffic cameras, cities will artificially lower speed limits to catch impatient people, or people trying to get to work on time, or whatever. 25 mph for everywhere at all times, or else you will fund the police department with your fines.

      • Sure, so long as it's "impossible" to hack a transmitter to spoof the speed of every other nearby car...
    • Then again, already years ago I spoke to a guy from NXP who worked on vehicle to vehicle communication, and according to him their main challenge was trust. When you receive data from another car, how do you know it's trustworthy? How do you secure the hardware and the communication protocols? That's what they were mainly working on (apart from designing the chips, of course). So not everyone is naive.
    • law enforcement only need to have a passive receiver to write speeding tickets.

      You mean, like a speed camera?

    • by tlhIngan ( 30335 )

      I spoke at a conference to someone developing a standard to vehicle to vehicle communication protocol. I really didn't like their answers about authentication and privacy. First, since you always broadcast your position and speed, law enforcement only need to have a passive receiver to write speeding tickets. Basically, the car is snitches on you as you drive. Second, latest design does not attempt authentication with proof, you are assigned static ID (think MAC) and everyone else assumes you are accurately

  • One of the flashiest things in Release 16 is V2X, short for "Vehicle to Everything." In other words, using 5G for cars to communicate with each other and everything else around them..

    Now my car can fake get/spread COVID-19 from 5G too [vox.com].

  • by Rick Schumann ( 4662797 ) on Saturday August 08, 2020 @09:31PM (#60381385) Journal
    I don't want any vehicle I own wirelessly connected to anything, ever, I want it completely isolated. That's the only way I can be reasonably sure it's not getting hacked by someone remotely.
    Laptops have had 'RF kill' switches for a long time now. Cars should have the same thing: a switch that kills all wireless connectivity, one that can't be overridden.
    It should be up to the driver whether their vehicle has the ability to connect wirelessly to anything.
    • What will you do when all cars are required to be connected to a network to use public roads?

      • LOL that's not going to happen, nice try though.
    • Its already too late for that: https://www.businessinsider.co... [businessinsider.com]

      • *snip*
        *solder center conductor to shield*
        So much for GPS.
      • by dfghjk ( 711126 )

        The article said absolutely nothing about cars being networked. From the article:

        "A Ford spokesperson later told Business Insider that in general, GPS units in Ford cars are not routinely pinging out their whereabouts as customers drive around. Rather, Ford cars have several on-board services such as "Sync Services Directions" (a navigation device that works with drivers' phones) and 911 Assist, which users have to switch on and opt into. And employers can use a service called "Crew Chief" to monitor their

  • 6G is right behind 5G

    Who really needs all this data transfer speed?

    If people are to be tracking chipped (plenty of mounting evidence of this intent) .... go figure. https://www.youtube.com/watch?... [youtube.com] & https://patentscope.wipo.int/s... [wipo.int] (tracking your body movements?)

    China already has a social credit status restricting what people can do and go to. They can even arrest you if you have your phone off to long.

    No everyone has a cell phone, so what is the obvious solution?

    Now they are talking of the possible

    • Don't tell me how much bandwidth I need.

      • by Khyber ( 864651 )

        Yea, 6G promises multi-terabit bandwidth. Does your mobile device, or even your regular PC, have that kind of bandwidth, even on the PCI-E lanes, or even for the system as a whole?

  • "two devices that might need to communicate directly rather than go through a base station first."

    Just like Wifi adhoc mode. What happened to it. I recall old Windows Wifi drivers had the ability for Wifi adhoc connections between two clients, without no base station involvement. I don't see references to it anymore.

    If two 802.11*-equipped cars -- that have never 'seen' each other -- drive down a highway, can they automatically setup a Wifi adhoc network and exchange data?

    • >If two 802.11*-equipped cars -- that have never 'seen' each other -- drive down a highway, can they automatically setup a Wifi adhoc network and exchange data?

      Yes. If they both have a higher layer protocol that permits such a thing. The 802.11 protocol will support it. If you had some funky crypto like some bizarre massive scale IBE (read that at is ain't going to happen), you could arrange for them to have keys ready before they talk. With a trusted third party (like *gulp* the government), you could u

      • Ah thanks -- d you know of any higher level protocols in use in the wild?

        I wasn't even aware of IBE (identity based encryption). Thanks for the pointer.

        Cameras and licence plate recognition mean an IBE scheme can use the license plate as the address. Or DMVs could work just like a CA -- renewing a car's communication certificates every year as part of vehicle registration.

        If all a vehicle does is broadcast safety messages "Plate MMH 443 at position X: accelerating on vector V at rate C " -- transponder st

        • There's an 802 spec for making announcements over 802 (802.21). This is primarily conceived so that you could (a) listen to an 802.11 AP and know what to do next (I.E. bother to connect, expect to find internet, need to use a bloody interstitial) and (b) Know if it is fit for handing over from a heterogenous network type like a cellphone connection. I should know, I was one of the founders of that spec.

          A secondary concept is that you have an ordering problem in secure communications between previously unpai

    • The Freifunk movement in Germany (and probably other places) uses it to mesh routers together by running special routing protocols over them. Usually they also open up a managed network for the clients to connect to.

      • Thanks for the pointer - just googled them -- it's great to know of the work they're doing. By managed network, did you mean the 'VPN to Sweden' mentioned in Wikipedia, or just a generic managed network that enables communication between peers?

        I wonder what the association delays would be for a new peer to mesh with a Freifunk router. So if multiple cars, each with a Freifunk router onboard, were driving up down the highway, how fast would the peers associate/disassociate?

        • Well since they use adhoc mode they don't disassociate or associate.
          The managed mode is only used for connecting clients. the "VPN to Sweden" is just one way to do Internet uplinks. Such an uplink makes the network immediately useful for many people.

    • Google. At least on Android. Android 4 included a re-written wifi driver with no support for adhoc. Even though "Issue #82" was open for years, and was one of their most upvoted bugs.

      Cisco. They stack every vote at the IEEE they have an interested in. And they are only interested in selling infrastructure. The initial adhoc spec is terrible, it's a thrown together afterthought. Building a stable adhoc mesh network essentially requires you to break it. 802.11s was kinda supposed to be a replacement. The spe

  • by Cylix ( 55374 )

    So the cellular companies have been sitting on bandwidth intended for vehicle coordination. They were about to lose it, but now we have promised rolled into 5g.

    I have a feeling vehicular interconnection alerts still won’t manifest.

    The idea was for things like crash detection, road congestion and all manner of p2p would be facilitated. Much like rural broadband funding this has all been made under promises that will never be delivered upon.

  • by Casandro ( 751346 ) on Sunday August 09, 2020 @03:15AM (#60381887)

    ... when they did ridiculous things like changing every signaling element to a new, but nearly identical one running over HTTP/2.
    On the other hand, all the improvements on the radio side could easily be included into LTE.

    Perhaps in the future we will see "5G" just like "3G", a standard with very little useful improvement that was mostly meant to push through patented technologies (like CDMA). Just like the "3G" standards its gap might be closed in rather quickly from both sides. On 2G we got EDGE, while 4G finally brought usable Internet.

  • Because it would be a shame if it wasn't and autonomous vehicles believed there was and invisible bus in front of them, slowing them to a crawl, or believed a diversion was telling them to go the wrong way up a one way street. Endless possibilities for griefing and worse.
    • Autonomous vehicles (AVs) are going to be frustrating enough for most people when they actually follow the rules of the road such as driving the speed limit! :) Can you imagine once AVs start getting to be a fair percentage of vehicles on the road and you have the rest of the drivers whom are used to going a couple of kph (or mph depending on where you are) above the limit stuck behind a line of AVs that are sticking exactly to the limit? Especially someone whom is late for work? Or someone in in AV and wou

  • Wow... and here I was thinking UX was a stretch.

  • This isn't a 5G vs 4G thing. Even 4G started massively focusing beyond the scope of the phone. Hell One-to-many simulcasting was introduced in 3GPP Release 13 5 years ago. And that was built on changes such as QoS and prioritization introduced in Release 9 back in 2008.

    Basically from the moment 4G was ratified in the late 90s the 3GPP was looking to expand it beyond phones and cellular data.

  • "5G incorporates millimeter waves, which are higher frequency radio waves (30 to 300 GHz) that don't travel nearly as far as traditional cell signals. Millimeter waves means it will be possible to build a network just for an office building, factory, or stadium. At those scales, 5G could function essentially like Wi-Fi networks."

    Meaning, it will even work worse through walls?

  • that's cute and all but what would be nice is if I could get some basic fucking phone service in the middle of the god damned city, looking specificly at you ATT how come your map shows all blue but nowhere I go do I get more than 1 bar of service and the only place I get 4G is at my house, where I don't fuckin need it

  • Here's the problem with this: it looks great on paper and works fine in the ivory tower of Silicon Valley or some other major city but once you get out into rural areas, this stuff ain't gonna work. And in a lot of places, rural American might only be 20 miles outside a major city.

  • by ElitistWhiner ( 79961 ) on Sunday August 09, 2020 @12:58PM (#60382765) Journal

    Peer2Peer, aka walkie-talkie, just became a viable plus more than Voice-cases. Mesh becomes ubiquitous. TacTek private frequency use-case for Fire, Police and importantly Emergency where any 5G device can be enabled. No more dedicated handheld hardware thru a gatekeeper server. Could be democratizing tech or Hyper surveillance. Big decade ahead.

    Peer2Starlink for battlefield mapping, localized ID and geo-fencing scenarios.

"A car is just a big purse on wheels." -- Johanna Reynolds

Working...