4 ms·

The problem is its not so easy to share spectrum. In reality you can only have a single transmitter/receiver on a given frequency band in a given area.

CamperBob2 7y ago
The problem is its not so easy to share spectrum. In reality you can only have a single transmitter/receiver on a given frequency band in a given area In reality that hasn't been true for decades, beginning with the initial work on spread-spectrum communications in the WWII timeframe. But the regulatory system has yet to catch up fully with the technology's potential.
- 420codebro 7y ago
  Using DSSS as a crutch in lieu of careful frequency planning is really short-sighted.
- madengr 7y ago
  Spread spectrum still has limitations. The more codes you add just brings up the noise floor for everyone else.
  - CamperBob2 7y ago
    What is this "noise floor" you speak of? https://www.reddit.com/r/amateurradio/comments/49yuwx/til_gps_signal_are_so_weak_they_are_actually/ https://www.reddit.com/r/amateurradio/comments/49yuwx/til_gp... The whole idea behind SS and UWB techniques is that the noise floor is under your control, rather than Boltzmann's.
    - madengr 7y ago
      You would need PRN codes with zero cross-correlation; no such thing. So when you add more users, they appear as noise in your de-spread bandwidth. Add more users, your noise floor goes up, degrading your SNR. There are optimum codes to use, such and Gold (GPS) and Walsh (IS-95 CDMA), but they still have non-zero cross-correlation, so your users are limited. Still no free lunch.
      - CamperBob2 7y ago
        I didn't say it was a free lunch. ~99% of a free lunch, which is what proper spreading codes buy you, is good enough. Point being, the obstacles are regulatory, not technical.
        madengr 7y ago
        They are still technical. The spreading doesn’t buy you anything in terms of channel capacity, just processing gain over narrow band interference, and the downside is now you need precise power control to keep all those codes at the same noise level. 4G & 5G uses OFDM, essentially distributing the data over slow, narrow band channels to get around the multi-path issues with trying to use even wider band CDMA. Then MIMO to actually increase the capacity, which only works when you have multi path. This gives a good summary, specifically page 32, but CDMA is essentially dead soon. http://www.pitt.edu/~dtipper/2720/2720_Slides9.pdf http://www.pitt.edu/~dtipper/2720/2720_Slides9.pdf
        CamperBob2 7y ago
        Dude, your entire post history calls to mind the question, "What am I doing on a site called 'Hacker News'?" Channel capacity is not the problem. Do you own a radio capable of VHF/UHF reception? Turn it on. Tune it to a randomly-chosen frequency between, say, 600 and 1000 MHz. What do you hear? Now, repeat this exercise 5,000 times, and let me know how many of those iterations yield anything but white noise from your receiver's own front end and IF strip. There's your "channel capacity." This problem is purely regulatory. Everything should be frequency-agile at this point in time, but it's not. Never mind CDMA or TDMA, we haven't even picked the lowest of the low-hanging FDMA fruit yet.
dzhiurgis 7y ago
You don't share per-se, you operate network same way many MVNO's do. There's one network (ran by one gov agency) and rest of the players are basically MVNO's, doing customer support, marketing and negotiating peering. Just like traditional ISPs work in many western countries.

CamperBob2 7y ago

The problem is its not so easy to share spectrum. In reality you can only have a single transmitter/receiver on a given frequency band in a given area In reality that hasn't been true for decades, beginning with the initial work on spread-spectrum communications in the WWII timeframe. But the regulatory system has yet to catch up fully with the technology's potential.

420codebro 7y ago
Using DSSS as a crutch in lieu of careful frequency planning is really short-sighted.
madengr 7y ago
Spread spectrum still has limitations. The more codes you add just brings up the noise floor for everyone else.
- CamperBob2 7y ago
  What is this "noise floor" you speak of? https://www.reddit.com/r/amateurradio/comments/49yuwx/til_gps_signal_are_so_weak_they_are_actually/ https://www.reddit.com/r/amateurradio/comments/49yuwx/til_gp... The whole idea behind SS and UWB techniques is that the noise floor is under your control, rather than Boltzmann's.
  - madengr 7y ago
    You would need PRN codes with zero cross-correlation; no such thing. So when you add more users, they appear as noise in your de-spread bandwidth. Add more users, your noise floor goes up, degrading your SNR. There are optimum codes to use, such and Gold (GPS) and Walsh (IS-95 CDMA), but they still have non-zero cross-correlation, so your users are limited. Still no free lunch.
    - CamperBob2 7y ago
      I didn't say it was a free lunch. ~99% of a free lunch, which is what proper spreading codes buy you, is good enough. Point being, the obstacles are regulatory, not technical.
      - madengr 7y ago
        They are still technical. The spreading doesn’t buy you anything in terms of channel capacity, just processing gain over narrow band interference, and the downside is now you need precise power control to keep all those codes at the same noise level. 4G & 5G uses OFDM, essentially distributing the data over slow, narrow band channels to get around the multi-path issues with trying to use even wider band CDMA. Then MIMO to actually increase the capacity, which only works when you have multi path. This gives a good summary, specifically page 32, but CDMA is essentially dead soon. http://www.pitt.edu/~dtipper/2720/2720_Slides9.pdf http://www.pitt.edu/~dtipper/2720/2720_Slides9.pdf
        CamperBob2 7y ago
        Dude, your entire post history calls to mind the question, "What am I doing on a site called 'Hacker News'?" Channel capacity is not the problem. Do you own a radio capable of VHF/UHF reception? Turn it on. Tune it to a randomly-chosen frequency between, say, 600 and 1000 MHz. What do you hear? Now, repeat this exercise 5,000 times, and let me know how many of those iterations yield anything but white noise from your receiver's own front end and IF strip. There's your "channel capacity." This problem is purely regulatory. Everything should be frequency-agile at this point in time, but it's not. Never mind CDMA or TDMA, we haven't even picked the lowest of the low-hanging FDMA fruit yet.

dzhiurgis 7y ago

You don't share per-se, you operate network same way many MVNO's do. There's one network (ran by one gov agency) and rest of the players are basically MVNO's, doing customer support, marketing and negotiating peering. Just like traditional ISPs work in many western countries.