3 ms·

The rule of thumb I use is that the utilisation of the spectrum is 5%. This is based on a conference that I went to (20 years ago granted) where an Australian

by femto 6y ago

The rule of thumb I use is that the utilisation of the spectrum is 5%. This is based on a conference that I went to (20 years ago granted) where an Australian group had actually measured the utilisation of the spectrum. This was in the context of "whitespace" radio, where they were investigating the opportunistic use of licensed but unused spectrum. Their answer was that from a technical perspective it's worth doing as 95% of the spectrum is available for opportunistic use, the main barriers being non-technical. Personally I can't see a problem with a "free for all". As space-time (MIMO) coding comes to dominate, WiFi access points will be increasingly focusing their transmissions on specific spatial regions leading to less interference on average, even if the devices are sharing a frequency. It's in the transmitter's self-interest to maximise channel capacity by focusing as much power as possible on the receiver of interest and not wasting power by spraying it around in an omni-directional manner. This self-interest will avoid a "tragedy of the commons", provided the rules say that everyone has the same maximum power, forcing people to use their radiated power wisely.

gruez 6y ago
>It's in the transmitter's self-interest to maximise channel capacity by focusing as much power as possible on the receiver of interest and not wasting power by spraying it around in an omni-directional manner. This self-interest will avoid a "tragedy of the commons", provided the rules say that everyone has the same maximum power, forcing people to use their radiated power wisely. That breaks down when you factor in that beamforming devices are expensive, so not all devices will opt to use it. The devices that don't use them will still blaring at full power, causing interference to others.
- femto 6y ago
  But if they are limited the same radiated power (an important caveat placed on the "free for all", in the form of a class license) the MIMO system will be able to get more power focused on its receiver, and within that spatial region overpower the omnidirectional interferer. The omnidirectional transmitter will lose more by not being MIMO than the MIMO system will lose from the interference. MIMO isn't expensive these days with the availability of integrated chip sets, and the cost is only dropping. It will get to the point where apart from broadcasters no one runs a non-MIMO system because the cost of not being MIMO (loss of channel capacity) will exceed the cost of the MIMO hardware.
  - gruez 6y ago
    >The omnidirectional transmitter will lose more by not being MIMO than the MIMO system will lose from the interference. The problem is that you're not competing against another laptop/router that wants to maximize throughput, you're competing against a IoT device (eg. smart fridge, surveillance camera, lowend tablet, etc.) that doesn't give a shit about throughput and will happily tolerate a 50Mb/s (or slower) connection, while taking up the same timeslice/radiative power as every other device.
    - femto 6y ago
      Granted that there will be some impact on the MIMO from the Iot device, but it will be lopsided as the MIMO can null out the IoT transmission. It can be looked at in a couple of ways: the IoT is having to spread its power more thinly than the MIMO, so the IoT tx will have less influence on the MIMO receiver than the MIMO tx will, or the MIMO's spatial coding gives a degree of isolation which benefits both the IoT and MIMO devices. There will be a problem if there are a large number of close by IoT devices up against the MIMO, as their smallish contributions will sum up to something big. In that case the MIMO would need to improve its isolation from the IoT device by increasing its number of antennas or the quality of its processing. As a side issue (doesn't really negate your argument), excessive transmission power is a cost for battery powered devices, so at least some of this IoT devices will be motivated to use their transmitted power more wisely. To raise a new front in the argument :-), there is a case to be made that the ability of a device to cause interference isn't actually dominated by the power of its transmission, but by the information content of its transmission. Low information content means the signal is easier to predict and a smart enough receiver can predict the signal and cancel it out, leaving only the desired signal.

by femto 6y ago

gruez 6y ago

>It's in the transmitter's self-interest to maximise channel capacity by focusing as much power as possible on the receiver of interest and not wasting power by spraying it around in an omni-directional manner. This self-interest will avoid a "tragedy of the commons", provided the rules say that everyone has the same maximum power, forcing people to use their radiated power wisely. That breaks down when you factor in that beamforming devices are expensive, so not all devices will opt to use it. The devices that don't use them will still blaring at full power, causing interference to others.

femto 6y ago
But if they are limited the same radiated power (an important caveat placed on the "free for all", in the form of a class license) the MIMO system will be able to get more power focused on its receiver, and within that spatial region overpower the omnidirectional interferer. The omnidirectional transmitter will lose more by not being MIMO than the MIMO system will lose from the interference. MIMO isn't expensive these days with the availability of integrated chip sets, and the cost is only dropping. It will get to the point where apart from broadcasters no one runs a non-MIMO system because the cost of not being MIMO (loss of channel capacity) will exceed the cost of the MIMO hardware.
- gruez 6y ago
  >The omnidirectional transmitter will lose more by not being MIMO than the MIMO system will lose from the interference. The problem is that you're not competing against another laptop/router that wants to maximize throughput, you're competing against a IoT device (eg. smart fridge, surveillance camera, lowend tablet, etc.) that doesn't give a shit about throughput and will happily tolerate a 50Mb/s (or slower) connection, while taking up the same timeslice/radiative power as every other device.
  - femto 6y ago
    Granted that there will be some impact on the MIMO from the Iot device, but it will be lopsided as the MIMO can null out the IoT transmission. It can be looked at in a couple of ways: the IoT is having to spread its power more thinly than the MIMO, so the IoT tx will have less influence on the MIMO receiver than the MIMO tx will, or the MIMO's spatial coding gives a degree of isolation which benefits both the IoT and MIMO devices. There will be a problem if there are a large number of close by IoT devices up against the MIMO, as their smallish contributions will sum up to something big. In that case the MIMO would need to improve its isolation from the IoT device by increasing its number of antennas or the quality of its processing. As a side issue (doesn't really negate your argument), excessive transmission power is a cost for battery powered devices, so at least some of this IoT devices will be motivated to use their transmitted power more wisely. To raise a new front in the argument :-), there is a case to be made that the ability of a device to cause interference isn't actually dominated by the power of its transmission, but by the information content of its transmission. Low information content means the signal is easier to predict and a smart enough receiver can predict the signal and cancel it out, leaving only the desired signal.