3 ms·

How could the 200MHz spacing between the frequency allocations be insufficient? I can't imagine the oscillators in either system are specified for deviations as

How could the 200MHz spacing between the frequency allocations be insufficient? I can't imagine the oscillators in either system are specified for deviations as high as +-100MHz over their lifetime.

xxpor 5y ago
You're right: they're perfectly fine. This whole thing is FUD from someone... just not sure who yet. My understanding is this band has been used for Europe for a few years already.
- p_l 5y ago
  With effectively 400 MHz guard band, though.
- onphonenow 5y ago
  Yeah, and all this drama vs actually you know, testing things in the 10 years they've had is rediculous by the FAA. Seriously, just test it. You have thousands of flights a day. On a day with visual flight rules put a transmitter at 4.0ghz, max power, worst case radiation pattern and land a 737, 787, A320 and A321 etc as a test. There is something going on here that we don't know about.
jdiez17 5y ago
I was also curious about this, so I looked at the Airworthiness Directive[1] referenced in the example NOTAM. Basically, the argument is that the radio altimeter receivers "must detect faint signals reflected off the ground to measure altitude" and [C-band 5G signals] "could significantly degrade RA functions [...] if the altimeter is unable to sufficiently reject those signals". So it seems it is basically a question of how wide is the RF input filter bandwidth of RAs on commercial aircraft. [1] https://rgl.faa.gov/Regulatory_and_Guidance_Library/rgad.nsf/0/625b2b44ac41e39a862587a600628feb/$FILE/2021-23-12.pdf https://rgl.faa.gov/Regulatory_and_Guidance_Library/rgad.nsf...
- p_l 5y ago
  I'm not good with analog circuits, but someone mentioned in previous thread that designing a 200 MHz wide filter that would have "brick wall" response to frequencies outside it is non-trivial, and technically the 200MHz is just the transmit range, not receive range...
  - aDfbrtVt 5y ago
    That's why virtually every RF frontend oversamples sufficiently to do the sharp filtering in DSP.
    - pgorczak 5y ago
      Although when receiving a weak signal with strong out of band interference, you may still need a lot of dynamic range to successfully filter it out digitally.
      - VLM 5y ago
        3rd order IMD performance can become an issue. 3rd ord IMD for non-EE could be described as low power mixers like mixing low power signals and usually the low power signal comes from your intentionally designed radio oscillator, but, whack it with enough incoming RF power and it will go non-linear as its designed to do and start mixing with that incoming signal, mostly making massive levels of interference. You can always improve 3rd order IMD by simply running the oscillator and mixer at higher power levels (which all turns to heat), usually not the best strategy for battery powered gear.
  - guerby 5y ago
    This is all ridiculous, there's still a 200 MHz band guard between the FAA band and the 5G proposed band. Here is what a $1 ESP wifi dongle has to follow: https://en.wikipedia.org/wiki/IEEE_802.11 https://en.wikipedia.org/wiki/IEEE_802.11 "The mask requires the signal to be attenuated a minimum of 20 dB from its peak amplitude at ±11 MHz from the center frequency" So 2 dB/MHz filter. I let you do the math. FAA is just ridiculous here if they let old junk radio hardware handle safety landings for airplanes, but well after 737 max what do you expect...
upofadown 5y ago
That's a 200MHz spacing at a 4400MHz working frequency. For some sort of physical filter this stuff tends to work in terms of percentages. That is only something like 5%. So pretty close. Normally you have lots of space to waste at higher frequencies for guard bands and you can allocate compatible uses for next door bands. Still, when you get a powerful transmitter right next door there are often problems. The previous tenants of that spectrum (C band satellite downlinks) used to have a problem with interference from radar altimeters. So I suppose there is some sort of karmic process here where the service that used to cause interference is now the service that is getting hit with interference.
VLM 5y ago
Phase noise is obviously the problem. Plenty of microwave broadband gear isn't all that narrowband to begin with... Note there's a pretty big difference between "it should be good enough for 99.99% of planes" and "we can guarantee there won't be even one crash even under unlikely conditions". Aerospace is already VERY comfy doing ADS-B at 1090 and DME/TACAN from 960-1215 and the old GPS L1 allocation at 1575.42 and the military L2 allocation at 1227.60 (admittedly not many civies using that LOL) and someday probably L4 and remember ASR-11 airport radars blast at 25 KW power over 2700-2900 MHz. But, you know, the straw that broke the camels back and all that. Most things will be fine. A small amount of things will operate degraded, which is the point of the NOTAM. Some things (like very few) will all fail the same time 5G is deployed, which will be a headache. Going forward every installation of every device on the ground or in the plane will be tested against 5G and it'll be just fine. Most likely for a week or a month everyone will have to treat their radar altimeter as potentially broken until proven otherwise, then it'll be fine and they'll drop the notam. Some equipment will indeed fail and connectors will be tightened or even devices replaced, and it'll be fine.
- aDfbrtVt 5y ago
  I’m not sure I follow how phase noise is related to an interfering channel emitting meaningful radiation 100MHz out of band. If all of these transmitters are compliant then it’s a non-issue. I’ve also never seen a transmitter where phase noise was more of a factor than IF in the context to out of band emissions. Could you elaborate on how phase noise would cause an issue in this scenario? N.B. Most of my experience is in optical systems, my wireless knowledge is limited.

How could the 200MHz spacing between the frequency allocations be insufficient? I can't imagine the oscillators in either system are specified for deviations as high as +-100MHz over their lifetime.

xxpor 5y ago

You're right: they're perfectly fine. This whole thing is FUD from someone... just not sure who yet. My understanding is this band has been used for Europe for a few years already.

p_l 5y ago
With effectively 400 MHz guard band, though.
onphonenow 5y ago
Yeah, and all this drama vs actually you know, testing things in the 10 years they've had is rediculous by the FAA. Seriously, just test it. You have thousands of flights a day. On a day with visual flight rules put a transmitter at 4.0ghz, max power, worst case radiation pattern and land a 737, 787, A320 and A321 etc as a test. There is something going on here that we don't know about.

jdiez17 5y ago

I was also curious about this, so I looked at the Airworthiness Directive[1] referenced in the example NOTAM. Basically, the argument is that the radio altimeter receivers "must detect faint signals reflected off the ground to measure altitude" and [C-band 5G signals] "could significantly degrade RA functions [...] if the altimeter is unable to sufficiently reject those signals". So it seems it is basically a question of how wide is the RF input filter bandwidth of RAs on commercial aircraft. [1] https://rgl.faa.gov/Regulatory_and_Guidance_Library/rgad.nsf/0/625b2b44ac41e39a862587a600628feb/$FILE/2021-23-12.pdf https://rgl.faa.gov/Regulatory_and_Guidance_Library/rgad.nsf...

p_l 5y ago
I'm not good with analog circuits, but someone mentioned in previous thread that designing a 200 MHz wide filter that would have "brick wall" response to frequencies outside it is non-trivial, and technically the 200MHz is just the transmit range, not receive range...
- aDfbrtVt 5y ago
  That's why virtually every RF frontend oversamples sufficiently to do the sharp filtering in DSP.
  - pgorczak 5y ago
    Although when receiving a weak signal with strong out of band interference, you may still need a lot of dynamic range to successfully filter it out digitally.
    - VLM 5y ago
      3rd order IMD performance can become an issue. 3rd ord IMD for non-EE could be described as low power mixers like mixing low power signals and usually the low power signal comes from your intentionally designed radio oscillator, but, whack it with enough incoming RF power and it will go non-linear as its designed to do and start mixing with that incoming signal, mostly making massive levels of interference. You can always improve 3rd order IMD by simply running the oscillator and mixer at higher power levels (which all turns to heat), usually not the best strategy for battery powered gear.
- guerby 5y ago
  This is all ridiculous, there's still a 200 MHz band guard between the FAA band and the 5G proposed band. Here is what a $1 ESP wifi dongle has to follow: https://en.wikipedia.org/wiki/IEEE_802.11 https://en.wikipedia.org/wiki/IEEE_802.11 "The mask requires the signal to be attenuated a minimum of 20 dB from its peak amplitude at ±11 MHz from the center frequency" So 2 dB/MHz filter. I let you do the math. FAA is just ridiculous here if they let old junk radio hardware handle safety landings for airplanes, but well after 737 max what do you expect...

upofadown 5y ago

That's a 200MHz spacing at a 4400MHz working frequency. For some sort of physical filter this stuff tends to work in terms of percentages. That is only something like 5%. So pretty close. Normally you have lots of space to waste at higher frequencies for guard bands and you can allocate compatible uses for next door bands. Still, when you get a powerful transmitter right next door there are often problems. The previous tenants of that spectrum (C band satellite downlinks) used to have a problem with interference from radar altimeters. So I suppose there is some sort of karmic process here where the service that used to cause interference is now the service that is getting hit with interference.

VLM 5y ago

Phase noise is obviously the problem. Plenty of microwave broadband gear isn't all that narrowband to begin with... Note there's a pretty big difference between "it should be good enough for 99.99% of planes" and "we can guarantee there won't be even one crash even under unlikely conditions". Aerospace is already VERY comfy doing ADS-B at 1090 and DME/TACAN from 960-1215 and the old GPS L1 allocation at 1575.42 and the military L2 allocation at 1227.60 (admittedly not many civies using that LOL) and someday probably L4 and remember ASR-11 airport radars blast at 25 KW power over 2700-2900 MHz. But, you know, the straw that broke the camels back and all that. Most things will be fine. A small amount of things will operate degraded, which is the point of the NOTAM. Some things (like very few) will all fail the same time 5G is deployed, which will be a headache. Going forward every installation of every device on the ground or in the plane will be tested against 5G and it'll be just fine. Most likely for a week or a month everyone will have to treat their radar altimeter as potentially broken until proven otherwise, then it'll be fine and they'll drop the notam. Some equipment will indeed fail and connectors will be tightened or even devices replaced, and it'll be fine.

aDfbrtVt 5y ago
I’m not sure I follow how phase noise is related to an interfering channel emitting meaningful radiation 100MHz out of band. If all of these transmitters are compliant then it’s a non-issue. I’ve also never seen a transmitter where phase noise was more of a factor than IF in the context to out of band emissions. Could you elaborate on how phase noise would cause an issue in this scenario? N.B. Most of my experience is in optical systems, my wireless knowledge is limited.