> Given the neighboring aviation band (from 4.2 to 4.4 GHz) allocated exclusively for vital aeronautical radionavigation equipment
They are afraid of (a) heavy use in the 3.7 - 3.98 GHz band (fundamental emissions) and (b) "accidental" broadcasts in 4.2 to 4.4 GHz range (spurious emissions). The study [1] on p. 12 in §6.1.1 basically assumes the 5G equipment will blast an interfering signal on altimeter band at full power. I think any frequency allocation request can be shut down this way.
And I don't think 5G network will work properly if even a fraction of the assumptions in that report are true. If you look at https://en.wikipedia.org/wiki/5G_NR_frequency_bands, you will see that bandwidths are in 10s of MHz and the sub-carrier spacing [2] is as low as 15kHz. If 5G radios will be off by 15kHz (as opposed to 202000 kHz from 3.98 to 4.2 GHz as the aviation industry is implying), 5G network transmissions will not work properly.
>If 5G radios will be off by 15kHz (as opposed to 202000 kHz from 3.98 to 4.2 GHz as the aviation industry is implying), 5G network transmissions will not work properly.
But that's not really how RFI tends to work - you can be perfectly aligned to your target frequency and generating the signals you expect there, but still be generating mixing or LO spurs, image response, third-order intermodulation products and so on which can be at totally different frequencies.
> mixing or LO spurs, image response, third-order intermodulation products
Do I understand it correctly that it will only be a fraction of the output power and equation (6-1) in https://www.rtca.org/wp-content/uploads/2020/10/SC-239-5G-In... must have a coefficient so that P_source would become P_source*k_spurious where k_spurious < 1 or even k_spurious << 1?
Edit: by not working properly I did not mean that the signal on the target frequency will have severe interference but that just 15kHz apart there could be another target frequency that must be free of interference or otherwise multiple operators will not be able to use the 5G spectrum (12f, ie 180kHz seems to be an accepted guard band as per [1] and FCC seems to have mandated 23MHz guard band [2] which is 127x the norm) or even a single operator will not be able to use OFDM effectively. Am I still wrong in my understanding?
Edit 2: with the interfence evaluation from [1], almost all evaluated guard bands result in -20dBm interference which would make proper k_spurious=0.01. My understanding remains that the aviation industry report used overzealous assumptions and overestimated the interference emission power by at least 100x.
[1] E. Memisoglu, A. B. Kihero, E. Basar and H. Arslan, "Guard Band Reduction for 5G and Beyond Multiple Numerologies," in IEEE Communications Letters, vol. 24, no. 3, pp. 644-647, March 2020, doi: 10.1109/LCOMM.2019.2963311.
As far as I can see, based on a very quick skim reading, the study is essentially divided into a consideration of impacts from fundamental and spurious emissions from 5G transmissions.
For the fundamental emissions, the study considers primarily the impact of receiver front-end overload on the operational margins for the radar altimeters. The referenced ITU document suggests that radar altimeters are specified with "modest selectivity", e.g. 24dB/octave below 4.2GHz up to a maximum of 40dB. I believe that a power spectral density envelope derived from the ITU specifications is the basis of the fundamental emission calculation, not assuming the entire power will be dissipated spuriously.
This is nowhere near my area of expertise, but it seems like those selectivity assumptions (I have no basis for opinion about them) in conjunction with large gains from beamforming are what creates the perception of risk.
> Given the neighboring aviation band (from 4.2 to 4.4 GHz) allocated exclusively for vital aeronautical radionavigation equipment
They are afraid of (a) heavy use in the 3.7 - 3.98 GHz band (fundamental emissions) and (b) "accidental" broadcasts in 4.2 to 4.4 GHz range (spurious emissions). The study [1] on p. 12 in §6.1.1 basically assumes the 5G equipment will blast an interfering signal on altimeter band at full power. I think any frequency allocation request can be shut down this way.
And I don't think 5G network will work properly if even a fraction of the assumptions in that report are true. If you look at https://en.wikipedia.org/wiki/5G_NR_frequency_bands, you will see that bandwidths are in 10s of MHz and the sub-carrier spacing [2] is as low as 15kHz. If 5G radios will be off by 15kHz (as opposed to 202000 kHz from 3.98 to 4.2 GHz as the aviation industry is implying), 5G network transmissions will not work properly.
[1] https://www.rtca.org/wp-content/uploads/2020/10/SC-239-5G-In...
[2] https://info-nrlte.com/tag/subcarrier-spacings/
reply