Ouput mask in test harness and calculated are not matching

[alexcpn]

Hi, After going through WINNF-TS-1014.pdf (https://winnf.memberclicks.net/assets/work_products/Specifications/WINNF-TS-1014.pdf) and following all the values recommended for Noise level etc and all the values parsed out from the ULS database, there is still a big discrepancy in the calculated mask for Bandwidth 160 Mhz, Frequency 6025, (5945.0, 6105.0)).

I have put everything in this google sheets, with sheet titled FreeSpace 15 that has all the values and other sheets that take the values from the ULS data snapshot found here: https://github.com/Wi-FiTestSuite/6GHz-AFC/blob/17656cc709795a704bbbd7eeda5bfbb67b2560ef/ULS_database/FCC/weekly/l_micro_211024_modified_PA.zip

https://docs.google.com/spreadsheets/d/16Tmn1HcEwS3lRDUA7TrkoUus0Mi17ZpD9ts3dXawXtY/edit?usp=sharing

I am trying for a few days to try to correlate but I am not sure what I am missing. It will be good if there is some working for at least one full request and response in the test vector on how it is calculated.

Request in question https://github.com/Wireless-Innovation-Forum/6-GHz-AFC/blob/main/src/harness/inquiries/AFCS.FSP.1.json Mask - https://github.com/Wireless-Innovation-Forum/6-GHz-AFC/blob/main/src/harness/masks/AFCS.FSP.1_mask.json#L572-L595

Also in this mask, I found incumbent frequency for both channel 15 and channel 145 in the 160 MHz band. Hence I am not able to make out the wide varying eirp from 8.7 dBm for channel 15 to -18 dbm for channel 143

mask value | channel number  | chn no and freq   -- | -- | -- { |   |   nominalValue: 8.7, | 15 | {134: [(15, 6025, (5945.0, 6105.0)) upperBound: 10.7 |   |   }, |   |   { |   |   nominalValue: 19.3, | 47 | {134: [(47, 6185,(6105, 6265)) upperBound: 21.3 |   |   }, |   |   { |   |   nominalValue: 29.2, | 79 | {134: [(79, 6345, (6265, 6425)) upperBound: 31.2 |   |   }, |   |   { |   |   nominalValue: -18.0, | 143 | {134: [(142, 6665, (6585, 6745)) upperBound: -16.0 |   |  

[AEgbert]

Also in this mask, I found incumbent frequency for both channel 15 and channel 145 in the 160 MHz band. Hence I am not able to make out the wide varying eirp from 8.7 dBm for channel 15 to -18 dbm for channel 143

This difference is largely due to the different PSD limits in those channels (looking at the frequency-based limits). I posted more details on this subject just now in response to #39, which you may find helpful. At the moment, I don't have specific details on why the PSD limit differences are there, but I'm looking into it. I suspect part of the difference is because the incumbent bandwidth size differs within those two channels.

there is still a big discrepancy in the calculated mask for Bandwidth 160 Mhz, Frequency 6025

Looking at your linked spreadsheet, I suspect part of the discrepancy you've noted may be related to your choice of noise floor. Note that the -110.0 dBm/MHz noise floor from R2-AIP-02 needs to be scaled to the frequency in question:

N (dBm): the actual FS receiver noise level as per R2-AIP-02 scaled to the frequency range I/NRx, Est and I/NVRx-j, Est calculations are performed. [Definition of N within R2-AIP-16]

I don't think this fully addresses the difference between your "Freespace 15" spreadsheet tab and the values in the mask, but I'm looking for input from some of the people who were more involved with the mask generation.

[alexcpn]

Thanks for the clarification related to scaling

using NF_scaled = NF_nominal + 10*log10(f_scaled/f_nominal) and assuming that f_nominal= 1 Mhz, the NoiseFigure has increased 110.0 dBm/MHz + 10 * log10(6025 / 1) = -72.2 dBm/MHz and Transmit Power has also increased to -29.2 dBm; But still far from the 10.7 dBm of the expected upper bound.

Can co-channel exclusion explain this discrepancy? I am however not able to figure out how to calculate that out

[AEgbert]

Here's an update I've received so far:

There are two incumbent microwave paths that terminate within the location uncertainty footprint specified in FSP-1. This situation requires special care to compute the potential interference, as the receiver gain towards the AFC device can vary greatly within this footprint. How the receiver antenna gain is determined for each evaluation point in the uncertainty area may be the source of much of the discrepancy in results.

I'll provide another comment when I've received more information. I've also passed along your suggestion to provide an example derivation for one of the test vectors, which I hope to discuss with our group soon.

[alexcpn]

@AEgbert any update regarding this? I tried to calculate for one more (FS3) based on ITM. This time I calculated also the Effective gain from Boresight and Azimuthal differences between the Receiver, Transmitter and AccessPoint https://docs.google.com/spreadsheets/d/16Tmn1HcEwS3lRDUA7TrkoUus0Mi17ZpD9ts3dXawXtY/edit?usp=sharing (third tab, FS3-ITM). I also tried to calculate the Adjacent channel impression mask, but not really sure if I applied that correctly.

I chose FS3 as I could get a clean Receiver and Transmitter with Antenna data in the common data

If we could get one calculation fully done, then it will help in clarifying what we are missing

[AEgbert]

@alexcpn Apologies for the delay. I've been working on an example derivation for FSP-30 as time permits. I hope to eventually be able to provide that example, but do not have an estimated time on when it will be completed.

Looking at your results for FSP-3, I believe your calculation is done with an incumbent device that is not the actual limiting device for the channel you're evaluating. Working backwards from the PSD data in the FSP-3 mask, I would expect the incumbent that is limiting the power of (GOC 134, Channel 143) is a 10 MHz BW receiver operating at or near 6645 MHz.

For the bandwidth used by the incumbent you've identified, the PSD response reflects a limit of:

"frequencyRange": {
  "lowFrequency": 6660,
  "highFrequency": 6670
},
"maxPsd": {
  "nominalValue": 21.6,
  "upperBound": 23.0
}

Applying a 21.6 dBm/MHz limit to a 160 MHz channel would result in a 43.6 dBm EIRP, which is closer in line with what your calculation provides. (In practice, a result like this would be capped at the FCC EIRP limit.)

In summary, I think the discrepancy in your FSP-3 results is largely from the incumbent that was used to calculate your channel power limit, and that a more restrictive incumbent constraint exists on that channel.

[alexcpn]

@AEgbert thanks; I will recheck FSP-3 again and update and try calculating for FSP-30 . Before that - how can I give a comment on the document (6 GHz Committee Functional Specification WG Functional Requirements for the U.S. 6 GHz Band WINNF-TS-1014-V1.3.0) There is a small error in the R2-AIP-23 sub bulleting

Note - It is very difficult to calculate R2-AIP-17 - the near-field adjustments. In FSP-3 and in most cases the receiver is not so close to the transmitter AP ( $d_{limit}$) that this calculation is needed.

But if it is needed the calculation via linear interpolation by looking up two Excel sheet data seems pretty difficult.

[alexcpn]

from the PSD data in the FSP-3 mask, I would expect the incumbent that is limiting the power of (GOC 134, Channel 143) is a 10 MHz BW receiver operating at or near 6645 MHz.

These are the two MW incumbent receivers* from ULS data I have got; One at 13 km (WQYB266) and one (WPVY281) at 1332 km away. Ignoring the latter as it is too far away (is there any range beyond which an incumbent can be ignored, say > X km)

We have WQYB266 which is operating at 6665 MHz. This is what is updated already in the FS3 ITM tab in the google sheet

Note that for calculating the Response for GC 134, CFI 143 we do not use this operating frequency (6665) anywhere in the calculation R2-AIP-16. We just use the MW Receiver's location data and the Elevation data (for ITM loss) towards the AP location. (At least for

This is what I tried to do in the FS3 ITM tab using itm-logic library to calculate ITM loss)

---

[INFO] For coordinates (33.18055555555555, -97.56055555555555), distance 13.538712332900133
[INFO] Data exists for 6585.0 and 6745.0
[INFO] 2444898, WPVY281, 1, 6695.00000000, 6695.00000000, 7, 1
[INFO] 2444898, WPVY281, 2, 6585.00000000, 6585.00000000, 7, 1
[INFO] 2444898, WPVY281, 10, 6645.00000000, 6645.00000000, 7, 1

Too far

[INFO] For coordinates (33.177749999999996, -97.54677777777778), distance 1332.7376518443828
[INFO] Data exists for 6585.0 and 6745.0
[INFO] 3836940, WQYB266, 1, 6665.00000000, 6665.00000000, 2, 1
[INFO] 3808671, WQXP873, 2, 6635.00000000, 6635.00000000, 3, 1

*To check which is the Receiver, cross-check with the Antenna table

INFO] uid,antenna_number,call_sign,(transmit)location_number,height_to_center_raat,gain,diversity_gain,path_number,antenna_type_code,antenna_make,antenna_model,btob_rx_dish_gain,            btb_tx_dish_gain,reflector_height,reflector_width,tilt,azimuth
[INFO] 3836940, 1, WQYB266, 2, 24.4, 39.3, None, 1, R, Commscope, VHLP6-6WA, None, None, None, None, 0.0, 0.0

Since you are working on FSP-30 (much thanks to that), I will also try FSP-30 so that it will be easier for you to reply. Thanks again.

[AEgbert]

Based on my (unverified) parsing of ULS data, I'm seeing a TX/RX path with Unique System Identifier 2444898, path number 10, TX Call sign of WPVY281 (location number 1, antenna number 10), RX Call sign of WRCB370 (location number 7, antenna number 1), and emission record of 10M0D7W at 6645 MHz. The separation distance between this ULS RX lat/lon (33.180556, -97.560556) and the nominal SPD lat/lon (33.180553, -97.560701) looks to be about 14 meters (not kilometers). I believe the third line the first code block in your previous comment matches this path.

I expect (but am not certain) that this RX location is the limiting incumbent for channel (GOC 134, CFI 143).

[AEgbert]

Also, I'll pass your correction to TS-1014 on to the group that maintains that document--thanks for the feedback.

[alexcpn]

Based on my (unverified) parsing of ULS data, I'm seeing a TX/RX path with Unique System Identifier 2444898, path number 10, TX Call sign of WPVY281 (location number 1, antenna number 10), RX Call sign of WRCB370 (location number 7, antenna number 1), and emission record of 10M0D7W at 6645 MHz. The separation distance between this ULS RX lat/lon (33.180556, -97.560556) and the nominal SPD lat/lon (33.180553, -97.560701) looks to be about 14 meters (not kilometers). I believe the third line the first code block in your previous comment matches this path.

I expect (but am not certain) that this RX location is the limiting incumbent for channel (GOC 134, CFI 143).

You are right; I calculated in meters, outside the function I printed it as km and calculated wrongly. Thanks!

For GC 134, CFI 143, now these are the two receivers,

Now will check both itm and freespace from the AP to both these receivers as per the distance criteria and take the most restrictive value to protect this band.

My guess is freespace path loss contribution from the nearest will be most restrictive as you mentioned.

With the freespace path loss it is coming closer; I need to re-caclulate now the effective gain for this receiver and check. Just adding this as it is taking time

--------------------------------------------------------------------------------
uuid: 3836940
call_sign: WQYB266   
frequency_assigned: 6665.00000000
receiver_loc: (33.177749999999996, -97.54677777777778)
receiver_to_ap_dist_m: 1332.7376518443828
accesspoint_loc: (33.180553, -97.560701)
accesspoint_ht_m: 3
ap_freq: (6585.0, 6745.0)
ap_bw: 160
r_height_to_center_raat: 24.4
r_gain: 39.3
r_diversity_gain: None
r_azimuth: 0.0
r_antenna_make: Commscope
r_antenna_model: VHLP6-6WA
r_loc_num: 2
r_path_num: 1
r_ant_num: 1
r_diameter_ft: 6.0
r_diameter_m: 1.83
transmitter_loc: (33.06447222222222, -97.71686111111111)
t_height_to_center_raat: 30.5
t_gain: 39.3
t_diversity_gain: None
t_azimuth: 51.6
t_antenna_make: Commscope
t_antenna_model: VHLP6-6WA
t_loc_num: 1
t_path_num: 1
t_ant_num: 1
--------------------------------------------------------------------------------
uuid: 2444898
call_sign: WPVY281   
frequency_assigned: 6645.00000000
receiver_loc: (33.18055555555555, -97.56055555555555)
receiver_to_ap_dist_m: 13.538712332900133
accesspoint_loc: (33.180553, -97.560701)
accesspoint_ht_m: 3
ap_freq: (6585.0, 6745.0)
ap_bw: 160
r_height_to_center_raat: 15.2
r_gain: 38.7
r_diversity_gain: 35.5
r_azimuth: None
r_antenna_make: COMMSCOPE
r_antenna_model: PAR6-59WA RF
r_loc_num: 7
r_path_num: 10
r_ant_num: 1
r_diameter_ft: 6.0
r_diameter_m: 1.83
transmitter_loc: (32.98216666666667, -97.78955555555555)
t_height_to_center_raat: 24.4
t_gain: 38.7
t_diversity_gain: None
t_azimuth: 44.1
t_antenna_make: COMMSCOPE
t_antenna_model: PAR6-59WA RF
t_loc_num: 1
t_path_num: 10
t_ant_num: 10

Note my code ignored WRCB370