1. D1.0 PHY Comments Discussion
Month Year
doc.: IEEE yy/xxxxr0
D1.0 PHY Comments Discussion
Date:
Authors:
Slide 1
Slide 1
John Doe, Some Company

2. Introduction (1)
In the TGaf LB1 based on the D1.0 specification there were numerous comments recommending PHY changes because of the following reasons:
Make use of the ac specification in the af instead of the n specification
802.11ac provides feature advantages over n
Different down-clocking of signals in order to reduce the bandwidths results in 3 different preamble/symbol lengths: 20/40MHz, 10MHz, and 5 MHz
This complicates receiver design and may create coexistence problems
Slide 2

3. Introduction (2)
When compared to the 2.4GHz WLAN implementations (11b/g/n), range extension modes are not included in the specification
802.11a/g/n include DSSS rates that extend range
802.11af range advantage over 2.4GHz systems is diminished
5/10/20/40 MHz non-overlapping structure of the channels may create coexistence problems
The TVWS channels are 6 MHz wide while the proposed 11af bandwidths are multiples of 5 MHz
For example two 5 MHz channels and a single 10 MHz channel are not overlapping
Channel alignment or use of bandwidths that are multiples of 6 MHz may be investigated (6/12/.. MHz instead of 5/10/.. MHz bandwidths)
Slide 3

4. 802.11af Based on the 802.11ac Specification
The ac specification has the following mechanisms that provide advantage over the n specification
Unified beamforming protocol
Multi-User MIMO protocol
Multichannel Protocol (80MHz + 80 MHz)
Suitable for spotty TVWS channel availability in order to provide higher throughputs
802.11ac is in the internal LB TG review of Draft 0.1
Complication that the amendment is not finalized
However, the PHY and MAC features are mostly complete
Slide 4

5. FFT Sizes (1)
For the outdoor applications it is desirable that the delay spread of 3.2 (short GI) us (normal GI) is tolerated for outdoor environments [1]
This results in a tone spacing of approximately 39 KHz
In a 5 MHz channel this would result in a 128 point FFT
Need to vary the number of FFT tones per BW in order to keep the preamble length the same for all bandwidths
The resulting FFT sizes for 2.5/5/10/20 MHz bandwidths would be:
64/128/256/512
1/8 down-clocked ac
Slide 5

6. FFT Sizes (2)
FFT sizes for 5/10/20/40 MHz bandwidths and half the delay spread tolerance (1.6 us – 3.2 us) would be:
64/128/256/512
1/4 down-clocked ac
Tone spacing of approximately 78 KHz
Carrier frequency estimation robustness needs to be investigated for narrower tone spacings
Slide 6

7. Signal Bandwidths
FCC emission requirements are very strict resulting in a mask 30+ dB stricter than the mask (adjacent channel, while alternate adjacent requirements are even stricter)
To meet the FCC requirements it should be investigated which bandwidth can actually be realized in a 6 MHz TVWS channel
4 MHz?
5 MHz?
6 MHz?
other ?
Wider bandwidths may be multiples of the above bandwidths
Different countries have different channel bandwidth regulations
Slide 7

8. Range Extension
Simple approach would be to include DSSS rates with ¼ down-clocking:
Lower PAPR requirement
Lower SINR requirement
Another approach that can be explored is to make OFDM rates more robust
To be investigated
Slide 8

9. References
[1] /0251r0 “Outdoor Channels Models for TGah”

10. Straw Poll #1
Do you support using ac specification as a basis for the af specification development?
Yes:
No:
Abs:
Slide 10

11. Straw Poll #2
Do you support using 128/256/512 FFT sizes for 5/10/20 MHz bandwidths?
Actual bandwidths may vary slightly (see slide 8 for examples, TBD)
Yes:
No:
Abs:
Slide 11

12. Straw Poll #3
Do you support using 64/128/256/512 FFT sizes for 5/10/20/40 MHz bandwidths?
Actual bandwidths may vary slightly (see slide 8 for examples, TBD)
Yes:
No:
Abs:
Slide 12

13. Straw Poll #4
Do you support including range extension modes in the af specification (some examples on slide 7, TBD)?
Yes:
No:
Abs:
Slide 13

14. Straw Poll #5
Do you support 1/8 down-clocked version of the ac PHY?
Yes:
No:
Abs:
Slide 14