1. Additional Test Cases for MAC calibration
May 2013
Sept 2014
Additional Test Cases for MAC calibration
Date:
Authors:
Name
Affiliation
Address
Phone
Gwen Barriac
Qualcomm
Qingjiang Tian
Simone Merlin
George Cherian
Osama Aboul-Magd (Huawei Technologies)

2. September 2014
Context
So far we have agreed upon several simple MAC calibration tests
See document
These tests try to calibrate basic MAC featrues
Overhead, deferral, backoff, etc.
Our ultimate goal is to be able to calibrate MAC functionality in the full simulation scnearios described in document
In order to move towards that goal, we just suggest some additional calibration tests that test MAC interactions when there are multiple STAs per AP
Qualcomm

3. Simulation Parameters
September 2014
Simulation Parameters
Parameter
SUGGESTED VALUES
CCA energy threshold
-62 in 20 MHz
Preamble Detection threshold
-82 in 20Mhz
Primary channels
All same
Traffic
Full-buffer
Aggregation
A-MPDU max aggregation size =2
No A-MSDU
immediate BA
MSDU Size
1500 Bytes
TXOP
4ms
Max number of retries 10
Rate adaptation
Fixed MCS; MCS=0
EDCA parametrs
Default params for best effort
Parameter
SUGGESTED VALUES
TX Power
AP: dBm/Antenna
STA: 15 dBm/antenna
Noise figures
5dB
# of TX/RX antenna assumed
1x1
Preamble detect threshold
-82dBm in 20 MHz
GI:
[long]
Data Premble:
[11ac] BW
20 Mhz
SINR to PER curves
AWGN
Note: We suggest using the worst SINR per MPDU to calculate the PER for that MPDU
Qualcomm

4. Test 5a: One BSS with multiple STAs
September 2014
Test 5a: One BSS with multiple STAs
Purpose:
Testing deferral on both energy detection and preamble detection within a single BSS
1. STA1,2 to STA1,4 & AP1 are within ED range of each other;
2. STA1,1, STA1,5 are outside of ED range but inside PD range of each other
APT6
APT7
APT8
APT9
APT10
APT1
APT2
APT3
APT4
APT5
Residential Scenario Floor 1
Qualcomm

5. Test 5a cont. One BSS with multiple STAs
September 2014
Test 5a cont. One BSS with multiple STAs
Assumptions:
One common pathloss file used by all companies
One common PER curve used by all companies
RTS/CTS: Off
Output: MAC throughput cdf:
Run the simulation at least 20 times with different random seeds, each run should simulate at least 4 seconds
Compare 5%-tile,50%-tile,90%-tile, mean throughput of all STAs
Qualcomm

6. September 2014
Test 5b
Same as Test 5a, but with RTS/CTS ON.
Qualcomm

7. Test 6a: Two BSSs with multiple STAs
September 2014
Test 6a: Two BSSs with multiple STAs
Purpose:
Testing deferral on both energy detection and Preamble detection when there are 2 BSSs
Qualcomm

8. Residential Scenario Floor 1
September 2014
APT6
APT7
APT8
APT9
APT10
APT1
APT2
APT3
APT4
APT5
Residential Scenario Floor 1
STAi,2 to STAi,4 & APi are within ED range of each other;
STAi,1, STAi,5 are outside of ED but inside PD range of each other
STA1,5 and STA2,1 are within ED range of each other
Qualcomm

9. September 2014
Test 6a: Assumptions
Assumptions:
One common pathloss file used by all companies
One common PER curve used by all companies
RTS/CTS: off
Deferral to OBSS packets:
Nodes defer to any received packet received from an OBSS
Option 1: Nodes may choose to stop decoding packet if it is from OBSS.
Option 2: Nodes decode any packet they lock on to regardless of whether it’s from an OBSS or not.
Qualcomm

10. Test 6a: Output Output: MAC throughput c.d.f.
September 2014
Test 6a: Output
Output: MAC throughput c.d.f.
Run the simulation at least 20 times with different random seeds
5%-tile,50%-tile,90%-tile, mean throughput of all STAs
Qualcomm

11. September 2014
Test 6b:
Same as Test 6a except that RTS is ON
Qualcomm

12. September 2014
Straw poll
Do you support to add the calibration tests in document 1220r0 to the Simulation Scenario document 980r2?
Qualcomm

13. September 2014
Appendix
Qualcomm

14. Node Location Matrix September 2014 Node Index Coordinate X(meter)
Coordinate Y(meter)
AP1 5
AP2 15
STA1,1
0.5
0.2
STA1,2 4
STA1,3 6
STA1,4
STA1,5
9.5
9.8
STA2,1
10.5
10.2
STA2,2 14
STA2,3 16
STA2,4
STA2,5
19.5
19.8
Reference Coordinates (0,0) is at the
left bottom corner of apartment 1
Qualcomm

15. Pathloss Model TGax model No Antenna gain No shadowing
September 2014
Pathloss Model
TGax model
No Antenna gain
No shadowing
PL= *log10(fc/2.4e9)+20*log10(min(d,5))+(d>5)*35*log10(d/5) +5*W
d is the distance between two nodes
If two nodes are in different apartments, 2 Walls are assumed
Qualcomm

16. Path Loss Matrix for Test5
September 2014
Path Loss Matrix for Test5
AP1
STA1,1
STA1,2
STA1,3
STA1,4
STA1,5
0.00
64.58
46.43
63.02
66.24
62.89
75.11
49.44
66.16
52.45
Qualcomm

17. Path Loss Matrix For Test6
September 2014
Path Loss Matrix For Test6
AP1
AP2
STA1,1
STA1,2
STA1,3
STA1,4
STA1,5
STA2,1
STA2,2
STA2,3
STA2,4
STA2,5
0.00
86.21
64.58
46.43
76.71
85.45
86.97
92.01
63.02
66.24
62.89
75.11
91.50
92.54
91.49
96.21
49.44
66.16
78.15
87.66
86.28
92.53
52.45
75.34
84.61
84.70
78.09
87.72
57.07
75.25
Qualcomm