Is this the real case or a problem of my simulation errors in laa wifi coexistence evaluation?

[John Young]

 I ran laa-wifi-indoor  with ftpLambda 4, duration of 100s, wifi cca threshold default to -62 dBm, and ranging laa eD from -62dBm to -82 dBm. Drs and ReservationSignal were disabled. The num of all antenna are configured to be 1 to obtain a SISO configuration.  RngRun was configured to be ranged in 1 to 5. And the result was averaged.  

The results and some indices are in the following.  Laa's latency increased as it took more time to wait. However, wifi's latency increased as well. It seemed laa gained better snr that led to less tx time but wifi suffered from laa's backoff. I am not sure what led to such a result.  And my results seemed to contradict some proposals in 3GPP meeting.  Can someone give me some ideas or suggestions?

delay(ms)

A -62.000000:

70.2706

A -72.000000:

74.2961

A -82.000000:

77.7293

B -62.000000:

116.852

B -72.000000:

118.663

B -82.000000:

120.357

tx-time(ms)

-62.000000:

 enB  7097.61  AP  5500.76  sta  33.2552

-72.000000:

 enB  6975.58  AP  5594.87  sta  33.3811

-82.000000:

 enB  6889.76  AP  5745.39  sta  33.5776

#retries

Enb -62.000000:

271.4

Enb -72.000000:

264.6

Enb -82.000000:

262.3

wifi

-62.000000 AP  11860.5   STA  0.24

-72.000000 AP  12411.6   STA  0.2

-82.000000 AP  12699   STA  0.25

#backoff

-62.000000:

 enB  840.5  AP  1228.15  sta  2.15

-72.000000:

 enB  838.15  AP  1267.8  sta  2.11

-82.000000:

 enB  837.2  AP  1290.7  sta  2.17

[Tom Henderson]

On 12/03/2016 09:02 PM, John Young wrote:
> I ran laa-wifi-indoor with ftpLambda 4, duration of 100s, wifi cca
> threshold default to -62 dBm, and ranging laa eD from -62dBm to -82 dBm.
> Drs and ReservationSignal were disabled. The num of all antenna are
> configured to be 1 to obtain a SISO configuration. RngRun was
> configured to be ranged in 1 to 5. And the result was averaged.
>
> The results and some indices are in the following. Laa's latency
> increased as it took more time to wait. However, wifi's latency
> increased as well. It seemed laa gained better snr that led to less tx
> time but wifi suffered from laa's backoff. I am not sure what led to
> such a result. And my results seemed to contradict some proposals in
> 3GPP meeting. Can someone give me some ideas or suggestions?

You are averaging a lot of pairwise communications, and probably
obscuring some details of what is going on by doing so. I would suggest
that you separately track the latency of each AP -> STA combination as
you vary eD and, for those that seem to increase as a result of the eD
change, take a closer look at why that pair either finds the channel
busier or undergoes more retransmissions. Also, I suggest to check -67
and -77 values to see if they follow the trend.

- Tom

[John Young]

Thanks, I will follow your advice.

在 2016年12月6日星期二 UTC+8下午2:50:05，Tom Henderson写道：

[Vinicius Dantas]

How did it go? Any other results?