[STDS-802-11-TGM] Estimated Throughput calculation

Thread Links	Date Links
Thread Prev	Thread Next	Thread Index	Date Prev	Date Next	Date Index

Hello Matt,

We are resolving comment 8203 in TGmc. (See doc 11-16/820r1):

8222

3623.40

R.7

"EstimatedThroughputInbound and EstimatedThroughputOutbound for each AC of a current or
potential link to another STA using Equation (R-1)", but Equation (R-1) is just about estimating EstimatedThroughput and there is no indication how the Inbound and Outbount estimates are derived from this

Change "timatedThroughput =" at line 45 to "EstimatedThroughputInbound = EstimatedThroughputOutbound ="

GEN

Discussion:

3624.33 clarifies that this equation is calculating outbound (downlink) throughput only “is the estimated portion of airtime that is available for outbound transmissions” (note “inbound” and “outbound” are meaningless unless it is specified from the perspective of which device. I believe the intended perspective is from the network or AP.)

1076.48 muddies the waters: “The Estimated Air Time Fraction subfield is 8 bits in length and contains an unsigned integer that represents the predicted percentage of time, linearly scaled with 255 representing 100%, that a new STA joining the BSS will be allocated for PPDUs carrying Data of the corresponding AC for that STA.”

This is ambiguous. Does “will be allocated” mean if the STA makes a TSPEC request for admission control or HCCA? I think not, because these optional standardised features are not much deployed. I think, instead, it relates to non-standardized internal queuing process for MSDUs in the downlink.

I don’t believe the “inbound” (uplink) rates can be inferred from the “outbound” (downlink) rates. They depend on the STA’s success in gaining channel access, which is a complex function of instantaneous aggregate load.

The line at 1904.06 indicates that R.7 is for calculating outbound throughput. There is no such statement for inbound throughput.

Proposed Resolution:

Revised.

The equation R-1 is usable to calculate only outbound (traffic from the AP) throughput, this is because it depends on an estimated airtime fraction indicated by the AP that will be applied to the outbound traffic.

The following changes remove any indication that R-1 may be used to calculate inbound throughput.

At 3623.39 change: “can determine values for EstimatedThroughputInbound and EstimatedThroughputOutbound for each AC”

to “can determine a value for EstimatedThroughputOutbound for each AC”

At 3623.45 change “timatedThroughput” to “EstimatedThroughputOutbound”

At 3623.50 change “EtimatedThroughput” to “EstimatedThroughputOutbound”

Proposed Resolution:

Rejected. It is clear from 3623.45 that both of these values are calculated “Using Equation (R-1)”.

Status: Discussed, but no consensus for making a change or rejection.

In my mind, equation R-1 is a means of calculating throughput given a bunch of “global parameters”, one of which is

Estimated airtime fraction. The latter is derived from the information in the ESP element

(“EST_AIRTIME_FRACTION is the estimated portion of airtime that is available for outbound transmissions for this link as indicated in the ESP element received” 3624.33).

The question is how to calculate the inbound rate.

The assumption behind the comment is that we can assume that the inbound rate is equal to the outbound rate – but

I see no reason why this should be the case, and it would defeat having separate inbound and outbound parameters

on all the SAP interfaces.

If we know the inbound airtime fraction, R-1 can tell us how to calculate throughput. So how do we determine

the inbound airtime? The first proposed change left this value undefined, while clarifying that it was not calculatable

by R-1. The second proposed resolution hopes the problem goes away.

Can you provide an alternative resolution?

Best Regards,

Adrian P STEPHENS

Tel: +44 (1793) 404825 (office)
Tel: +1 (971) 330 6025 (mobile)

----------------------------------------------
Intel Corporation (UK) Limited
Registered No. 1134945 (England)
Registered Office: Pipers Way, Swindon SN3 1RJ
VAT No: 860 2173 47

IF YOU WISH to be Removed from this reflector, PLEASE DO NOT send your request to this CLOSED reflector. We use this valuable tool to communicate on the issues at hand.