January 2010doc.: IEEE 802.11-10/0104r2
IEEE P802.11
Wireless LANs
Date: 2010-01-19
Author(s):
Name / Affiliation / Address / Phone / email
Susan R Dickey / Caltrans / California PATH
ITS, UC Berkeley
RFS, Bldg 452
1357 S. 46th St.
Richmond, CA 94804 / (510) 665-3664 /
Wayne Fisher / ARINC, Inc / 2551 Riva Road, Annapolis, MD / 410-266-4958 /
Monday (AM2), January 18, 2010
16 people in attendance (affiliation given in parentheses at first mention of speaker in minutes.)
Lee Armstrong (USDOT), TGp Chairman, called the meeting to order at 10:30 am. Since the TGp Secretary Susan Dickey (Caltrans) was delayed at the airport, Wayne Fisher, TGp Technical Editor, began taking the minutes for this time slot.
Lee gave the opening presentation in: IEEE 802.11-10/0068r0. This week’s agenda is in IEEE 802.11-09/1315r1. Lee reviewed the IEEE 802 and 802.11 Policies and Rules, policies on patents, attendance, voting, the website for documents( and the protocol for our meetings. Lee requested that anyone making a submission should download the official document template from the IEEE 802.11 website(
The goal for this week’s sessions is to complete comment resolution for TGp SB 001 and go to recirculation ballot. The latest Master Spreadsheet for the TGp sponsor ballot is IEEE 802.11-09/1200r8.
Lee asked for a motion to approve the November minutes, IEEE 802.11-09/1256r2. A motion was made by Stuart Kerry (OK Brit), seconded by Justin McNew (Kapsch) and approved by unanimous consent.
Lee presented the Agenda for this session, from 11-10/1315r2. He discussed the strategy for Comment Resolution in Sponsor Ballot.This is different from the previous Letter Ballots; all comment resolutionsmust be tracked in the spreadsheet with complete resolutions without reference to other, supporting documents.Jon Rosdhal (CSR) made a few clarification remarks on SB Comment Resolution. Any reference must be a complete reference source (including URL, etc.).John Kenney (VSC3) asked what about a comment referencing another comment resolution? Jon Rosdahl said that only rejected comments are forwarded to REVCOM and each comment must stand alone. The TGp Task Group has been assigned as the 802.11p Comment Resolution Committee and is responsible for resolving all of the 802.11p SB Comments.Stuart also noted that teleconferences can be used with full powers of comment resolutions and voting.
Lee talked about the teleconference meeting results; teleconference minutes are in documents IEEE 802.11-10/004740, 0016r0 and IEEE 802.11-09/1310r0, 1327r0. CIDs 1001 and 1008 were resolved in the Dec 3 telecon (11-09/1310r9). CIDs 1010, 1015 and 1016 were resolved in the Jan 7 telecon (11-10/0016r0). Lee said we had very few resolutions during the teleconferences but many discussions of proposed resolutions in submissions.He discussed what submissions will be presented and set up the list of documents in the agenda. More submissions will be added to the agenda as they become available.
The agenda was adopted by unanimous consent.
Le e reviewed the draft status: speculative Draft 9.0, including resolution of editorial comments and of the comments resolved during the teleconferences,has been placed in the Members Area. We will vote on approving it at the Monday evening session.
Justin McNew then began a discussion of submission 11-10/0045r2, with the addition of the comments on the PICs that were assigned to Justin at the last teleconference. Changes to proposed comment resolutionsmade during the discussion were added to 11-10/0045r3.
With respect to CIDs 1005 and 1006, Justin suggested three alternative resolutions:
- Add supported rates back into Timing Advertisement Frame
- Make support of all data rates mandatory
- Make minimum rate set mandatory, and provide explanation of how other rates are subject to higher-level negotiations, e.g. as in the WAVE service advertisement.
John Kenney said, from the vehicle safety point of view, safety messages are broadcast at pre-defined rates, and he is OK with living with 3 data rates already mandatory in clause 17.Jon Rosdahl commented that there is a minimum rate set for Control Frames, and there may also be a minimum rate set for Data Frames. After discussion, Justin said he will generally follow alternative 3 above, and complete resolutions to these CIDs once he has figured out where in the document to insert the explanatory text.
For CIDs 1011 and 1012, the proposal of the commenter is to take out a footnote and put the contents in the body of the document.Justin proposed to agree with both these comments, and there was no objection from the group.
With respect to CID 1014, after clarification and discussion with John Kenney, Justin proposed using the phrase “frame classes” instead of “classification” as suggested by the commenter.
With respect to CID 1017, John K mentioned an alternative resolution to delete that section of the PICs that was proposed in 11-09/1242r0. This document is on the agenda for tomorrow morning, but John is not sure if he will be ready to present it by then. It was decided to defer the discussion of this comment to be dealt with when 11-09/1242r0 is considered.
CIDs 1026 and 1068 dealt with the same text. Justin’s resolution is intended to deal with both comments. John K said that he crafted his resolution to make sure it does not apply to a legacy device, and he is not sure that is covered in the document if Justin’s resolution is adopted. Justin added another sentence to the resolution in 11-10/0045r1 to take this into account.
For CID 1034, the same resolution previously adopted in the master spreadsheet for 1008 was copied in, and then checked against the draft to make sure it was correct.
For CID 1037, Justin maintained that PIC is not necessary for this functionality, because it is an informative statement. For CID 1042, John K said that this comment started everyone on the road to submission 11-10/1242r0. John K said that one way to add the PICs we need may be by modifying A4.4.4.2 in the baseline document. Justin said this can be done in reference to resolution of CID 1017, and that in his submission he will go ahead and decline CIDs 1042-1051.
With reference to CID 1058, Justin said there is no technical change in what will be done for the TGp timing capabilities code, but the presentation style is made compatible with TGv for greater clarity, making it easier for TGv to add timing capability code 2.
With respect to 1059, Jon R pointed out that clean up of the OUI is being done in 11mb to take care of what the commenter is complaining about. This is outside the scope of 802.11p, and TGp should not have to make changes anywhere we haven’t already changed things, there is no need for TGp to duplicate TGmb’s work. The commenter should be encouraged to address his concerns to TGmb. If TGp would complete the comment resolution expeditiously, then it can be rolled in by TGmb. Jon thinks that, although it is good to have thorough comment resolutions to satisfy the commenter, we could be more succinct and get through the process faster.
With respect to CID 1064, Justin altered his resolution to take account of John K’s concerns about a “shall” in section 11.1; see wording in 11-10/0045r3.
Wtih respect to CID 1067, there was discussion on the TSF Timer. A system may not need to maintain a TSF Timer during OCB operations.
With respect to CID 1063, 1064, 1065, 1066, 1067, 1068, 1069, resolutions from 11-10/0045r2 were accepted.
For CID 1094, Justin agreed after discussion with John K that TA IE should be Optionalin the Timing Advertisement frame.
With respect to CID 1100, Justin pointed out to Wayne that Table 7.37a is supposed to have the values in columns CWmin and CWmax that are the same in our document as in Table 7.37 in the base document.
Lee recessed the TGp session, since we were out of time for the morning. We resume Monday evening at 7:30 pm.
Monday (PM1), January 18, 2010
10 people in attendance
Justin resumed his review of 11-10/0045r2 at CID 1100. With reference to CID 1118, Justin explained that he has returned the text to UTC instead of UT0, since UT0 was not actually correct. With reference to CID 1125, John K questioned the use of 0 as the indication for unknown time error. Justin changed it in 11-10/0045r3 to be a value of all 1s.
For Richard Roy’s comments, Justin disagrees with all of those remaining starting with 1126 except for 1151. Justin would agree with 1151, since he thinks the additional phrase “is not a member of a BSS” is redundant. John K is worried that making the change in 1151 might create ambiguity. Justin agreed to take all of these comments, including 1151, out of 11-10/0045r3, to be dealt with separately.
Justin uploaded 11-10/0045r3 to the server, as a record with change bars of the discussion, and created a clean copy with deferred CIDs removed as 11-10/0045r4.
Motion: To resolve CIDs 1011, 1012, 1014, 1026, 1034, 1037, 1042, 1043, 1044, 1045, 1046, 1048, 1050, 1051, 1057, 1058, 1059, 1060, 1061, 1063, 1064, 1065, 1066, 1067, 1068, 1069, 1070, 1090, 1094, 1100, 1116, 1117, 1118, 1123, 1124, 1125 according to the resolutions in 11-10/0045r4.
Moved: Justin McNew (Kapsch)
Seconded: George Vlantis (ST Microelectronics)
8 Agree
0 Disagree
0 Abstain
Francois Simon (USDOT) then presented 11-09/1320r2. In a discussion of 1002, George and Lee said that it should be agreed, not declined, since the reference was not there already, and needs to be added to clause 2. Francois said this change must apply to 1122 as well.
The comments in submission 11-09/1320r2 had been discussed previously in teleconferences. We went over each CID; Lee made the above changes and additional changes to 11-09/1320r3, most of them involving changing the language in the resolution column to use the correct agree/counter/disagree” descriptions. Other changes were made to take comments out of 11-09/1320r3 and have John K present resolutions for them in another submission. Comments 1043-1051, which had already been resolved in submission 11-10/0045r4 were also removed from 11-10/1320r2.
We discussed CID 1004. John K said he was considering accepting a comment that suggested adding the acronym OCB in the text, but after discussion he agreed it was not a good idea, and we agreed to “disagree” for CID 1004.
After the changes Lee made, 11-10/1320r3 contained resolutions for the following comments: FS-1, 1002, FS-2, FS-3, 1020, 1122, 1103, 1104, 1036, 1105, 1022, 1023, 1038, 1095, FS-4, 1096, 1097, 1024, FS-5, 1025, 1098, 1106, 1099, 1003, 1101, 1102, 1107, 1108, 1009, 1039, FS-6, 1040, FS-7, 1041, 1062, 1013, 1109, FS-8, 1033, 1092, 1004, 1021, and FS-9. (FS-x refers to additional editorial corrections not addressed elsewhere.)
Motion: to accept the Recommended Resolutions to these comments [as listed at the beginning of 11-09/1320-r3] and the recommended changes to P802.11p D90 as noted [in 11-09/1320-r3] and instruct the editor to make these changes to the latest draft of P802.11p.
Moved by: Francois Simon (USDOT)
Seconded: Wayne Fisher (ARINC)
9: Agree
0: Disagree
0: Abstain.
George Vlantis then presented 11-10/0015r0, making changes into 11-10/0015r1 (first change was to correct title to say Sponsor Ballot instead of Letter Ballot). There was a discussion about 1112, 1113, 1114, with Justin saying that no other group goes beyond what is in the regulations. George said that this is no longer true, and Justin said he would not bring the comments back again. Justin and George also discussed CID 1119 and clause J.2.2. Justin said if we set dot11MultiDomainCapabilityEnabled, for example, to true, then we have to set Country Class in the Timing Advertisement frame, and he thinks this is a big mistake. Justin also said we should not be declaring OCBEnabled to be true in a band that is a standard WiFi band in Europe. Furthermore Justin thinks having any reference to MIB attributes in Annex J is clearly wrong. George agreed to remove 1119, 1120, 1121 and 1156 from this submission.
Motion: To accept resolutions for CIDs 1001, 1018, 1019, 1031, 1032, 1112, 1113, 1114, 1154, 1155 as in 11-10/0015r1.
Moved: George Vlantis (ST Microelectronics)
Second: Justin McNew (Kapsch)
9: Agree
0: Disagree
0: Abstain
George then presented 11-10/0037r2, which describes how to resolve CID 1091 by changing our draft to make no changes to 17.3.8.8, and to adjust the headings and PICs statements to be consistent with that. This is consistent with the changes 11mb is making to remove mention of temperature from 802.11. John K commented that the principle behind this overall change is that the temperature requirement is not part of the standard, but part of a specification given to the supplier.
Motion: To accept the resolution for CID 1091 as in 11-10/0037r2.
Moved: GeorgeVlantis (ST Microelectronics)
Seconded: Wayne Fisher (ARINC)
9: Agree
0: Disagree:
0: Abstain
John K announced that he has been sitting in on TGae and gave a presentation IEEE 802.11-10/0095r0 on vehicle safety communication as a use case for the need to be able to assign lower priority to management frames in some situations.
Lee recessed TGp at 9:30.
Tuesday (AM2), January 19, 2010
10 people in attendance
Lee reconvened the TGp session at 10:30 am. Sue mentioned that she had uploaded IEEE 802.11-10/0115r0, which is a cross-reference of CIDs that have been a subject of motions. Lee went over the agenda. John K said he was ready to present 11-10/0088r3. Carl Kain (Noblis) is working on resolutions of comments in Clause 17. George said he has also been working on resolutions for 1073, 1076 and 1091; he and Carl will work together on this.
John K presented 11-10/0088r3, removing a comment from 11-10/0088r2 dealt with in yesterday’s submissions. He reviewed the comments in that submission; most commentswere reviewed with little additional discussion from the group, which concurred with John K’s recommendations. In reference to 1128, he pointed out that acceptance of this comment does make a minor change in the language of the baseline, but without changing the actual intent, so he recommends to agree with the commenter.
There was some discussion of the suggested addition of “over links” language in comments 1131, 1132 and 1134 on clause 5.2.10. Carl and Francois, as well as John K, disagreed with the commenter that this is an improvement. John K said he thinks the commenter is trying to establish that a BSS is a set of links, not a set of STAs, and that this is not something TGp wants to do. With reference to CID 1133, Carl and John K discussed whether our transmitter/receiver language was correct; the consensus was that we should stay with our original language and disagree with the comment.
There was a discussion with contributions from Lee, John K and Justin about behavior when switching the MIB attribute dot11OCBEnabled, and how this would affect a STA’s possible interactions with a BSS to which it had been joined before setting dot11OCBEnabled. Justin said he would revisit this in a comment resolution he is working on.
Motion: To accept resolutions to 802.11p D9.0 Sponsor Ballot comments 1035, 1093, 1110, 1115, 1128, 1131, 1132, 1133, 1134, and 1135 as shown in 11-10/0088-03.
Moved: John Kenney (VSC3)
Seconded: George Vlantis (ST Microelectronics)
Agree: 7
Disagree: 0
Abstain: 0
Justin then presented IEEE 802.11-10/0114r0, containing comments from clauses 7, 9 and 11 that were deferred from yesterday. With respect to the proposed resolution of CIDs 1005 and 1006, there was some discussion of whether requiring management frames to be sent at the lowest PHY mandatory rate would put an initial burden on implementers. Justin commented that the ACK time is always dependent on the lowest mandatory rate. John asked if we have to say anything, or if we can let what is already in the baseline be the rule? Justin pointed to clause 9.6, and said since we don’t have a BSS Basic Rate Set, we need to specify. George pointed to page 281 in the 2007 baseline document, and said the rule for selecting the data rate in the case of an ACK or a CTS is to pick the highest date rate in the data rate set (BSS Basic Rate Set or mandatory set) that is less than or equal to the rate of the frame you have just received. The group then looked at the changes that were part of 802.11n that are now being rolled in to 11mb, and decided that the paragraphs describing what to do when operating with dot11OCBEnabled true (and thus no BSS Basic Rate Set from beacons) should be put as the second and third paragraphs of 9.6.0a (11n). Language for this was inserted into 11-10/0114r1.
There was further discussion about the “over links” terminology, the apparent intent to eliminate the MIB attribute dot11OCBEnabled, and the suggestions for elaboration of timing distribution and primitives in other comments with proposed resolutions in 11-10/0114r1. The consensus of the group was to concur with Justin’s explanations in his comment resolutions as to why we disagree with these proposed changes and how they would represent an extension of scope for TGp.
Motion: To accept the resolutions to CIDs 1005, 1006, 1126, 1129, 1130, 1136, 1137, 1138, 1139, 1140, 1141, 1142, 1143, 1144, 1145,1146, 1147, 1148, 1149, 1150, 1151, 1152, 1153 as proposed in 11-10/0114r1.
Moved: Justin McNew (Kapsch)
Seconded: John Kenney
Agree: 6
Disagree: 0
Abstain: 0
Lee recessed TGp at 12:25 pm.
Tuesday (PM1), January 19, 2010
13 people in attendance
Lee re-convened the TGp session at 1:35 pm. Lee reviewed the status of the Master Spreadsheet. He is working on including the resolutions from this morning’s timeslots. Green shading indicates agree or agree in principle, white we haven’t looked at it yet, orange disagree. Sue has uploaded 11-10/0115r1, including the CIDs of comment resolutions from this morning’s time slot. Carl uploaded IEEE 802.11-10/0123r0 containing comment resolutions for Clause 17, and presented them to the group.
With respect to CIDs 1071 and 1072, John K suggested re-wording to avoid redundancy and make the meaning more clear, suggesting “If a STA has dot11ACRtyp=2, it shall meet the enhanced receiver performance specifications given in Table 17.13a.” George pointed out that we should only need to say optional once, as to whether the enhanced requirements in Table 17.13a apply. Joe Lauer (Broadcom) also pointed out problems with the wording in the paragraph, and agreed with Francois that we need to specify “no less than,” not just “meet.” After discussion, a wording was agreed on and Carl inserted it in to 11-10/0123r1.
Carl then explained to John K, with reference to 1073 and 1076, that there are two tests, so his comments are from a false premise and the text should remain the same.
Motion: To accept comment resolutions to CIDs 1071, 1072, 1073, 1074, 1075, 1076, 1109, and 1111, as proposed in 11-10/0123r1.
Moved: Carl Kain (Noblis)
Seconded: Justin McNew (Kapsch)
Agree 8
Disagree 0
Abstain 0
George then presented IEEE 802.11-10/0122r0, containing comment resolutions on comments from Appendix J that were deferred yesterday evening. The idea of adding clause J2.2.2 originally came from Peter Ecclesine (CISCO) ; George has discussed these comment resolutions with Justin and with Alastair Malarky (Mark IV). There was discussion of whether we need to require dot11OCBEnabled to be true on all channels in the 5.850 to 5.925 GHz band in the United States, and the consensus was that this requirement should not change, but that the specification of the other MIB elements should not be done by TGp. Rewordings were entered into 11-10/0122r1. George said that J.2.3 should be eliminated;we should not be specifying MIB attributes in an ISM band. This requires renumbering J.2.4 as J.2.3. For the new J.2.3, text is similar to the changes for J.2.2, in specifying that dot11OCBEnabled should be set to true, but not specifying any other MIB attributes. Again in the discussion of CID 1156, the consensus was that the requirement for dot11OCBEnabled to be true should be retained in the bands where we expect 802.11p to be operative.