16 July 2012 Ad-Hocdoc:IEEE 802.11-12/0901r0

IEEE P802.11
Wireless LANs

802.11 ISD SGAd-hoc Meeting Minutes – 16 July 2012
Date: 2012-07-16
Author(s):
Name / Affiliation / Address / Phone / email
Dwight Smith / Motorola Mobility / Grapevine, TX, USA / 817-714-2127 /

Chair: Stephen McCann (RIM)

Acting secretary: Dwight Smith (Motorola Mobility)

Monday, 16July 2012, 09:00 to 11:00 – Madeleine CD

Call to order and agenda

Meeting called to order on Monday, July 16th2012 by Stephen McCann at 9:00 am.

This is an ad hoc meeting as the plenary has not yet occurred. As such we can not take any votes or make decisions. He also reminded participants to record their attendance. This session counts as extra credit for the July 802.11 meeting.

The chair reviewed the meeting guidelines including the IEEE patent policy.

The chair then reviewed the agenda (doc 12/0757r0) which covers three sessions (Mon ad-hoc, Tue AM2 and Thu AM2) to be held this week.

-An additional presentation was identified and a revision of the agenda was produced

-The revised agenda (doc 12/0757r1) was approved by unanimous consent.

Presentation:12/0854r0 –ISD Use Cases (Zhuang Yan – Huawei)

- Describes use cases with three different environments

1) High Density (e.g. Railway Station) where AP provides information on coffeeshop services and location

2) Long Range Discovery with multiple APs – Enterprise scheme with open and secure zones to support employees and others

3) Discovery based on Location Detection –example of a projector in a conference room

- In each case high-level scenario and requirements were provided

Questions/Comments

  • (Motorola Mobility)Mentioned that the range of issues extends to possible pre-association and post-connection methods. We will need to look at the elements of the methods that we can deal with and possibly separate from those that may be doable elsewhere.
  • (chair) uses material to get sense that we are looking at pre-association approaches.
  • (chair) asks whether the AP should be an involved agent in the discovery or whether it just passes it on?
  • (Clair Global) indicates that multicast DNS and similar cause quite a bit of overhead that can affect basic service delivery such as video streams
  • (FOKUS) raises point that a tunneled request through AP could raise security issue or whether a device even needs to associate
  • (Marvell) needs to consider density for all use cases – overhead like probes in those environments and with 11u data services could create various problems
  • (Cisco) would like AP to be transparent – could provide info about a printer or projector but details needed after association instead of AP needing to know about everything
  • (chair to Cisco) – would a device ask AP about a music service or pass it on
  • (Cisco) pass it on to something in network - could maintain some type of cache
  • (FOKUS) but that then requires some awareness of content
  • (Cisco) would be based on simple matches not on inspection of dialog
  • (Clair Global) use case of train station similar to venue with rock concert – heavy congestion in small area
  • (Marvell) traditional tunnel requires end-to-end where the service discovery could be multicast and end up in a lot of places. Also the handling of peer-to-peer could introduce other loading
  • (presenter) avoid multicast distribution of local AP has something
  • (Cisco) service discovery tunnels should have a constrained distribution scheme to avoid the flooding issue

Presenter did not have any straw polls

Presentation:12/0898r1 –Fast Initial Service Discovery – an Enabler for Self-Growing (Marc Emmelman – FOKUS))

- Provides a mechanism for service discovery – additional information in doc 12/0899r0

- Uses a cognitive decision making approach as services added in some networks

- Example is sensor network which needs to add a new report with a different delivery scheme which the sensors can then discover in their environment

- Network-based decisions help device determine the routing

Questions/Comments

  • (Marvell) – to get to network would seem we need more tunneling to reach these elements – could this add load where we don’t need it
  • (presenter) – looking to provide useful information at AP so other stations can find things
  • (Marvell) – growth could cause a lot of traffic – we need to be cautious
  • (Clair Global) – enterprise class networks can disable to lower data rates – smaller cells – so devices looking without getting beacons could be sending out quite a bit of data to support discovery.
  • (Huawei) – since this a study group the material should be helpful in characterizing problems and potential approaches to help establish the scope for future work.
  • (presenter) this is addressing potential areas not to be restrictive
  • (Qualcomm) – pre-association methods means that STA does not really know details of the AP capabilities – so we may want to provide some information to improve the data exchange
  • (chair) need to be careful to avoid producing a full session negotiation effort to support the service discovery
  • (Qualcomm) issue of white space snnounceemts and such via this vehicle
  • (presenter) white space domain already has its setup and shift approaches.

Presenter requested a straw poll. Chair asked if doc 12/0899 is relevant to the vote. Presenter indicated the doc has a bit different purpose.

-Are you in favour of adding the self-growing use cases to the ISD SG Use Case Document?

  • Yes - 6
  • No - 6
  • Abstain – 33

Review of Use Case document

Chair reviewed the current status of the Use Case Document (doc 12/0433r3). It has captured many use cases along the way as well as a rip of requirements that were presented. It is not yet a true ISD doc and should help us to develop the PAR and not a definitive set of requirements. Contribution 12/0684 was missing and the chair will be revised in the future.

So Chair asks group as to how we get to the scope of the group. Should we rework the requirements or go to the PAR scope directly? The sense was that working the requirements could take away from getting at the scope.

So chair agreed to work on the PAR. He presented the current draft version to get the group current.

Development of Group Scope

Chair started a drafting session to address the scope as presented in the agenda (doc 11/757r1). There was a lot of discussion and wordsmithing of the primary point and its bullets. This material will be held, as-is, to be continued later this week.

Adjourn

The meeting was adjourned at 11am.

Minutespage 1Dwight Smith, Motorola Mobility