16ng at 70th IETF meeting WG
----------------------------

Slot: Tue, 2007-12-04 17:40-18:40
Room: (Vancouver, Canada, The Westin Bayshore)

============================
Notes taken by:
   Ji Hoon Lee
============================

============================
Administrator (10 min)		Chairs
	- Blue sheets
	- Jabber scribe
	- Agenda bashing
	- WG report
============================

Document:
http://www3.ietf.org/proceedings/07dec/slides/16ng-3.ppt

============================
16ng problem statements (10 min)	Junghoon Jee
	- ID: draft-ietf-16ng-16ng-ps-goal-03
	- GOAL: Ready for IESG submission
============================

Document
http://www3.ietf.org/proceedings/07dec/slides/16ng-4.ppt

Status
- Rev 03 was submitted reflecting the AD's comments and resolution.

Conversation
Daniel Park: This document is in IETF LC status. If you have any comments,
  send it to IETF mailing list as well as 16ng WG mailing list.
Alexandru Petrescu: Are you going to be referring 802.16 or 802.16e?
Junghoon Jee: Current draft are referring both documents. 
  Idle/sleep mode are not in the base 802.16 document.


============================
IP transmission over Ethernet CS (20 min)	Hongseok Jeon
	- ID: draft-ietf-16ng-ip-over-ethernet-over-802.16-03
	- GOAL: ready for WGLC
============================
Document:
http://www3.ietf.org/proceedings/07dec/slides/16ng-1.ppt

Status
- More feedback needed from people outside 16ng.

Conversation
Daniel Park: We had a WGLC before. In my sense, it is needed 
  a little bit more work on this draft. But, the lastest version 
  of this draft is trying to resolve all issues related with LC. 
  Do we need one more LC?
Jari Arkko: How much changed?
Daniel Park: multicast CIDs, etc. Ok, we wish one more short WGLC.  
Jari Arkko: The length of LC is not important. Please get sufficient reviews.
Junghoon Jee: Reviewing of this document from 802.16 WG is required, but it's missing.
Daniel Park: Ok, we'll arrange a conference call with 802.16 WG.
 

============================
IP transmission over IPv4 CS (10 min)		Basavaraj Patil
	- ID: draft-ietf-16ng-ipv4-over-802-dot-16-ipcs-01
	- GOAL: resolution to issues
============================

Document:
http://www3.ietf.org/proceedings/07dec/slides/16ng-0.ppt

Status
- Rev 01 is available now
- Major issue from the last WG meeting was MTU size.
- Update on resolved issue on MTU
  + Recommendation needs to be changed: default is 1500 octets.

Conversation
- MTU issue
Kent Leung: Why 1440?
Basavaraj Patil: That's actully why I'm arguing againt what authors wrote here,
  saying that they should change the recommendation of default being 1440 to 1500.
  If you have any tunneling mechanism which may require that MS reduces MTU,
  then you can actually decide. It depends on type of network and topology of deployment.
Wesley George: I agree. It might be helpful to say alternatives. Another recommendation 
  is trying to raise MTU at transport portion. For exmaple, use jumbo frame like Gigabit 
  Ethernet as a transport mechanism, instead of doing something which might constrict MTU.
Dave Thaler: How does AR tell the MTU size in IPv4? You still have problems.
Basavaraj Patil: Agree. I wasn't thinking about that.
Gabriel Montenegro: Path MTU discovery is mentioned in both IPv4 and IPv6 cases. 
  is there something wrong? 
Basavaraj Patil: It's an option. I don't know whether implementation supports or not.
Kent Leung: IPv4 Path MTU discovery requires ICMP.
Alexandru Petrescu: There may be an 802.16 link-layer message carrying MTU.
Jari Arkko: DHCP option 26 can be used.
- ARP/NAT issues
Dave Thaler: ARP behavior is an implementation issue. This may not be needed.
Gabriel Montenegro: there was a consensus removing ARP stuff. We'll talk about ARP.
  Why do we mention NAT?
Basavaraj Patil: I think it's not even required in Appendix.


============================
IP over IEEE802.16 deployment scenario (10 min)	Ji Hoon Lee
	- ID: none
	- GOAL: initial discussion
============================

Document:
http://www3.ietf.org/proceedings/07dec/slides/16ng-2.ppt

Status
- no draft

Conversation
Alexandru Petrescu: MTU issue can be discussed here.
Kent Leung: How does WiBro's link model emulate point-to-point link?
Ji Hoon Lee: It doesn't emulate. It's just shared link model.
Gabriel Montenegro: Is there any IPv6 pilot deployment?
Ji Hoon Lee: I don't have information, but IPv6 products are ready.
Basavaraj Patil: I wonder how WiBro dealt with many problems 
  as we discussed here, since it used shared prefix model.
Unknown: What's this draft trying to solve? Is there any missing part in current drafts?
Gabriel Montenegro: This is in the charter and we debated before. 
  Please discuss how to structure this document on the mailing list.


============================
Closing					Chairs
============================