MIP4 WG Minutes
===============

MONDAY, July 28, 2008
1300-1500 Afternoon Session I


2. Document Status

WG Documents:

    draft-ietf-mip4-rfc3344bis
    - Waiting for Shepherd Writeup
      
    draft-ietf-mip4-generic-notification-message
    - New version needed

    draft-ietf-mip4-nemov4-fa
    - Please review and comment

    draft-ietf-mip4-nemov4-dynamic
    - Re-submit, review and comment

    draft-ietf-mip4-gen-ext
    - Review and discuss based on
      draft-deng-mip4-host-configuration-00.txt

    draft-ietf-mip4-rfc2006bis
    - Waiting for final revision

    draft-ietf-mip4-dsmipv4
    - IESG Processing, waiting for processing at the same time as DSMIPv6

RFCs Published:
 
    draft-ietf-mip4-nemo-v4-base
    - RFC 5177

    draft-ietf-mip4-mobike-connectivity
    - RFC 5266

    draft-ietf-mip4-vpn-problem-solution
    - RFC 5265	


3. DHCP-based Host Configuration
   draft-deng-mip4-host-configuration-00.txt
   
    Henrik - Why use the MAC address instead of the NAI

    Hui - Captured and analyzed DHCP messages. They all use the MAC address, so 
	recommended using the MAC address

    Henrik - The use of MAC address is common, but does not prevent other forms
	of identies from being used

    Pete - 3G cellular links do not have the MAC address

    Henrik - It should be possible to use NAI as the client identifier

    Kent - Using the NAI for client identifier is common too

    Kent - On the use of broadcast IP address. HA is supposed to replicate the
	broadcast packets and sent to every mobile node. Is that applicable here?

    Hui - Yes

    Kent - This is a bad idea. HA should not be broadcasting these DHCP messages.
	So prefers using unicast addressing for the DHCP messages.

    Kent - Previous drafts spoke about informatiom from both the home and visited
	networks. This is addressed in the current draft.

    Ahmad - What is the intended status? The messages are already there. We are 
	only describing how to use these messages with MIP4

    Henrik - Informational might make sense.

    Jari - Are there examples of visited network information that you would want
	to delivered to the MN

    Kent - To get back to Jari on that

    Henrik - Why not use DHCP directly in the visited network to obtain 
	information from the local visited network?

    Pete - There might be an issue with sending a DHCP message locally broadcast 
	at the same time as sending one to the home network

    Henrik - Do the local DHCP request before setting up a tunnel with the home 
	agent

    Pete - There is an issue with FA mode. There is no local address for the MN to
	use.

    Ahmad - There was a draft from WiMAX about sending a broadcast/multicast 
	packets locally. Maybe that draft could be considered

    Pete - We need to look at the general issue of local breakout for broadcast 
	packets.
   

4. Generic Notification
   draft-ietf-mip4-generic-notification-message-06
   
    Ahmad & Sri - Some confusion. 

    Henrik - There is agreement already to make the changes to the draft. There is
	no point in re-hashing the discussion. Lets move on

    Vidya - Is integrity protection optional?

    Sri - No 
   

5. Better Than Nothing MIP4 Fast Handovers
   draft-doswald-robert-mip4-btn-fmipv4-00.txt

   This document proposes a complement to fast handover mechanisms
   according to RFC 4881 and RFC 4988.  Those require the presence
   of Foreign Agents; while the mechanism proposed here uses a direct
   connection to the Home Agent.
   
    Sri - What is the motivation for this? Any use-case?

    Alistair - Looking for a solution to improve handovers without a FA

    Kent - Can you give an example of such a network?

    Alistair - Has a use-case for mobiles using an existing infrastructure.
	(Some sort of street cleaning robots. )

    Kent - What kind of access technologies?

    Alistair - We are using WiFi and some 3G cellular technologies

    Henrik - This looks like delay tolerance is the most important thing
	here rather than fast handovers