1300 - 1310 10 minutes Chairs Administrativia * Behcet presented agenda as per slides 1310 - 1325 15 minutes Helen MLD/IGMP Tuning * Helen presented MLD/IGMP tuning as per slides and stressed the following points among others: * Aim is to optimize MLD/IGMP to be applicable to wireless or mobile networks * Limit the changes without affecting overall protocols and without introducing interoperability issues * Listed the proposed tuning options (pg. 4) * Hitoshi - * Unicast general query (listed on pg 4) is not well specified in RFCs, and may require some protocol extensions * General query and tuning response delay (listed on pg. 4) aslo may require some protocol extensions * Helen - * All tunings require protocol changes, but some are small and some are very large. Most of the items listed on pg. 4 will only require small protocol changes but one may rquire bigger changes * Hitoshi - * Okay. We can discuss detailed impacts later. * Helen went back to slides pg. 5-9 and completed the presentation without any more questions from the WG. 1325 - 1340 15 minutes Hitoshi MLD/IGMP Tuning * Hitoshi presented MLD/IGMP tuning as per slides and stressed the following points among others: * Apologized for not uploading his latest draft to meet IETF 79 deadlines * Showed simulation results using WiFi and IPv6 with 50 MNs, and also showed simulation results using WiMax * Tracking of membership state (pg. 5) is important and perhaps should even be considered to be a "MUST" * First tuning strategy (pg. 8) is to base tuning on query interval and query response interval * Second tuning strategy (pg. 9) is to based tuning on unicast query interval and unicast query response. But this may require a protocol extension * and then described three more tuning strategies (pg. 10-12) * Thomas Schmidt - It is good that several link types were considered by Hitoshi. However, still has not learned much about how MLD will be impacted by link types * Hitoshi- Maybe you want cross layer analysis, but MLD does not support cross link * Thomas Eckert (?) - Would feel happier if done with protocol extensions that are more explicit. For example if queries are supressed then may confuse the overall network (e.g. snoopers). So extensions should be explicit. * Hitoshi- This draft does not aim to change protocol * Thomas Schmidt - If someone wants deploy multicast enabled network, then you want to know which link types are being deployed. * Hitoshi- We cannot make perfect values for every type of links. That was the group consensus previously. 1340 - 1400 20 minutes Chairs Discussion on the charter item * Behcet presented some slides and stressed the following points among others: * Decision moment has now come for MLD/IGMP work item and we have to choose which of the two drafts presented we want to adopt * Stig - Charter does not allow extensions only tuning * Hitoshi - Explicit tracking has been defined somewhere else and we are just proposing to use it * Stig - There is a difference between manually entering values and have the protocol automatically adapting to the link * Behcet - Asked WG if we should adopt one draft or another, or should we merge the drafts? * Juan Carlos Zuniga - It is not clear what you are asking the WG to vote on since you seem to be asking exclusive questions * Stig/Behcet - Is there consenus to adopt draft-wu as a WG draft? * No one raised hands * Stig/Behcet- Anyone opposed? * 3 people raised hands * Stig/Behcet - Is there consensus to adopt draft-asaeda as a WG draft? * 2 hands raised * Stig/Behcet- Anyone opposed? * 1 hand raised * Stig/Behcet - Is there consensus in favor of merging the 2 drafts as a WG draft? * 3 hands raised * Stig - Perhaps we cannot conculde anything at this point. * Jari - 1 conclusion is that there is not much interest in the group before we can go further. If there is not enough interest we can drop the documents. If there is interest then we can push ahead. * Behcet - Another proposal is to move the IGMP/MLD "tuning" to the next charter * Jari - how many people would work on extensions? * 3 people raised hands * Behcet - So we will keep in the current charter and get some in depth reviews, as Jari suggested, before considering moving this work to the recharter * Jari - It is also okay to come back and say that there is no tuning required, but that perhaps an extension is required (as part of re-chartering work) 1400 - 1410 10 minutes Chairs Rechartering Discussion * Jari - Since the based protocol has now been submitted to IESG, we can now look at re-chartering and * Stig/Behcet- presented slides for rechartering items with 3 new proposed items * Thomas - Unahppy with narrow focus on sticking to PMIP. It was okay for the first charter of the group. But now it makes sense to consider for FMIP and other protocols, but of course still supports PMIP * Jari- Can you give some examples for the need to go beyond PMIP? * Thomas - Context transfer and direct routing could work beyond PMIP with other mobility protocols * Jari - I am a bit of trouble working on general solutions as it may overlap effort with other WGs * Rajiv Koodli - HI/HACK apply to MIP and PMIP. So if multicast applies to both then we should not limit to PMIP * Jari - We can specify that multicast is targeted to PMIP but can apply to MIP so that we still have a focus * Hitoshi - I prefer to keep focus on PMIP as well * Stig - Yes, we can target PMIP as main focus but also may work with other mobility protocols * Jari - Requirements documents are not mandatory * Behcet - Please send any comments on re-charter text today 1410 - 1420 10 Minutes Thomas Multicast Listener Extensions for MIPv6 and PMIPv6 Fast Handovers * Thomas presented his slides and stressed the following points among others: * Defines a coherent mulitcast membership transfer based on MLD/IGMP state records (easy handling) * New multicast mobility signalling * Carlos - Do you require support from MN to support this? * Thomas - Only for FMIP. Not for PMIP. * Rajiv - For example, protocol is between HI/HACK and does not need MN support for PMIP 1420 - 1430 10 Minutes Carlos Rapid acquisition of the MN multicast subscription after handover * Carlos presented his slides and stressed the following points among others: * Protocol extension to accelerate the MAG's knowledge of the MN's multicast subscription after HO * Thomas - Why do you consider predective and reactive HO? * Carlos - The new MAG might sense attachment of MN before old MAG detects detachment * Behcet - We need to take this discussion off line because of time pressures * Rajiv - We didn't you use existing protocols? * Carlos - We thought that we needed to stick to just PMIP 1430 - 1440 10 Minutes Juan Carlos Support Multicast Services Using Proxy Mobile IPv6 * Juan Carlos presented his slides and stressed the following points among others: * Merging of draft-zuniga and draft-contreras * We gave more details of deployment scenarios and PMIP protocol impacts * Thomas - Why don't you merge with the access router solution as one more option? * Juan Carlos - It is a similiar case as was raised for the last presentation on predictive and reactive HO 1440 - 1450 10 Minutes Shuai Multicast Source Mobility Support in PMIPv6 Network * Shuai presented verballly his slides * Thomas - Are you looking at source specifc or any source multicast? * Shuai - Now we are looking at any source multicast, and in the future we will look at source specific * Thomas - This may be a problem as it is group specific 1450 - 1500 10 Minutes Gyu Myoung Multicast for Content Delivery * Gyu presented his slides and stressed the following points among others: * Previously we stressed device, user mobility but we should also consider content mobility (content moving) * Proposal: Content mobility (multicast server handoff, content moving)