Minutes of the Mobile Ad-hoc Networks WG (manet) 55th IETF Proceedings Thursday, Nov 21 at 1300-1500 Minutes taken by Justin Dean ============================== CHAIRS: M. Scott Corson Joseph Macker AGENDA: (10 min) Agenda Bashing and Announcements (20 min) Proposed Manet Way Forward (10 min) Strawman Proposal for Charter and Milestone Update (10 min) Related IRTF Activity Announcement (10 min) Open Discussion of Proposal and Issues (15 min) TBRPF Update (15 min) DSR Update (15 min) OLSR Update (15 min) AODV Update Closing Joe Macker presented the agenda and opened it up for bashing. The agenda was accepted as is with no bashing. Joe Macker had some issues regarding the manet mailing list that he wished to remind members about. - There are 1969 members currently on the manet mailing list. - Only members are allowed to post to the mailing list. Somewhat standard practice, but not a perfect way to reduce spam, etc. - Manet members should be aware when switching account names so that they may still post effectively without moderator intervention. - The newer ietf.org mailing list is reasonably automated; please use the automated features to manage your own accounts whenever possible. - NS2, GLOMOSIM, QUALNET, OPNET simulation software questions have recently increased in volume. Often simulation environment specific questions are not appropriate for the mass manet mailing list, members are urged to contact the authors of the code or appropriate simulation mailing list for instruction questions, FAQs, etc. MANET Way Forward (for details see presentation in proceedings) Joe Macker presented an overview of a proposed way forward for the manet WG that was being worked on in conjunction with the ADs. Recent perceived WG status is that manet prototype implementation, designs, and experimentation are very active. However, the WG has experienced significant “research creep” into the ongoing work by continuing to consider broad work issues in the manet problem space. It is the intention of the chairs and the ADs to update the charter of the WG to scope down to a focused engineering mode of work. The core approach is: · reduce the near term scope of the present WG · formulate an IRTF venue for manet research items, split off related longer term research work from IETF · complete several mature unicast protocol ID work as EXPERIMENTAL RFCs · formulate a focused problem statement(s) · convert core work items to an engineering mode and target common Proposed Standard effort(s) Alex Zinin: Routing co-AD (for details see presentation in proceedings) Alex Zinin next discussed the proposed rechartering effort. Alex mentioned that he had posted a note to the list summarizing some of the issues as they have emerged a few weeks back. The summary is that very useful work is being done in manet, but the WG is operating under a wide problem statement. The perception is there are many competing protocols, a lack of WG-wide consensus, and no signs of convergence. All of the above == research mode. The goals of the proposed charter update would be to convert the WG to the IETF mode: Clear and focused problem statement Focused WG charter, clear milestones WG members working closely towards consensus Design IETF MANET protocol STD Work that needs more research (yes, it is somewhat subjective) will move to the IRTF. Proposed Steps: First, we create an IRTF sub-group for MANET and move LANMAR, FSR, ZRP, TORA and similar work into this forum. We would scope the charter to push DSR, AODV, OLSR, and TBRPF to EXP and when done the WG will better define the engineering problems statement. IF rough agreement is reached on the problem statement we will update the charter with the development of IETF protocol(s) that will go STD tracks. Plan of actions We will discuss this more here and on the mailing list (nothing so far). We will announce the decision within a month after this meeting. Questions: Q: Will the rechartering approach cause delay in pushing forward EXP RFCs for certain protocols already under consideration? A: No, that is not intended. Q: Will multicast work in manets be carried out in ietf or irtf? A: Lots of papers on this subject already, yet it remains to be decided what is appropriate for engineering. Stay Tuned. ------------------------------------------------------------- At this point, the Joe Macker polled the room for consensus on re-chartering direction. Room consensus was attained to re-charter based upon the approach presented.. Further mailing list input will be taken before reaching any decisions. Strawman Recharter Presentation - Joe Macker (for details see presentation in proceedings) Joe Macker presented some text and line items for consideration in an updated charter/milestone effort. More discussion of this will take place on the mailing list. Strawman Near Term Milestones Present: Restructure WG to be more narrowly focused, split off important related IRTF work Jan 2003 SUBMIT core reactive protocols for EXPERIMENTAL RFC (if not already done) Jan 2003 Revisit WG goals and problem statements Mar 2003 SUBMIT core proactive protocols for EXPERIMENTAL RFC Jun 2003 Develop and approve focused MANET WG problem statements and scoped engineering goals. Scott Corson: IRTF MANET research subgroup Scott presented an overview of the manet-related IRTF plan. Scalability is the main research theme and some of the goals will be to foster discussion and better understand the limitations of existing approaches (e.g., simple flat reactive and proactive). The group will look at proposing/developing approaches that demonstrate superior performance (e.g., Hybrid/hierarchical forms and radically new concepts). More details will be formulated. It has been a concern that meetings may not be long enough and it is likely that meetings may take place in concert with ongoing events to ease travel as much as possible. A new mailing list will be set up and announced on mailing list Q: Will multicast be in IRTF or IETF? A: Depends on the problem statement that the IETF works out and it may be a mixed case given appropriate rationale of problems statements. Q: What is small/medium networks? Maybe 10-100 nodes but what about movement models? What type of movement scale are we using? A: We are aware that dynamic link/motion models, traffic flow models, network surge conditions, and other factors affect any scalability assessment besides number of router nodes. We need some discussion of this in problem and applicability statements. Q: Other IETF groups have scalability in the charters so why are we getting rid of it in this charter? A: It is not the intention to artificially limit the potential for large area scalability, however we are not making it a requirement for engineering maturity and applicability reasons. Comment: IRTF should be looking at the limits of current protocols work as well. This may be fed back to justify changing things in IETF. Q: Would like to see IRTF and IETF groups work together closely; meetings at same conference? A: Yes but the IRTF participants need freedom to do their research as well. Q: Need to keep problem statement simple. Where does security belong? What about IETF issuing standard without security in it? A: Need to discuss closer with the routing security working group. Comment: I see a need for simplicity of the protocol. Comment: I think it would be extremely useful to get RFCs out that can be used for 50-100 node scenarios. Comment: HIP ipsec can help to bring improved security to manet. Other comment: Systems are being develop that should work with wireless networks. Chair: Great comments, but we need to move one please bring further discussions to the list. TBRPF Update: Richard Ogier *see online presentation TBRPF is now at Draft revision 6. It has been rewritten to improve readability. Hello message has been modified to include relay priority similar to OLSR MPR preference. It was clarified that tbrpf computes the equivalent of MPRs. There is now a relay priority in each hello.. OLSR MPR willingness is same type of thing and this presents the possibility of merging with OSLR. Other details (see presentation). There is a new SRI IPR statement registered at ietf.org. DSR Update: Dave Johnson (see on-line for details). The final DSR ID for EXPERIMENTAL consideration is not finished yet. We are trying to add a few things to back into draft. Stuff that was in version 3 of the ID. OLSR Update: Thomas Clausen (see on-line for details). A new version of OLSR was update (now version 8). Posting to ID editor was munged and the draft was posted to the list in the interim. There are no major changes. The feedback was that most people intending to use OLSR didn’t want all the features. They wanted something like version 3 or 4. AODV Update: Elizabeth Belding-Royer AODV is now at ID version 12. There we no major changes in this revision. There was some clarification of local proactive repair. http://sourceforge.net/projects/aodvimpl/ for aodv questions. http://moment.cs.ucsb.edu/AODV/aodv.html