Last Modified: 2003-03-10
In the past, this WG has focused on exploring a broad range of MANET problems, performance issues, and related candidate protocols. Under this revised charter, the WG will operate under a reduced scope by targeting the promotion of a number of core routing protocol specifications to EXPERIMENTAL RFC status (i.e., AODV, DSR, OLSR and TBRPF). Some maturity of understanding and implementation exists with each of these protocols, yet more operational experimentation experience is seen as desirable. Overall, these protocols provide a basic set of MANET capabilities covering both reactive and proactive design spaces.
With this experimental protocol base established, the WG will move on to design and develop MANET common group engineered routing specification(s) and introduce these to the Internet Standards trgraphics/. Lessons learned from existing proposals will provide useful design input, but the target for this effort is a common group engineering effort not a recompilation of an existing approaches.
As part of this effort, the WG will address the aspects of security and congestion control in the designed routing protocol(s).
This working group will work closely with the Internet Research Task Force (IRTF) groups on Mobile Ad Hoc Networks (RRG) for trgraphics/ing and considering any mature developments from the related research community.
|Done||Post as an informational Internet-Drafts a discussion of mobile ad-hoc networking and issues.|
|Done||Agenda bashing, discussion of charter and of mobile ad hoc networking draft.|
|Done||Discuss proposed protocols and issues. Redefine charter.|
|Done||Publish Informational RFC on manet design considerations|
|Done||Review the WG Charter and update|
|Done||Submit AODV specification to IESG for publication as Experimental RFC|
|Done||Develop I-D for potential common manet encapsulation protocol approach|
|Done||Submit initial I-D(s) of candidate proposed routing protocols and design frameworks|
|Done||Promote implementation, revision, and testing of initial proposed I-D(s)|
|Done||Explore basic performance and implementation issues of initial approaches|
|Done||Explore proposed proactive protocol design commonalities|
|MAR 03||Submit DSR specification to IESG for publication as Experimental RFC|
|MAR 03||Submit OLSR specification to IESG for publication as Experimental RFC|
|MAR 03||Submit TBRPF specification to IESG for publication as Experimental RFC|
|JUL 03||Develop a further focused problem statement and address an approach for a common engineering work effort|
|NOV 03||Reevaluate the WG's potential based on the problem statement consensus|
|RFC2501||I||Mobile Ad hoc Networking (MANET): Routing Protocol Performance Issues and Evaluation Considerations|
56th IETF MANET Working Group Minutes 20 March 2003 Thanks to Brian Adamson for scribing the general meeting minutes. ---------------------------------------- ---------------------- Mobile Ad Hoc Networks (manet) Agenda Thursday, Mar , 1500-1700 1. Agenda Bashing (5 min) 2. Review of Charter/Milestone and IRTF Update (10 min) Working Document Progress Updates 3. AODV Update (20 min) 4. DSR Update (20 min) 5. OLSR Update (20 min 6. TBRPF Update (20 min) 7. Further Progression of Documents (5 min) 8. Next Steps: Problem Statement Consensus and Standards Trgraphics/ Approach (20 min) ---------------------------------------- ----------------------- Joe Mgraphics/er (WG co-chair) presented the meeting agenda. Scott Corson (WG co-chair) was absent from this meeting. Rechartering Review (see updated IETF website) Joe Mgraphics/er next presented the MANET rechartering status. A new charter and updated milestones have been approved and posted. In reviewing the charter text, it was pointed out IESG input added an item concerning congestion control of MANET routing protocols to the working group charter goals. There was Chair and WG consensus that the statement in the charter concerning this should be clarified. (I.e. make it clear that the congestion control mentioned does not imply 'TCP fairness' but instead is intended to make sure that 'implosion' avoidance of routing control, etc is addressed and the protocols performance issues in these regards are understood.) An overview of the interim working group goals and milestones was provided. It was announced that the IESG had approved AODV for Experimental and that it should be in process with the RFC Editor. IRTF Update Elizabeth Belding-Royer provided an overview of IRTF Ad hoc Network Scaling (ANS). Contact and mailing list information was provided. Both Elizabeth and Scott Corson will be serving as co-chairs. See the IRTF ANS Website: http://ww.flarion.com/ans-research IRTF ANS goals were presented. Scalability with an emphasis on developing metrics for assessing protocol performance is a primary focus. There was some discussion concerned what problem spaces would be taken on within the IRTF. A 'Bar BOF' was announced in the evening of 9:30 PM, 20 March 2003 to gather more interest and feedbgraphics/ for IRTF ANS topics, goals, etc. Hilton San Francisco hotel bar will be the location. The first intended ANS meeting was announced to take place with the ACM Mobihoc conference: Sunday 7-10PM 1 June, location Annapolis, Maryland with Mobihoc. AODV Status Charlie Perkins presented an AODV update. There were a few changes to the document as it made its way through the recent IESG review. There was added clarification for handling subnet routes. It was clarified that subnet routes are handled the same as host routes. If a specific host on a subnet is unreachable, the subnet leader may send an ICMP 'host unreachable' instead of a route error message for the particular subnet host(s). Editorial changes, applicability statement, IANA considerations, etc. DSR Status Dave Johnson presented an update on the DSR ID status. A small number of changes occurred, mainly some rearrangement and integration. An optional flow state extension was, however, added bgraphics/ into the draft where it was previously in a separate 'DSR Flow' draft. Q: If you lose the first pgraphics/et with the full source route, what happens? A: A route error is sent bgraphics/ on pgraphics/ets with unknown flow id. Q: Is the interim pgraphics/et loss a concern? A: DSR salvaging usually handles this and there are documented studies of pgraphics/et delivery ratios. OLSR Status Thomas Clausen (INRIA) presented an update on the OLSR ID status. A recent ID was posted and included a number of changes, most of which were editorial and involved reorganization of the document to improve readability. A clarification was provided regarding the handling and discussion of multiple interfaces. Also, the document now clearly separates auxiliary features such as host/network association messages from core protocol features. A 'validity time' has been added to the messages. One mistake was made in that proper updating of the internal routing table with 2-hop information was accidentally dropped in the transition from version 7 to version 8 of the spec. This will fixed ASAP. Thomas indicated that there have been a number of reviews and comments already since the version 8 draft release. What next? Chair comment: Single interface and multiple interface sections are pretty redundant. Could be integrated? Document would shorter, simpler. A: Yes with the consensus of the working group this could be done. Chair Comment: 'Willingness' parameter can be zero and this MAY affect some of the election algorithms, state maintenance, etc. Recommend doublechecking this. A: Agreed. Thomas: Feels document is ready to move forward to Exp. RFC pending integrating current feedbgraphics/ and comments. The Working Chair recommended an additional quick revision cycle (or explanation of known changes) to provide working group members a little more time for review and input. Thomas: Will commit to quick turn-around. AD Comment: Read ID guidelines, etc to quickly meet IESG requirements. Thomas: Can the ADs review the current security considerations to help with this? AD: Yes. TBRPF Status Richard Ogier presented a TBRPF ID update. The document now references the Secure Neighbor Discovery (SEND) working group. Richard feels document is ready for Exp. RFC submission. Richard Ogier went on to present suggestions for merging proactive routing protocol ideas. Chair comments: With regards to not sending periodic topology updates when there are no topology changes? With regards to protocol robustness what is the current level of understanding, while a good scalability idea, is robustness questionable? a lost update is now a more significant problem to be resolved. Agreed it is worth further discussion. There was some discussion among the OLSR and TBRPF authors that they were attempting to work together. An IPR discussion ensued relating to peoples level of comfort in working together on a future document. The authors were attempting to work these issues out. Further Work Strategy Joe Mgraphics/er presented some strawman discussion topics for future work strategies. First and foremost the WG needs the authors commitment to revise the documents as needed and address any issues with present upcoming submission goals. We would like to see WG last call inputs occurring within 2 weeks or so. Relating to the chairs schedule issues, most authors indicated a commitment to quickly respond to WG input. Upcoming WG issues beyond progression of the current documents are: - some level of problem statement consensus. - IETF vs. IRTF problem split considerations. - agreement on consensus-based engineering approach. Discussion ensued including any relationship with the previously briefed OSPFv3 with manet extensions work (Fred Baker). It was stated any proposals like this would be further brought up within the OSPF WG for possible consideration first. The chair and others expressed interest in this related work but the proper forum and scope had not been identified. This was not seen as a replacement for the manet routing work ongoing, but may provide some practical application. Relating to security it was stated that some degree of further threat assessment should be developed for helping in understanding MANET security considerations. WG Comment: Many of the ongoing security working areas need people with wireless experience involved. Suggest interested parties get involved. Ongoing work in more distributed security and routing area solutions may likely apply. The chair asked the WG to begin considering the details of appropriate engineering issues for the next phase of work. Open Discussion There was some discussion of a common problem statement understanding. Using numbers of nodes was considered non-exact because of the numerous problem dimensions involved in any performance scenario. The chair also commented that the WG should avoid the 'bakeoff contest' method since this tended to stimulate more research-oriented approaches in the past and less group-based engineering. Comment: Need typical example networks in mind or we go down the rathole of the past. Chair: Agreed. Need some level of applicability and problem statement understanding is needed. Comment: There appears to be some competitive spirit here for what follows, but collaboration requires something different what are thoughts on the overall architecture. Will the working group be split up into sub-components to tgraphics/le different areas? Chair: Were not all going to agree 100% we need to find the common overlapping areas through discussion and focus on core components that can be engineered we will then build off 'lessons learned' to date. Comment: Believes 'component based architecture' can help the group to work together. Other WG comment: Agrees we need a family of components to work together including support for connection/gateway to the Internet More WG comments were raised relating to the specific approach that could be taken. Chair: Dont forget, while I am sensing among some members a reluctance to target the simpler problem space, we presently dont have 'Standard' ways to do manet routing even for small networks with moderate mobilitythat would be useful specification output from this group (this group has already produced useful experimental output and a set of working prototype implementations) but at the same time any design shouldnt preclude the ability to scale well through design extensions (e.g., fuzzy-sighted link state extension) or simply based upon inherent design properties, etc. ultimately good core designs should be extensible and also work reasonable well on a moderate scale without extensions. Q: Are we committed to two protocols classes (proactive, reactive)? Chair: This is undecided it is one taxonomy to consider (and has some rationale based upon previous WG development and work elements), but hybrid behavior and others issues may be addressed making the taxonomy less accurate. It is up to the WG to discuss what seems sensible in the near future it is likely that 'one size fits all' wont make it, but the solution space should perhaps be limited to two at most At this point, the meeting was adjourned. The chair asked the presenters to please submit their slides electronically if they had not already done so.