Learning-Capable Communication Networks (LCCN)

Goal:
The Learning-Capable Communication Networks (LCCN) initiative aims to combine insights of several communication networking use cases involving application of various learning techniques. The goal is to work altogether following an experimental research approach to find a common baseline for a best learning architecture.

Context:
Operational procedures and protocols of today's communication networks typically use explicitly defined mechanisms and representations to reach the goals associated to their design. This practice results into numerous protocols having a restricted space for (self-)adaptability and sensitivity respective to their network context (e.g. network traffic conditions, failure conditions, etc). On the other hand, a wide spectrum of learning and optimization techniques is available such that networks could learn and optimize their behavior in the running context. 

Planning and milestones:
  • 2010/10/18: Submission of first version of the problem statement document
  • 2010/11/11: First get together meeting at IETF 79 Beijing
  • 2011/03/30: Preparation meeting at IETF 80 Prague  
  • 2011/07/27: Kick-off meeting at IETF 81 Quebec City
  • 2011/11/16: Side meeting at IETF 82 Taipei
How to cooperate?
  • Comment on the problem statement announced on the <irtf MINUS discuss AT irtf DOT org> mailing list
  • Share experience learned/results obtained when applying learning methods to representative use cases 
  • Send an email to one of the active initiative people (see section listing involved people https://sites.google.com/site/lccngroup/involved-people)
  • Subscribe to the LCCN mailing list : http://groups.google.com/group/lccn/
  • Attend LCCN sessions at the IETF/IRTF meetings (we also invite attendees to participate to IRTF activities and take these opportunities to meet and exchange on the topic addressed by this group).