Comment Re:And this is different from OSPF how? (Score 1) 64
you should read the original papers before saying that the authors play semantic games
http://www.idsia.ch/~gianni/antnet.html
the differences are many, here are some important ones. Full paths are explicitly sampled by control packets (the 'ants') (wich is quite different from locally observing link costs and then flooding them). Sampling full paths also involves some
core issues that do not find any counterpart in ospf: how often ants are generated (i.e., when, how often, do I need to refresh my local routing information?), which destinations should be sampled? (some destinations are more important than others...), what kind of stochastic decision policy should be implemented in the hop by hop decisions? (how to balance exploration vs. exploitation). The routing tables adaptively constructed through ant sampling do not identify one single shortest path, but rather a bundle of paths that can be used to automatically implement stochastic multi-path routing (and have therefore automatic load balancing). Last but not least, OSPF is based on the knowledge of the full topology of the network (at the level the router is), that implies keeping flooding of up-to-date link information throughout the network. On the other hand, AntNet's routing table are just based on the knowledge of the local topology, that is, of the connected neighbors. This also makes the algorithm quite robust for possible ant losses (or 'errors' in the estimate of the quality of the sampled paths).