On Sat, Apr 12, 2003 at 12:54:43PM -0700, Ian Clarke wrote: > Hui Zhang has done another simulation of our probabilistic caching > heuristic - you can find the results here: > > http://netweb.usc.edu/~huizhang/New_topic/inverse_hop_caching_II.pdf
Fascinating. Two questions: 1. Does inverse hop caching scale reasonably well? We can live with a 60% success rate for 20% saturation, if it doesn't get much smaller with a bigger network. 2. We can't use enhanced-clustering for security reasons. However, we could use the following algorithm: Until the RT is p% full, do inverse hop caching. After that point, cache probabilistically according to the node's perception of its key distribution (a crude measure would be the 256-buckets histogram, a more sophisticated one would be the ngrouting estimator that ian and mjr have been working on). It is important to bear in mind that the simulations may be hopelessly naive... network topology, insert patterns, and many other things are probably a lot different to the assumptions. However this is all rather interesting. > > -- > Ian Clarke [EMAIL PROTECTED] > Coordinator, The Freenet Project http://freenetproject.org/ > Founder, Locutus http://locut.us/ > Personal Homepage http://locut.us/ian/ -- Matthew Toseland [EMAIL PROTECTED]/[EMAIL PROTECTED] Full time freenet hacker. http://freenetproject.org/ Freenet Distribution Node (temporary) at http://80-192-4-36.cable.ubr09.na.blueyonder.co.uk:8889/T1LAykfe8-A/ ICTHUS.
pgp00000.pgp
Description: PGP signature
