The research began in December 2004, when I was hired to complete a java-based peering network simulation package.
After the completion I was familiar with basics of distributed data structures, peering network construction and maintenance protocols.
That mainly involved DHTs (distributed hash tables) and - the most simple version of peering protocol - Chord.

Then I had completed project extension (Spring 2005) which implied improvements of Chord, from reliability and efficiency viewpoints.
I've proposed (and evaluated via simulation) some basic replication strategies, connection multiplexing, which had been already suggested by other authors.
Also, I've researched into advantages of geolocation/spatial query routing and measured performance and reliability improvements via simulation.

Comparisons were performed with some hybrid and prototype systems like SETI@home and Avaki, so we had come to nice historical evolution view of the system design (Spring 2006).
Then, I've started with case studies of peering protocols, which included Tapestry, P-Grid and some other frameworks, (October 2006).
This led to research starting to converge to structural analysis, and search for possible improvements for existing implementation.

Then there was some search for possible applications, including distributed databases, knowledge storages and other areas like grid computing (Spring 2007).
At that moment I had to perform a case study related to possible network topology measures, which eventually led to some investigations in random graph models.
Also, I've started with possible problem formalizations, which made me into Spectral Graph Theory, lots of material related to spontaneous order and self-organisation (scattered conference reports during 2007).

Specifically, I am currently planning to release a report on results of modeling and performance measures of P-Grid protocol with expander topology support.
After this, I am planning to propose eigenvalue-based expander network synthesis algorithm with geographical/spatial model restrictions.
Also, I am looking forward to elaboration of problem related to structural optimization of content-delivery network, but this still has lots of incomplete details.

Also I've recently got quite profound acquaintance with data transmission protocols for large bandwidth-delay product networks, which may be used in CDN/grid/peering environments.
Also, you may check out my publications.

Thanks for the interest and happy reading,
Anton

---