The stochastic graph process solver (GROOVER) is a Monte Carlo intergation tool developed at the Dipartimento di Ingegneria e Scienza dell'Informazione (DISI) of the University of Trento; its target is to analyze Overlay Content Distribution Systems.
GROOVER explores the state space of a Markov chain whose states are represented by graphs: each graph contains the nodes that have been reached by the content and the edges used to distribute it.
GROOVER comes in two flavors:
- analysis of non real-time distribution process, such as File Distribution;
- analysis of real-time distribution process, such as Overlay Streaming.
In both cases the tool consider tree- and mesh-based architectures and can be easily configured to
obtain different scenarios: by properly adjusting the input parameters it is possible to obtain networks with 103-106 nodes, with different constraints on minimum and maximum outdegree.
D. Carra, R. Lo Cigno, E.W. Biersack, "Content Delivery in Overlay Networks: a Stochastic Graph Processes Perspective," in Proc. of IEEE Globecom 2006, Nov. 27 -- Dec. 1, San Francisco, CA, USA .
D. Carra, R. Lo Cigno, E.W. Biersack, "Fast Stochastic Analysis of P2P File Distribution Architectures," in Proc. of IEEE Globecom 2006, Nov. 27 -- Dec. 1, San Francisco, CA, USA.
D. Carra, R. Lo Cigno, E.W. Biersack, "Graph Based Modeling of P2P Streaming Systems," in Proc. of IFIP Networking 2007, May 14-18, Atlanta, Georgia, USA.