Within ICAR-CNR, I am involved in projects that aim to experiment and assess the use of nature-inspired algorithms and protocols for Grid and P2P networks.
These algorithms follow the swarm intelligence paradigm: a number of entities perform simple operations at the local level, but these operations are combined at the global level so as to
engender an advanced form of intelligence and achieve complex objectives.
Other people involved in these projects are:
available (source code available)!
Self-Chord is a P2P
system that inherits the ability of Chord-like structured
systems for the construction and maintenance of an overlay of
peers, but features enhanced functionalities deriving from
ant-inspired algorithms, such as autonomy behavior,
self-organization and capacity to adapt to a changing
As opposed to the
structured P2P systems deployed so far, resource indexing and
placement is uncorrelated with network structure and topology,
and resource keys are organized and managed by self-organizing
mobile agents through simple local operations driven by
The benefits vs. classical structured systems
like Chord are essentially three:
1) Peer indexes and resource keys are uncorrelated, which opens
the possibility to give a semantic meaning to keys and perform
class (or range) queries
2) A better load balance among the peers can be obtained
3) Dynamic properties (e.g., management of joining nodes) are
improved. For example, it is not necessary to reassign keys when
new peers or resources are added to the system: the mobile
agents will spontaneously reorganize the keys. More information can be found on
A. Forestiero, C. Mastroianni, M. Meo, Self-Chord: a Bio-Inspired Algorithm for Structured P2P Systems,
9th IEEE/ACM International Symposium on Cluster Computing and the Grid, CCGrid 2009, Shanghai, China, 18-21 May 2009.
A. Forestiero, E. Leonardi, C. Mastroianni, M. Meo,
Self-Chord: a Bio-Inspired P2P Framework for Self-Organizing Distributed Systems, IEEE/ACM Transactions on Networking, April 2010. Published online.
The prototype is available here!
So-Grid is a set of P2P algorithms than can be adopted for the construction of a Grid information system.
So-Grid assumes that Grid resources are pre-categorized in classes.
To facilitate discovery operations, a number of ant-inspired mobile agents replicate the descriptors of the resources and accumulate the descriptors of different classes in different regions of the network.
Grid users are allowed to issues discovery requests to retrieve a set of resources of a specified class, and successively can choose the most appropriate for their use.
The discovery procedure is semi-informed: at the beginning, the search messages travel the network with a random walk strategy. As soon as one of them approaches a representative peer for the class of interest (i.e., a peer that stores a large number of descriptors of that class) the procedure becomes informed and the message is directed towards that peer.
More information can be found on this paper:
A. Forestiero, C. Mastroianni, G. Spezzano, "So-Grid: A Self-organizing Grid Featuring Bio-inspired Algorithms". ACM Transactions on Autonomous and Adaptive Systems, vol. 3, n. 2, May 2008.
The paper is downloadble from here.
You can also visit the So-Grid site to run your own simulations and analyze the behaviour of the So-Grid algorithms.
Antares is a different P2P approach for the construction a Grid information system.
Antares is devised for a Grid in which resources are indexed by bit keys, which are calculated with a hash function.
Resources that share common characteristics (for example hosts having similar CPU power or memory capacity) are mapped to the same key value. Moreover, if a locality preserving hash function is adopted, similar resources are assigned similar key values.
The objective of ant-inspired agents is to replicate and move resource descriptors, and spatially sort them on the network according to their key values.
Thanks to this sorting, a search message, issued to discover the descriptors having a specified key, can be forwarded, step by step, towards a region that has accumulated a number of resource descriptors having the desired key. The search message follows a path following the gradient of resource keys: at each step the message is forwarded to the neighbour peer that minimizes the distance between the target key and the keys of the descriptors stored in the vicinity of the peer. Therefore, with Antares, the search procedure is completely informed.
More information can be found on the papers:
A. Forestiero, C. Mastroianni, G. Spezzano, "Antares: an Ant-Inspired P2P Information System for a Self-Structured Grid". BIONETICS 2007 - 2nd International Conference on Bio-Inspired Models of Network, Information, and Computing Systems, Budapest, Hungary, December 2007.
A. Forestiero, C. Mastroianni, "A Swarm Algorithm for a Self-Structured P2P Information System". IEEE Transactions on Evolutionary Computation, vol. 13, n. 4, pp. 681-694, August 2009.
You can also visit the Antares site to run your own simulations and analyze the behaviour of the Antares algorithms.