Research Article
P2P-Based Scalable Execution Platform for Algorithmically Transitive Network
@INPROCEEDINGS{10.1007/978-3-642-32615-8_11, author={Mikio Yoshida and Hideaki Suzuki and Hidefumi Sawai}, title={P2P-Based Scalable Execution Platform for Algorithmically Transitive Network}, proceedings={Bio-Inspired Models of Network, Information, and Computing Systems. 5th International ICST Conference, BIONETICS 2010, Boston, USA, December 1-3, 2010, Revised Selected Papers}, proceedings_a={BIONETICS}, year={2012}, month={10}, keywords={peer-to-peer network framework system data-flow architecture agents oneway RPC}, doi={10.1007/978-3-642-32615-8_11} }- Mikio Yoshida
Hideaki Suzuki
Hidefumi Sawai
Year: 2012
P2P-Based Scalable Execution Platform for Algorithmically Transitive Network
BIONETICS
Springer
DOI: 10.1007/978-3-642-32615-8_11
Abstract
“Algorithmically Transitive Network” (ATN) is a novel computational model based on a data-flow network, consisting of the following operations: a forward propagation propelled with node firing and token creation, a backward propagation caused by evaluating differential coefficients, and a topological alteration taken place by autonomous agents. In the research of the ATN, a simulation run on some parallel processing scheme is essential. As a flexible and powerful implementation scheme, the paper employs a P2P based distributed platform, and describes the mechanisms for simulation and P2P deployment of the ATN. The implemented platform has the following three features: flexible allocation of ATN nodes to the physical resources, unified description of communication between nodes, and several methods to realize high parallelism. The proposed scheme is also helpful to verify applicability of the employed P2P system.