P2PDisCo - Java Distributed Computing for Workstations Using Chedar Peer-to-Peer Middleware

Abstract

This paper introduces

Chedar: peer-to-peer middleware

“…In our research project [4] Chedar has been used for distributed computing [9] and data fusion [13] and extended also to mobile devices [10]. Peer-to-Peer Distributed Computing (P2PDisCo) software was built on top of Chedar to speed up the training of neural networks with evolutionary computing.…”

Mobile Chedar — A Peer-to-Peer Middleware for Mobile Devices

“…It also provides platform independence and quick adaptation to new hardware. Chedar has been programmed with Java 2 Standard Edition and is currently being used for speeding up the computations of NeuroSearch resource discovery algorithm [13] with P2P Distributed Computing application (P2PDisCo) [6] and for studying distributed data fusion in peer-to-peer environment [11].…”

P2PRealm - Peer-to-Peer Network Simulator

“…Combining both the internal code This was however not enough, because reducing execution time of one simulation case from a week to 3-4 days was still quite slow. As a solution, we started developing Peer-to-Peer Distributed Computing platform (P2PDisCo) [11] allowing the distribution of simulation cases to multiple machines.…”

“…This paper describes an end product of a process where emulators were first used for studying P2P algorithms and later re-implemented as an efficient simulator to decrease the time used for execution. Latest improvement of the simulator is the distributed execution on a Peer-to-Peer Distributed Computing platform (P2PDisCo) [11] allowing us to parameter sweep different features of neural network based P2P algorithms.…”

“…With multiple cases it is possible to do parameter sweeps and to eliminate the need of starting the simulator manually each time one wants to use multiple training environments. The execution of different cases can also be distributed on Peer-to-Peer Distributed Computing platform [11] further described in section 6.…”