A comparative study of adaptive risk vs. adaptive aggressiveness control in parallel and distributed simulation

This paper reports a new parallel and distributed simulation architecture for OMNeT++, an open-source discrete event simulation environment. The primary application area of OMNeT++ is the simulation of communication networks. Support for a conservative PDES protocol (the Null Message Algorithm) and the relatively novel Ideal Simulation Protocol has been implemented.Placeholder modules, a novel way of distributing the model over several logical processes (LPs) is presented. The OMNeT++ PDES implementation has a modular and extensible architecture, allowing new synchronization protocols and new communication mechanisms to be added easily, which makes it an attractive platform for PDES research, too. We intend touse this framework to harness the computational capacity of highperformance cluster computersfor modeling very large scale telecommunication networks to investigate protocol performance and rare event failure scenarios.

Download Full-text

Parallel and distributed simulation

Winter Simulation Conference Proceedings, 1995. ◽

10.1109/wsc.1995.478713 ◽

2005 ◽

Cited By ~ 4

Author(s):

R.M. Fujimoto

Keyword(s):

Distributed Simulation ◽

Parallel And Distributed Simulation

Download Full-text

Parallel and Distributed Simulation

Applied System Simulation ◽

10.1007/978-1-4419-9218-5_20 ◽

2003 ◽

pp. 457-486

Author(s):

Farshad Moradi ◽

Rassul Ayani

Keyword(s):

Distributed Simulation ◽

Parallel And Distributed Simulation

Download Full-text

LOAD BALANCING FOR PARALLEL PCS NETWORK SIMULATIONS

Journal of Interconnection Networks ◽

10.1142/s0219265900000123 ◽

2000 ◽

Vol 01 (03) ◽

pp. 173-193 ◽

Cited By ~ 2

Author(s):

AZZEDINE BOUKERCHE

Keyword(s):

Load Balancing ◽

Distributed Simulation ◽

Simulation Experiments ◽

Time Warp ◽

Conservative Scheme ◽

Simulation Techniques ◽

Parallel And Distributed Simulation ◽

Pcs Networks ◽

Network Simulations ◽

Stage 1

Parallel and distributed simulation techniques have been investigated in a number of studies to decrease the execution times of PCS network simulations. In this paper, we consider distributed simulation of PCS models using a two-state PCS simulation testbed which makes use of a conservative scheme at Stage 1, and of Time Warp at Stage 2, and focus upon the load balancing issue. We investigate and study several load balancing schemes for TDMA systems. Extensive simulation experiments were conducted on a cluster of workstations using a real suburban area serviced by an FCA-based PCS networks. Our results indicate clearly that careful load balancing scheme is important in the success of the PCS simulation model.

Download Full-text