Skip to Main Content
Network emulation allows one to perform experiments in a controlled laboratory environment by applying network characteristics to live traffic. With recent advances in parallel simulation, it is becoming feasible to emulate large-scale networks at a high level of fidelity. However, high performance computing platforms are required. When such facilities are not locally available, remote computing facilities might be used. This approach is problematic, however, because of the latency introduced in the real time emulation environment to access the remote parallel machine. A distributed server-based architecture is proposed that utilizes local low-fidelity emulators to provide real-time QoS information to distributed applications, coupled with a remote high-fidelity simulator that is used to update and calibrate the local low-fidelity emulators continuously. The accuracy of this server-based emulation system is evaluated using test cases involving a worm propagation scenario.