Provisioning End-to-End Quality of Service with Fast Importance Sampling based Traffic Engineering

摘要

Simulation is a popular tool of network traffic engineering. Knowing how the network performance will change with respect to the network parameters is the key to achieve a target performance (QoS). The traditional crude Monte Carlo (CMC) method is very time consuming, especially for rare event simulation. A previously proposed Importance Sampling based Traffic Engineering (ISTE) approach is faster than the CMC approach yet ISTE may still require additional sample performances aside from the initial sample performance. The Fast Importance Sampling based Traffic Engineering (FISTE) approach can carry out all the tasks of ISTE requiring only a single sample performance thus making FISTE even more time efficient.