The paper presents a novel approach to batch-mean estimation of Effective Bandwidth. The rationale of this work is to take advan- tage of the properties of renewal cycles to randomize the selection of the block sizes in the estimation process. At first, the theoretical foundation of the estimator is given; then, its performance is evalu- ated by comparing its variance to that of the traditional fixed block size estimator in a simple tandem queue scenario. Simulations high- lights the clear benefit achieved by the proposed estimator; such an advantage is particularly evident in case of short buffers, a desire scenario for delay-sensitive traffic flows.
A New Regenerative Estimator for Effective Bandwidth Prediction
PAGANO, MICHELE;PROCISSI, GREGORIO
2007-01-01
Abstract
The paper presents a novel approach to batch-mean estimation of Effective Bandwidth. The rationale of this work is to take advan- tage of the properties of renewal cycles to randomize the selection of the block sizes in the estimation process. At first, the theoretical foundation of the estimator is given; then, its performance is evalu- ated by comparing its variance to that of the traditional fixed block size estimator in a simple tandem queue scenario. Simulations high- lights the clear benefit achieved by the proposed estimator; such an advantage is particularly evident in case of short buffers, a desire scenario for delay-sensitive traffic flows.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.