TY  - RPRT
VL  - 3214
Y1  - 2012///
A1  - Bianchi, Alessandra
A1  - Bregni, Stefano
A1  - Crimaldi, Irene
A1  - Ferrari, Marco
AV  - none
M1  - technical_report
ID  - eprints1132
TI  - Analysis of a Hurst parameter estimator based on the modified Allan variance
N1  - Electronic preprint, University of Bologna, ISSN 2038-7954
UR  - http://eprints.imtlucca.it/1132/
N2  - In order to estimate the Hurst parameter of Internet traffic data, it has been recently proposed a log-regression estimator based on the so-called modified Allan variance(MAVAR). Simulations have shown that this estimator achieves higher accuracy and better confidence when compared with other methods of common use. Here we link it to the wavelets setting and provide an asymptotic analysis in the case the signal process is a fractional Brownian motion. In particular we show that the MAVAR log-regression estimator is consistent and asymptotically normal, providing the related confidence intervals for a suitable choice on the regression weights. Finally, we show some numerical examples.
KW  - Hurst parameter
KW  -  long-range dependence
KW  -  self-similarity
KW  -  modified Allan variance
KW  -  parameter estimation
KW  -  wavelets
KW  -  fractional Brownian motion
ER  -