Recent traffic measurement studies from a wide range of working packet networks have convincingly shown the presence of self-similar (long-range dependence LRD) properties in both local and wide area traffic traces. LRD processes are characterized (in the case of finite variance) by self-similarity of aggregated summands, slowly decaying covariances, heavy-tailed distributions and a spectral density that tends to infinity for frequencies approaching zero. This discovery calls to question some of the basic assumptions made by most of the research in control, engineering and operations of broadband integrated systems. At the time being, there is mounting evidence that self-similarity is of fundamental importance for a number of teletraffic engineering problems, such as traffic measurements and modeling, queueing behavior and buffer sizing, admission control, congestion control, etc. These impacts have highlighted the need for precise and computationally feasible methods to estimate diverse LRD parameters. Especially real-time estimation of measured data traces and off-line analysis of enormous collected data sets call for accurate and effective estimation techniques. A wavelet-based tool for the analysis of LRD is presented in this paper together with a semi-parametric estimator of the Hurst parameter. The estimator has been proved to be unbiased under fractional Brownian motion fBm and Gaussian assumptions. Analysis of the Bellcore Ethernet traces using the wavelet-based estimator is also reported.