Fine features extraction of burst signals using fractal lacunarity

摘要

Fine features extraction is the key in signals identification. According to the non-universality of fractal dimension in features extraction, a new method was presented to extract the transient features of burst signals, which was based on the fractal lacunarity. The method used amplitude as the scale to calculate lacunarity. Lacunarity was more effective in describing the transient features comparing with fractal dimension. Experiments in AWGN and multipath channel show that the new method is viable.