JSEP Fellowship: Fractal Noise Processes
Abstract
Three fractal stochastic processes were developed and defined. Their statistical properties were derived, and examples were provided. Fractal shot noise (FSN), which is formed when a homogeneous Poisson process is passed through a filter with a power-law decaying impulse response function, was defined. For certain parameters, FSN converges to fractional Gaussian noise, while for others it is a Levy-stable random process. The fractal shot-noise- driven Poisson process (FSNDP), formed when FSN becomes the rate function for a second Poisson process, was considered next. The FSNDP is a point process which has a power spectral density inversely proportional to frequency raised to an arbitrary power (between zero and unity) for certain ranges of parameters. Also developed, were two fractal renewal processes with power spectral densities similar to that of the FSNDP: a standard renewal process with arbitrary power between zero and unity, and a finite-valued, alternating renewal process where the power extends to two. Several applications of these three processes were examined in detail.
Document Details
- Document Type
- Technical Report
- Publication Date
- May 31, 1993
- Accession Number
- ADA268841
Entities
People
- George W. Flynn
- Malvin C. Teich
- Steven B. Lowen
Organizations
- Columbia University