Abstract
This paper describes a new algorithm for the generation of pseudo random numbers with approximate self-similar structure. The Simple Self-Similar Sequences Generator ( 4SG) elaborates on an intuitive approach to obtain a fast and accurate procedure, capable of reproducing series of points exhibiting the property of persistence and anti-persistence. 4SG has a computational complexity of O ( n ) and memory requirements of the order of log 2 ( N ) , where N is the number of points to be generated. The accuracy of the algorithm is evaluated by means of computer-based simulations, recurring to several Hurst parameter estimators, namely Variance Time (VT) and the Wavelets-based estimator. The Hosking and the Wavelets-based methods for the generation of self-similar series were submitted to the same tests the 4SG was analysed with, providing for a basis for comparison of several performance aspects of the algorithm. Results show that the proposal embodies a good candidate not only for on-demand emulation of arbitrarily long self-similar sequences, but also for fast and efficient online simulations.
Paper version not known (
Free)
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call