The modification of system modeling algorithm based on self-similar characteristics
Lieguang Zeng · Journal of Tsinghua University(Science and Technology) · 2007
An autoregressive process of order one(AR1) accumulative algorithm was developed to generate a self-similar source to reduce algorithm complexity in self-similar system modeling.The self-similarity characteristics,a Taylor expansion,and an identity transform of the first Pollaczek-Khinchin equation are used to simplify the non-analytic generating function for the service time and the overall algorithm.The algorithm significantly reduces the arithmetic complexity and improves realization compared with the original network resource algorithm.The modified algorithm has been validated with an Ethernet test that gave the same results in 1/14 of the calculation of time.