On the characteristics and origins of internet flow rates
Proceedings of the 2002 conference on Applications, technologies, architectures, and protocols for computer communications
An Introduction to Copulas (Springer Series in Statistics)
An Introduction to Copulas (Springer Series in Statistics)
| Hi-index | 0.00 |
A correlated flow of arrivals is considered in a case, where interarrival times {Xn} correspond to the Markov process with the continuous state space R+ = (0,∞). The conditional probability density function of Xn+1 given {Xn = z} is determined by means of q(x|z) = q(x|Xn = z) = σi=1kpi(z)hi(x), z,x∈R+, where {p1(z),..., pk(z)} is a probability distribution, p1(z)+...+pk(z) = 1 for all z ∈ R+; {h1(x),..., hk(x)} is a family of probability density functions on R+. This flow is investigated with respect to stationary case. One is considered as the Semi-Markov process J(t) on the state set {1, ..., k}. Main characteristics are considered: stationary distribution of J and interarrival times X, correlation and Kendall tau (τ) for adjacent intervals, and so on. Further one is considered a Markovian system on which the described flow arrives. Numerical results show that the dependence between interarrival times of the flow exercises greatly influences the efficiency characteristics of considered systems.