Per stream grouping number statistical apparatus based on self-adapting non-linear sampling method
A non-linear, self-adaptive technology, applied in the direction of digital transmission system, transmission system, data exchange network, etc.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0022] From the class of sampling functions given in the definition of P(c) select a specific function as follows:
[0023] f(c)=[(1+u) c -1] / u; 0
[0024] where u is a constant parameter. It is easy to prove that as long as b=1+u is set, the above formula can satisfy the definition of P(c). n ^ = [ ( 1 + u ) c - 1 ] / u is an unbiased estimate when the above formula is used as a sampling function. At this point we can get the exact relative error: It can be seen from this that the change of n has little effect on the relative error. When n tends to infinity, the relative error tends to
Embodiment 2
[0026] We synthesize traffic flows of different distributions. Suppose we measure a fully loaded OC-48 (2.5Gbps) link, and the measurement time window is one minute. We have generated 3 kinds of business flows that obey different business flow size distributions: Pareto distribution, whose shape parameter is 1.053, and scale parameter is 4; exponential distribution, whose position parameter _=500 (that is, the average business The flow size is 500); evenly distributed, the business flow size is between 1 and 1000. Use f(c)=[(1+u) c -1] / u; 0
[0027] distributed
Embodiment 3
[0029] We use the self-adaptive non-linear sampling method for the statistics of the real business flow under the OC-192 link. The result is as Figure 4 shown. The results show that ANLS provides good accuracy for both large and small flows.
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 