Pseudo-random functions
Our next experiment is just to make sure that the results obtained previously are still valid for the case of estimating a proportion.
As the figures below show, results are similar as previously: all the implementations for the xi and hash functions seem to provide the same standard error.
Parameter | Value |
---|---|
Packets | 10000 |
Drop probability | 10% |
Columns | 32 |
Rows | 32 |
Digest size | 32 |
Hash function | {EH3, BCH3, BCH5, CW2, CW4} |
Xi function | {CW2, CW4, Tabulated} |
Pcap | CAIDA |
Average function | mean |