180
D. Hurley-Smith and J. Hernandez-Castro
Fig. 10.3 Distribution of χ 2 scores for Quantis devices. (a) 16M. (b) 4M
10.5 Conclusion
There are many complex issues to consider when evaluating RNGs for use in
security systems. Device specifications, the use of off-device entropy pools, postprocessing, and output speed are all critical to the evaluation process. Each element
should be tested in isolation, but only the whole device may be certified, leading
D. Hurley-Smith and J. Hernandez-Castro
Fig. 10.3 Distribution of χ 2 scores for Quantis devices. (a) 16M. (b) 4M
10.5 Conclusion
There are many complex issues to consider when evaluating RNGs for use in
security systems. Device specifications, the use of off-device entropy pools, postprocessing, and output speed are all critical to the evaluation process. Each element
should be tested in isolation, but only the whole device may be certified, leading