Algorithms, key size and parameters report 2014


#1

The “Algorithms, key size and parameters” report of 2014 is a reference document providing a set of guidelines to decision makers, in particular specialists designing and implementing cryptographic solutions for personal data protection within commercial organisations or governmental services for citizens. This report provides an update of the 2013 cryptographic guidelines report (link below) on security measures required to protect personal data in online systems. Compared with the 2013 edition, the report has been extended to include a section on hardware and software side-channels, random number generation, and key life cycle management, while the part on protocols, for 2014 is extended and is a stand-alone study on cryptographic protocols (link included below). The EC Regulation 611/2013 (link below) references ENISA as a consultative body, in the process of establishing a list of appropriate cryptographic protective measures for personal data protection

Publication date: Nov 21, 2014

http://www.enisa.europa.eu/activities/identity-and-trust/library/deliverables/algorithms-key-size-and-parameters-report-2014


#2

Excellent read, for us we use future proofed algorithms which are identified
1: AES 265 -> cfb with unique IV’s
2: SHA512
3: RSA4096
Are the main ones, also
1: PBDFv2 with at least 5000 rounds
2: Rabin’s IDA and Shamir secret sharing

This allows folks to check the doc with these parameters.


#3

Also something that’s important is leaving the door open to future upgrade (i.e., bigger key sizes), as the current recommended sizes will inevitably become too small in the future.


#4

Yes its an important area, at the moment the key size is a single line change in the common lib. For RSA or other asymmetric keys this is OK as it is only a case of using the correct matching key regardless of size. For algorithmic changes though it’s more complex, we are debating versioning capabilities for wire protocols and encryption schemes (self encrypt is versioned). The concerns we are hopefully addressing are the easiest ability to upgrade protocols and wire formats as easily as possible. All doable, bit no way is simple. Very important though for sure.