Re: Biometric encryption
From: John Elsbury (johne_at_neveryoumind.co.nz)
Date: 05/20/03
- Next message: sponge: "Re: Web security problem"
- Previous message: Walter Roberson: "Re: Biometric encryption"
- In reply to: Matt: "Biometric encryption"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]
Date: Tue, 20 May 2003 00:17:00 GMT
On 19 May 2003 13:22:12 -0700, mjnews@mljenkins.com (Matt) wrote:
>I have been wondering about this for a while now... Many companies
>are selling products which claim to encrypt files, passwords, ...
>using biometrics. Do these systems actually encrypt against the
>biometric template, or are they using a key that is stored somewhere
>with the template and given only when a correct match is given.
I think the advertising claims are a (over?) simplification.
What I suspect happens is that there is an encryption key stored
somewhere, and that the biometric component is used to authenticate
access to the stored key. If I were designing such a product I would
also make provision for a master password, somewhere, to override the
inevitable problems when the bio's metric changes.
>
>How can you encrypt against a biometric when the given match template
>is not always the same? If these systems are simply storing a key
>with the template, couldn't someone obtain the key from the database
>and use it to gain access to whatever is being protected? Or is there
>a way to ensure the key is never given unless the correct template is
>supplied? I can only think of two ways to do this. The first is to
>encrypt the data to be protected (or a key) against the biometric
>reference/enrollment template. The second would be to only allow the
>data (or key) to be decrypted by the program that verifies the match
>and reference template.
>
>Any thoughts?
Good biometric systems would track variations between stored and
current values, plot trends, and maintain a sort of "validity window"
in the middle of the stored parameter range, then calculate a new
window every time the authentication is successful. This takes care
of drift / change over time.
As in so many other instances, encryption merely moves the problem to
the key management space. Given that the PC platform is inherently
insecure, you are only going to get a moderate level of security
using a PC.
>Matt
- Next message: sponge: "Re: Web security problem"
- Previous message: Walter Roberson: "Re: Biometric encryption"
- In reply to: Matt: "Biometric encryption"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]
Relevant Pages
|