I am taking a computer security class, and have a homework problem which has the following setup:
Now assume a dynamic version of the Bell-La Padula Confidentiality model that incorporates a suitable version of the low watermark property, where the updates involve the lub (least upper bound) function.
First, to be clear: I am not looking for an answer to this problem. (I have provided only a portion of the problem here for that reason.)
Now, we learned in class that the low-water mark principle applies to the Biba security model. I found the same result on Wikipedia and many other sites.
I understand that Bell-LaPadula has "no read up" and "no write down" principles, so if I have:
Top Secret > Secret > Confidential > Unclassified
Subject = [Secret, {P, Q}] Object A = [Secret, {P, Q}] Object B = [Confidential, {P}] Object C = [Top Secret, {P, Q, R}]
Then the subject can read A
, B
and can write to A
, C
. (Right?)
But, how can the low-water mark principle be applied to the Bell-LaPadula model? What effect would the low-water mark principle have in these cases?