Formal Analysis for Encryption- Based Multilevel Model for DBMS - Multilevel Security for Relational Databases - page 117

• Data attribute A i and security classification level attribute
C i (2 ≤ i ≤ n ):
•

i = defines that the data
t [ A 1 ] can be altered by users with tc security classifica-
tion level. [

with tE [ Ci

i

( )] KeytEA ([ ( ]) c

Ci i

tE C i can be modified by users with tc or
c i security classification levels. When t [ C i ] < t [ TC ], t [ A i ] ≠
null is defined as a tuple borrowed from the t ′[ A i ] of t ′ that
has t ′[ A 1 , C 1 ] = t [ A 1 , C 1 ] ∧ t ′[ TC ] = t ′[ c i ] = t [ c i ].
• Null value:
• t [ A i , C i ] = [null, c i ], ( c i < tc ) means that for each data attri-
bute A i , there are users with tc security classification level
that expect to borrow data owned by users with ci i secu-
rity classification level. Both t [ A i , C i ] = [null, null] and
t [ A i , C i ] = [null, tc ] mean that there are no data available
in the data attribute A i . The [null, null] case applies when
tc ∉ { L i , …, H i }; the [null, tc] case applies otherwise.

( )]

6.3 Integrity Properties

6.3.1 Entity Integrity

In the entity integrity (EI) properties, a multilevel relational database
instance r satisfies the entity integrity if the following requirement
exists [77]:
For each tuple t ∈ r:

A i ∈ AK ⇒ t [ A i ] ≠ null;

A i, A j ∈ AK ⇒ t [ C i ] = t [ C j ];

and

A i ∈ AK ⇒ t [ C i ] ≥ t [ C AK ].

The first requirement ensures that no tuple t ∈ r has a null value for
any attribute in the primary key attributes AK . The second require-
ment ensures that all the primary key attributes AK should have the
same security classification level in the tuple t ∈ r. The third require-
ment states that the security classification level of the nonprimary key
attributes must be greater than or equal to the security classification
level of the primary key attributes.