Information Technology Reference
In-Depth Information
3.1
Syntax
We start out by introducing a collection of principal expressions, ranged over by
P and Q . Letting A range over a countable set of simple principal names, the
abstract syntax of principal expressions is given as follows:
P ::= A
|
P & Q
|
P
|
Q
|
P for A Q
The principal P & Q represents a compound principal who makes exactly those
statements made by both P and Q . P
Q represents an abstract principal cor-
responding to principal P quoting principal Q . P for A Q represents a principal
P acting on behalf of principal Q : P for A Q is syntactic sugar for P
|
Q ,
where A is a principal that vouches for P 's authorization to make statements on
Q 's behalf [1].
For the logic itself, we let p range over a countable collection of primitive
propositions and define the abstract syntax for the logic as follows:
|
Q & A
|
ϕ ::= p
ϕ
|
ϕ 1
ϕ 2 |
ϕ 1
ϕ 2 |
ϕ 1
ϕ 2 |
ϕ 1
ϕ 2 |
P says ϕ
|
P
Q
P serves T
|
P
T Q
|
Q
|
P
T Q
Here, T ranges over sets of formulas that include only formulas of the forms
included on the first line (e.g., not involving
T , serves T
,or
T ); the forms
involving T are our extensions.
Primitive propositions are used to represent requests and permissions, while
the formula P says ϕ represents principal P making the statement ϕ . In turn,
P
Q represents a relationship between principals P and Q through which
statements of P can also be attributed to Q . The formula P
T Q , pronounced
as “ P is mimicked by Q on T ,” is our extension to the logic, inspired by Howell
and Kotz's restricted speaks for relation
T
[9]. This restricted mimics formula
represents a weaker relation than P
Q , in part because only P 's statements
from the set T can be attributed to Q . Finally, P serves T
Q (the restricted serves
relation) and P
T Q (the restricted inherits relation) are syntactic sugar for
P
|
Q
T A
|
Q and ( P
Q )
( Q
T P ), respectively.
3.2
Semantics
The semantics of the logic is based on Kripke structures. A Kripke structure is a
triple
,where W is a set of possible worlds, I is an interpretation
function that maps each primitive proposition to a set of worlds, and J is an
interpretation function that maps each primitive principal to a binary relation
over W .Weextend J to a function J over arbitrary principal expressions as
follows:
M
=
W, I, J
J ( A )= J ( A )
J ( P & Q )= J ( P )
J ( Q )
J ( P
Q )= J ( P )
J ( Q )
|
J ( P )
J ( Q )
=
{
( w 1 ,w 3 )
|∃
w 2 . ( w 1 ,w 2 )
( w 2 ,w 3 )
}
.
Search WWH ::




Custom Search