Information Technology Reference
In-Depth Information
A
E
1
→
1
⁄
a
•
(
a
B
+
d
C
)
A
;
(7)
B
E
1
→
1
⁄
a
•
1
⁄
c
B
•
1
⁄
b
B
•
(
a
A
++
d
C
b
C
)
B
;
C
E
1
→
1
⁄
d
•
1
⁄
b
C
•
(
b
B
++
c
B
a
A
)
C
.
A
E
2
→
a
B
•
1
⁄
a
B
+
a
A
;
B
E
2
→
a
A
•
1
⁄
a
A
+
b
C
•
1
⁄
b
C
+
d
C
+
a
B
+
b
B
(8)
;
C
E
2
→
b
B
•
1
⁄
b
B
++
c
B
b
C
.
N
n
(9)
A
E
3
→
(
a
B
•
1
⁄
a
B
•
1
⁄
x
B
)
•
1
⁄
t
B
++
x
A
t
A
;
N
n
B
E
3
→
(
a
A
•
1
⁄
a
A
•
1
⁄
x
A
)
•
1
⁄
t
A
+
…
N
n
…
+
(
b
C
•
1
⁄
b
C
•
1
⁄
x
C
)
•
1
⁄
t
C
+
x
B
+
t
B
;
N
n
C
E
3
→
(
b
B
•
1
⁄
b
B
•
1
⁄
x
B
)
•
1
⁄
t
B
++
x
C
t
C
.
B
E
1
⊗⊗
B
E
2
B
E
3
=
1
⁄
c
B
+
d
+
t
A
+
t
C
A
E
1
⊗⊗
A
E
2
A
E
3
=
d
C
+
t
B
C
;
;
(10)
C
E
1
⊗⊗
C
E
2
C
E
3
=
1
⁄
d
C
+
c
+
t
B
B
3.2 Fitness Function
We consider e-rule E4 together with node's fitness as a meta-rule that can be used to
modify etiquette rules to safely adapt to network situation or context, how the trust
establishment etiquette itself can be modified to progress trust. The rationale behind is
that of community leadership expressed in local preferences for decision making; if the
community is active, stable (conflict-free), and mutually trusted then the behaviour
choices of every node can be streamlined, and, on contrary, when community is inac-
tive, unstable or disappears, the behaviour choices of every node should follow largely
the principles of self-protection, survivability, etc. Natural metrics of community's
activity are the number of served workflows and the number of trusted or trust-in-
progress node pairs within the community. On the other hand, an autonomic node
should preserve certain balance between the amount of served workflows and the rela-
tive amount of community members producing/consuming these workflows. The ra-
tionale for this is to avoid DoS and DDoS scenarios, where relatively small fraction of
community is generating unrealitsically high volumes of workflows destined for a node.
We claim that an optimal fitness function can be found for a given communication
scenario, and given community, sets of etiquette rules, and concerns. Our goal is not
to find an optimal fitness function but to demonstrate how a fitness function can be
used as a mean for self-adaptation, in particular in autonomic adaptation of etiquette
rules to situation changes. For demonstration purposes only we shall use (11)
