Cryptography Reference
In-Depth Information
a
()
0
t
Internal state
⊕
Output
filter
out
(t
)
b
()
15
t
Buffer b
b
()
0
t
State a
()
t
b
⊕
⊕
ρ
λ
b
+
(1)
t
t
b
+
( )
15
b
( )
0
t
+
Update function
(a) Structure of MUGI
(b)
λ
-function
Buffer
64
a
()
0
t
a
()
1
t
a
()
2
t
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
S
S
S
S
S
S
S
S
64
64
64
64
b
()
4
t
M
64
17
()
10
t
b
<<<
F
⊕
64
64
⊕
F
⊕
M
C
C
⊕
⊕
64
2
64
a
( )
0
t
+
a
( )
1
t
+
a
( )
2
t
+
F-function
(c)
ρ
-function
(d) F-function
Fig. 1.
Structure of MUGI ,
λ
-function,
ρ
-function, and F-function
element
b
i
is 64 bits. The update function is described as a combination of
ρ
and
λ
functions and the update functions of state
a
and buffer
b
each of which
uses another internal state as a parameter. The update function
Update
,ofthe
entire internal state is expressed as
(
a
(
t
+1)
,b
(
t
+1)
)=
Update
(
a
(
t
)
,b
(
t
)
)=(
ρ
(
a
(
t
)
,b
(
t
)
)
,λ
(
a
(
t
)
,b
(
t
)
))
.
Here, we call a step in which the update function is applied a round. In the
following, we review the structure of
ρ
and
λ
, initialization, and the output
filter.
λ Function.
λ
is the update function of buffer
b
and uses a part of state
a
as
a parameter.
λ
is a linear transformation of
b
as shown in Fig.1(b).
ρ Function.
ρ
is the update function of state
a
. It is a kind of generalized
Feistel structure with two F-functions and uses buffer
b
as a parameter. Figure
1(c) shows the structure of the
ρ
function where
C
1
and
C
2
are public constants.
The F-function of MUGI uses the same S-box table and linear function,
M
,as
those used in AES. Figure 1(d) shows the structure of the F-function in the
ρ
-
function where
S
and
M
denote the S-box table and a 4
×
4 matrix, respectively.
Search WWH ::
Custom Search