Digital Signal Processing Reference
In-Depth Information
Chapter 5
Other Important Random Variables
5.1 Lognormal Random Variable
5.1.1 Density Function
A
lognormal variable X
is obtained through the exponential transformation of a
normal random variable
Y
that has a mean value of
m
Y
and a variance of
s
Y
,
X ¼
e
Y
;
(5.1)
or
x ¼
e
y
;
x>
0
;
(5.2)
where
x
and
y
are ranges of the variables
X
and
Y
, respectively.
From (
5.2
), we have only one solution:
y ¼
ln
x:
(5.3)
The relation (
5.3
) explains why the name “lognormal” comes for the variable
X
.
y ¼
ln
x
f
Y
ðyÞ
f
X
ðxÞ¼
:
(5.4)
dx
j
dy
ðyÞj
The derivative of the transformation (
5.2
) is:
d
x
d
y
¼
e
y
¼ x;
x>
0
:
(5.5)
Search WWH ::
Custom Search