Below we will look at the four most common Hilbert spaces that you
will have to deal with when discussing and manipulating
signals and systems.
ℝn
n
(reals scalars) and
ℂn
n
(complex scalars), also called
ℓ
2
0n-1
ℓ
2
0
n
1
x=
x
0
x
1
…
x
n
-
1
x
x
0
x
1
…
x
n
-
1
is a list of numbers (finite sequence). The inner product for our
two spaces are as follows:
-
Inner product
ℝn
n
:
<x,y>=yTx=∑i=0n-1
x
i
y
i
x
y
y
x
i
0
n
1
x
i
y
i
(1)
-
Inner product
ℂn
n
:
<x,y>=yT¯x=∑i=0n-1
x
i
y
i
¯
x
y
y
x
i
0
n
1
x
i
y
i
(2)
Model for: Discrete time signals on the interval
0n-1
0
n
1
or periodic (with period
nn) discrete time signals.
x
0
x
1
…
x
n
-
1
x
0
x
1
…
x
n
-
1
f∈
L
2
ab
f
L
2
a
b
is a finite energy function on
ab
a
b
<f,g>=∫abftgt¯dt
f
g
t
a
b
f
t
g
t
(3)
Model for: continuous time signals on the interval
ab
a
b
or periodic (with period
T=b-a
T
b
a
) continuous time signals
x∈
ℓ
2
ℤ
x
ℓ
2
is an infinite sequence of numbers that's
square-summable
<x,y>=∑i=-∞∞xiyi¯
x
y
i
x
i
y
i
(4)
Model for: discrete time, non-periodic signals
f∈
L
2
ℝ
f
L
2
is a finite energy function on all of
ℝ.
<f,g>=∫-∞∞ftgt¯dt
f
g
t
f
t
g
t
(5)
Model for: continuous time, non-periodic signals
Each of these 4 Hilbert spaces has a type of Fourier analysis
associated with it.
-
L
2
ab
L
2
a
b
→ Fourier series
-
ℓ
2
0n-1
ℓ
2
0
n
1
→ Discrete Fourier Transform
-
L
2
ℝ
L
2
→ Fourier Transform
-
ℓ
2
ℤ
ℓ
2
→ Discrete Time Fourier Transform
But all 4 of these are based on the same principles (Hilbert space).
Not all normed spaces are Hilbert
spaces
For example:
L
1
(
ℝ
)
L
1
(
ℝ
)
,
∥f∥1=∫|ft|dt
1
f
t
f
t
. Try as you might, you can't find an inner product that
induces this norm,
i.e. a
<·,·>
·
·
such that
<f,f>=∫|ft|2dt2=∥f∥12
f
f
t
f
t
2
2
1
f
2
(6)
In fact, of all the
L
p
ℝ
L
p
spaces,
L
2
ℝ
L
2
is the
only one that is a Hilbert space.
Hilbert spaces are by far the nicest. If you use or study
orthonormal basis
expansion then you will start to see why this is true.
"My introduction to signal processing course at Rice University."