The Bilinear Transform is a nonlinear
C→C
→
mapping that maps a function of the complex variable
s
s to a function of a complex variable
z
z. This map has the property that the LHP in
s
s (
ℜs<0
s
0
) maps to the interior of the unit circle in
z
z, and the
iλ=s
λ
s
axis maps to the unit circle
eiω
ω
in
z
z.
Bilinear transform:
s=αz−1z+1
s
α
z
1
z
1
Hz=H
a
s=αz−1z+1
H
z
H
a
s
α
z
1
z
1
iλ=αeiω−1eiω+1=α(eiω−1)(e−(iω)+1)(eiω+1)(e−(iω)+1)=2isinω2+2cosω=iαtanω2
λ
α
ω
1
ω
1
α
ω
1
ω
1
ω
1
ω
1
2
ω
2
2
ω
α
ω
2
, so
λ≡αtanω2
λ
α
ω
2
,
ω≡2arctanλα
ω
2
λ
α
.
Figure 1.
The magnitude response doesn't change in the
mapping from λ
λ to ω
ω, it is simply warped nonlinearly according to
Hω=H
a
αtanω2
H
ω
H
a
α
ω
2
, Figure 2.
This mapping preserves
∥L∥∞
L
errors in (warped) frequency bands. Thus optimal
Cauer
(
∥L∥∞
L
) filters in the analog realm can be mapped to
∥L∥∞
L
optimal discrete-time IIR filters using the
bilinear transform! This is how IIR filters with
∥L∥∞
L
optimal magnitude responses are designed.
The parameter α
α provides one degree of freedom which can be used
to map a single
λ
0
λ
0
to any desired
ω
0
ω
0
:
λ
0
=αtan
ω
0
2
λ
0
α
ω
0
2
or
α=
λ
0
tan
ω
0
2
α
λ
0
ω
0
2
This can be used, for example, to map the pass-band edge of
a lowpass analog prototype filter to any desired pass-band
edge in ω
ω. Often, analog prototype filters will be designed with
λ=1
λ
1
as a band edge, and α
α will be used to locate the band edge in
ω
ω. Thus an
Mth
Mth
order optimal lowpass analog filter prototype can
be used to design any
Mth
Mth
order discrete-time lowpass IIR filter with the
same ripple specifications.