2.

PROPERTIES OF TRANSFER FUNCTIONS

The Impulse Response function
A transfer function can be characterised by its effect on certain elementary reference signals. The
simplest of these is the impulse sequence defined by

1, if t = 0;
δt = (2.1)
0, if t = 0.
Its z-transform is δ(z) = 1. The output engendered by the impulse is described as the im-
pulse response function. The impulse response of the rational function β(z)/α(z) is the sequence
{ω0 , ω1 , . . .} of the coefficients of its series expansion
Another common reference signal is the unit step defined by

1, if t ≥ 0;
st = (2.2)
0, if t < 0.
The step response is the sequence of the partial sums of the impulse response. The sequence
converges to
β(1) β0 + β1 + · · · + βq
γ = ω(1) = = , (2.3)
α(1) α0 + α1 + · · · + αp
which is described as the steady-state gain of the transfer function.

1
 D.S.G. POLLOCK: Lectures in Lodz

1
0.75
0.5
0.25
0
−0.25
−0.5
−0.75

0 10 20 30

Figure 1. The impulse response of the transfer function b(z)/a(z) with a(z) = 1.0−0.673z +
0.463z 2 + 0.486z 3 and b(z) = 1.0 + 0.208z + 0.360z 2 .
 D.S.G. POLLOCK: Lectures in Lodz

Stability of a Transfer Function

The stability of a rational transfer function β(z)/α(z) can be investigated using its partial fraction
decomposition. Assume that β(z)/α(z) is a proper rational function in which the denominator is
factorised as
r
α(z) = (1 − z/λj )nj , (2.4)
j=1

where nj is the multiplicity of the root λj , and j nj = p is the degree of the polynomial. Then,

β(z)  
r nj
cj,k
= ; (2.5)
α(z) j=1 (1 − z/λj )k
k=1

and the task is to find the series expansions of the partial fractions.
Consider a partial fraction that contains a distinct (unrepeated) real root λ. The expansion is

c
= c{1 + z/λ + (z/λ)2 + · · ·}. (2.6)
1 − z/λ

For this to converge for all |z| ≤ 1, it is necessary and sufficient that |λ| > 1; and this is necessary
and sufficient for the satisfaction of the bounded-input bounded-output (BIBO) stability condition.

2
 D.S.G. POLLOCK: Lectures in Lodz

When α(z) has conjugate complex roots λ and λ∗ , the partial fraction has a quadratic denominator:

gz + d c c∗
= + . (2.7)
(1 − z/λ)(1 − z/λ∗ ) (1 − z/λ) (1 − z/λ∗ )

Then c = (gλ + d)/(1 − λ/λ∗ ) and c∗ = (gλ∗ + d)/(1 − λ∗ /λ) are also conjugate complex numbers.
The expansion of (2.6) applies to complex roots as well as to real roots:

c c∗   ∗
 ∗ ∗ 2

+ = c 1 + z/λ + (z/λ) 2
+ · · · + c 1 + z/λ + (z/λ ) + · · ·
1 − z/λ 1 − z/λ∗
∞ (2.8)
= z t (cλ−t + c∗ λ∗−t ).
t=0

The complex quantities can be represented in terms of exponentials:

λ = κ−1 e−iω , c = ρe−iθ , λ∗ = κ−1 eiω , c∗ = ρeiθ . (2.9)

Then, the generic term in the expansion becomes

t −t ∗ ∗−t t t
 i(ωt−θ) −i(ωt−θ)

z (cλ +c λ ) = z ρκ e +e
(2.10)
= z 2ρκ cos(ωt − θ).
t t

3
 D.S.G. POLLOCK: Lectures in Lodz

The expansion converges for all |z| ≤ 1 if and only if |κ| < 1, which is a condition on the complex
modulus of κ. But, |κ| = |λ−1 | = |λ|−1 ; so it is confirmed that the necessary and sufficient condition
for convergence is that |λ| > 1.

In the case of a repeated root with a multiplicity of n, a binomial expansion is available that gives

1 z n(n + 1)  z 2 n(n + 1)(n + 2)  z 3

=1+n + + + ···. (2.11)
(1 − z/λ)n λ 2! λ 3! λ

If λ is real, then |λ| > 1 is the condition for convergence. If λ is complex, then it can be combined
with the conjugate roots in the manner of (2.10) to create a trigonometric function and, again, the
condition for convergence is that |λ| > 1.

The general conclusion is that the transfer function is stable if and only if all of the roots of the
denominator polynomial a(z), which are described as the poles of the transfer function, lie outside
the unit circle in the complex plane.

It helps to represent the poles (denominator roots) and zeros (numerator roots) of the transfer
function graphically by showing their locations within the complex plane. It is more convenient to
represent the poles and zeros of β(z −1 )/α(z −1 ), which are the reciprocals of those of β(z)/α(z),
since, for a stable and invertible transfer function, these must lie within the unit circle.

4
 D.S.G. POLLOCK: Lectures in Lodz

Im
i

Re
−1 1

−i

Figure 2. The pole–zero diagram for the transfer function b(z −1 )/a(z −1 ) corresponding to
the impulse response function of Figure 1. The poles are marked by crosses and the zeros by
circles.
 D.S.G. POLLOCK: Lectures in Lodz

The Response to a Sinusoidal Input

Consider mapping the signal sequence {xt = cos(ωt); t = 0, ±1, ±2, . . .} through the transfer func-
tion with the coefficients ψ0 , ψ1 , . . . , ψq . The output is


q

y(t) = ψj cos ω[t − j] . (2.12)

j=0

The trigonometrical identity cos(A − B) = cos A cos B + sin A sin B enables us to write this as
   
y(t) = ψj cos(ωj) cos(ωt) + ψj sin(ωj) sin(ωt)
j j (2.13)
= α cos(ωt) + β sin(ωt) = ρ cos(ωt − θ).

Here we have defined


q 
q
β 
−1
α= ψj cos(ωj), β= ψj sin(ωj), and ρ = α2 + β 2 , θ = tan . (2.14)
j=0 j=0
α

There is a gain effect whereby the amplitude of the sinusoid is increased or diminished by a factor
of ρ and there is a phase effect whereby the peak of the sinusoid is displaced by a time delay of θ/ω
periods. The frequency of the output is the same as the frequency of the input.

5
 D.S.G. POLLOCK: Lectures in Lodz

Aliasing of Sampled Processes

In a discrete-time system, there is a problem of aliasing whereby frequencies (i.e. angular velocities)
in excess of π radians per sampling interval are confounded with frequencies within the interval
[0, π].

Consider the case of a pure cosine wave of unit amplitude and zero phase whose frequency ω lies
in the interval π < ω < 2π. Let ω ∗ = 2π − ω. Then

 ∗

cos(ωt) = cos (2π − ω )t
= cos(2π) cos(ω ∗ t) + sin(2π) sin(ω ∗ t) (2.15)
= cos(ω ∗ t);

which indicates that ω and ω ∗ are observationally indistinguishable. Here, ω ∗ ∈ [0, π] is described
as the alias of ω > π.

Imagine that a person observes the sea level at 6am. and 6pm. each day. He should notice a very
gradual recession and advance of the water level; the frequency of the cycle being f = 1/28 which
amounts to one tide in 14 days. In fact, the true frequency is f = 1 − 1/28 which gives 27 tides in
14 days. Observing the sea level every six hours should enable him to infer the correct frequency.

6
 D.S.G. POLLOCK: Lectures in Lodz

1.0

0.5
1 2 3 4

−0.5

−1.0

Figure 5. The values of the function cos{(11/8)πt} coincide with those of its alias
cos{(5/8)πt} at the integer points {t = 0, ±1, ±2, . . .}.
 D.S.G. POLLOCK: Lectures in Lodz

Spectral Representation of a Stationary Stochastic Process

Any stationary stochastic process can be expressed as a weighted combination of the infinity of
sines and cosines of which the frequencies lie in Nyquist frequency interval [0, π]. Thus, if xt is an
element of such a process, then it can be represented by
 π
xt = cos(ωt)dA(ω) + sin(ωt)dB(ω) . (2.16)
0

Here, dA(ω) and dB(ω) are the infinitesimal increments of stochastic functions defined on the
frequency interval that are everywhere continuous but nowhere differentiable. It is assumed that
the increments A(ω) and B(ω) are uncorrelated with each other and with preceding and succeeding
increments.
Such processes can be representing the in terms of complex exponentials. Thus
eiθ + e−iθ −i iθ 1
cos θ = and sin θ = (e − e−iθ ) = (eiθ − e−iθ ). (2.17)
2 2 2i
These enable equation (2.16) to be written as
 π  iωt 
(e + e−iωt ) (eiωt − e−iωt )
xt = dA(ω) − i dB(ω)
0 2 2
 π  (2.18)
{dA(ω) − idB(ω)} {dA(ω) + idB(ω)}
= eiωt + e−iωt .
0 2 2

7
 D.S.G. POLLOCK: Lectures in Lodz

On defining
dA(ω) − idB(ω) dA(ω) + idB(ω)
dZ(ω) = and dZ ∗ (ω) =
2 2

and extending the integral over the range [−π, π], this becomes

 π
xt = eiωt dZ(ω), (2.19)
−π

which is commonly described as the spectral representation of the process generating yt .

The spectral density function of the process is the function f (ω) that is defined by

E{dZ(ω)dZ ∗ (ω)} = E{dA2 (ω) + dB 2 (ω)} = f (ω)dω. (2.20)

The increment f (ω)dω is the power or the variance of the process that is attributable to the elements
in the frequency interval [ω, ω + dω]; and the integral of f (ω) over the frequency range [−π, π] is
the overall variance of the process. A white noise process {εt } with a variance of σε2 has a uniform
spectral density function fε (ω) = σε2 /(2π).

8
 D.S.G. POLLOCK: Lectures in Lodz

The Frequency Response

Let {ψ0 , ψ1 , . . .} be the impulse response of the transfer function. Then, the effects of the transfer
function upon the spectral elements of the process defined by (2.19) are shown by the equation
   
yt = μj x(t − j) = μj eiω(t−j) dZx (ω)
j j ω
   (2.21)
= eiωt μj e−iωj dZx (ω).
ω j

The effects are summarised by the complex-valued frequency-response function


μ(ω) = μj e−iωj = |μ(ω)|e−iθ(ω) . (2.22)

The final expression entails the filter gain |μ(ω)| and the phase response θ(ω):

∞ 2  
∞ 2
|μ(ω)|2 = μj cos(ωj) + μj sin(ωj)
j=0 j=0
  (2.23)
μj sin(ωj)
θ(ω) = arctan  .
μj cos(ωj)

9
 D.S.G. POLLOCK: Lectures in Lodz

0
−π −π/2 0 π/2 π

Figure 3. The gain of the transfer function depicted in Figures 1 and 2.

 D.S.G. POLLOCK: Lectures in Lodz

−π

−π −π/2 0 π/2 π

Figure 4. The phase plot to accompany Figure 3.

 D.S.G. POLLOCK: Lectures in Lodz

The frequency response is just the discrete-time

 Fourier transform of the impulse response function.
Equally, it is the z-transform ψ(z −1 ) = j ψj z −1 evaluated at the points eiω that lie on the unit
circle in the complex plane.
As ω progresses from −π to π, or, equally, as z = eiω travels around the unit circle, the frequency-
response function defines a trajectory in the complex plane, which becomes a closed contour when
ω reaches π. The points on the trajectory are characterised by their polar co-ordinates. These are
the modulus |ψ(ω)|, which is the length of the radius vector joining ψ(ω) to the origin, and the
argument Arg{ψ(ω)} = −θ(ω) which is the (anticlockwise) angle in radians which the radius makes
with the positive real axis.
The spectral density of the output function fy (ω) of the filtered process y(t) is given by

fy (ω)dω = E{dZy (ω)dZy∗ (ω)}

= μ(ω)μ∗ (ω)E{dZx (ω)dZx∗ (ω)} (2.24)
= |μ(ω)|2 fx (ω)dω.

10
 D.S.G. POLLOCK: Lectures in Lodz

Im

Re

Figure 5. The path described in the complex plane by the frequency response function
corresponding to the gain and phase functions of Figures 3 and 4. The trajectory originates,
when ω = 0, in the point on the real axis marked by a dot and it travels in the direction of
the arrow, of which the tip is reached when ω = π/4.