38
|
2 Fundamentals of Information, Signal and System Theory
transformability can be achieved by an additional convergence term. This leads us to
the Laplace transform.
2.4.2 Continuous Laplace Transform
The Laplace¹⁵ transformation also belongs to the class of integral transformations and
is an extension of the Fourier transformation. The integral core is similar to that of the
Fourier transform (cf. Equation 2.49):
K(ξ, t) = K(p, t) = e−pt
(2.55)
where p is a complex variable,
p = σ + jω .
(2.56)
Here σ is a real number and j is the imaginary unit. Equation 2.55 inserted into Equa-
tion 2.48 yields the mathematical operation of the Laplace transformation:
S(p) = ∫s(t)e−ptdt .
(2.57)
If we separate the variable p in Equation 2.57 into its components according to Equa-
tion 2.56 the similarity of the Laplace transform with the Fourier transform becomes
even clearer:
S(σ, jω) = ∫s(t)e−jωte−σtdt .
(2.58)
The term
e−σt
(2.59)
corresponds to an additional damping term, whereby also signals become transform-
able which do not satisfy the Dirichlet condition according to Equation 2.54. Thus
Equation 2.54 for the Laplace transform can be extended to
∫|s(t)e−σt|dt < ∞.
(2.60)
The damping effect of Equation 2.59 occurs only when the argument of the exponen-
tial function is negative overall. Since we are dealing with causal signals in signal
processing (s(t) = 0, ∀t < 0), so the negative time domain is not considered, σ must
be > 0. This region in the complex p-plane is called the convergence region of the
Laplace transform. There, polynominals with arbitrary power of t are transformable.
For exponentially increasing signals of type s(t) = u(t)eat with a > 0, one further re-
strictions for the convergence region of σ have to be considered and the dampening
15 Pierre-Simon Laplace (1749–1827), french mathematician, physicist and astronomer, Minister of the
Interior under Napoleon.