3.3 The One-Sided Laplace Transform 177
Solution
Even thoughδ(t)is not a regular signal, its Laplace transform can be easily obtained:
L[δ(t)]=
∫∞
−∞
δ(t)e−stdt=
∫∞
−∞
δ(t)e−s^0 dt=
∫∞
−∞
δ(t)dt= 1
Since there are no conditions for the integral to exist, we say thatL[δ(t)]=1 exists for all values
ofs, or that its ROC is the wholes-plane. This is also indicated by the fact thatL[δ(t)]=1 has no
poles.
The Laplace transform ofu(t)can be found as
U(s)=L[u(t)]=
∫∞
−∞
u(t)e−stdt=
∫∞
0
e−stdt=
∫∞
0
e−σte−jtdt
where we replaced the variables=σ+j. Using Euler’s equation, the above equation becomes
U(s)=
∫∞
0
e−σt[cos(t)−jsin(t)]dt
and since the sine and the cosine are bound, then we need to find a value forσso that the expo-
nentiale−σtdoes not grow astincreases. Ifσ <0, the exponentiale−σtfort≥0 will grow and
the integral will not converge. On the other hand, ifσ >0, the integral will converge ase−σtfor
t≥0 decays, and it is not clear what happens whenσ=0. Thus, the integral exists in the region
defined byσ >0 and all frequencies−∞< <∞(the frequency values do not interfere in the
convergence). Such a region is the open right-hands-plane, and is called the ROC ofU(s).
In the region of convergence, the integral is found to be
U(s)=
e−st
−s
|∞t= 0 =
1
s
where the limit fort=∞is zero sinceσ >0. So the Laplace transformU(s)= 1 /sconverges in
the region defined by{(σ,):σ >0,−∞< <∞}, or the open (i.e., thejaxis is not included)
right-hands-plane. This ROC can also be obtained by considering that the pole ofU(s)is ats= 0
and thatu(t)is casual.
We can find the Laplace transform of signals using symbolic computations in MATLAB. For the
unit-step and the delta functions, once the symbolic parameters are defined, the MATLAB function
laplacecomputes their Laplace transforms as indicated by the following script.
%%%%%%%%%%%%%%%%%
% Example 3.2
%%%%%%%%%%%%%%%%%
syms t s
% Unit-step function