MD CT C2 E02 Solutions
MD CT C2 E02 Solutions
m(t) = Ac cos(ωm t + ϕ)
ωc + n × ωm
for every integer n, where ωc is the frequency of the carrier and ωm is the frequency of the
modulating signal (in both cases they are angular frequencies in rad/s). In this case ωc = 2πfc
and ωm = 2πfm . The amplitude of each delta is π times the coefficient Jn (β), in this case Jn (5).
The number of harmonics in the effective bandwidth is Me = 2⌊β⌋ + 3 = 13 harmonics. Taking
into account that
J0 (5) = −0.18, J1 (5) = −0.33, J2 (5) = 0.05
J3 (5) = 0.37, J4 (5) = 0.39, J5 (5) = 0.26, J6 (5) = 0.13 · · ·
and that J−m (β) = (−1)m Jm (β), the Fourier transform of the modulated signal will have the form
of Figure 2.1 (Only the Me = 2⌊β⌋ + 3 = 13 central frequencies are represented, which contain
the effective bandwidth).
a) The bandwidth is 2B Hz. The power spectral density, for a generic modulation index a, is that
of the figure
S(jω) +0.26Ac π
+0.15Ac π +0.15Ac π
+0.05Ac π +0.05Ac π
··· ···
ωc − 5ωm ωc − 3ωm ωc ωc + ωm
−0.18Ac π
−0.26Ac π
−0.33Ac π
−0.37Ac π
Figure 2.1: Frequency response of the modulated signal for Exercise 2.2.
SS (jω)
A2c
2 π 2
aAc
........ 6
... AM ........ 6
...
..... ... ..... ...
2CM
.. ........ ..
.. .. ........ ..
..
...... .. ...... ..
..
.. ................ ..
..
..
.. ................ ..
..
. .
−ωc − W −ωc −ωc + W +ωc − W +ωc +ωc + W ω
b) The bandwidth is 2B Hz. The power spectral density is that of the figure
SS (jω)
A2c
......... .. AM ......... ..
...... ... ...... ...
4
.. ........ .
... .. ........ .
...
.. ...... ....... .. .. ...... ....... ..
.. .......... .. .. .......... ..
−ωc − W −ωc −ωc + W +ωc − W +ωc +ωc + W ω
c) The bandwidth is B Hz. The power spectral density is that of the figure
SS (jω)
In all cases PM = 21 is considered, which is the power of the modulating signal, CM = 1 the
maximum value in module of the signal m(t), and W = 2πfm .
SS (jω)
A2c
2 π
A2c
6 AM 4 a2 6
6 6 6 6
A2c
b) The power of the modulated signal is PS = 2 M
P . The power spectral density is that of the
figure, where AM = π2
SS (jω)
A2c
AM 4
6 6 6 6
c) The power of the modulated signal is PS = A2c PM . The power spectral density is that of the
figure, where AM = π2
SS (jω)
AM A2c
6 6
a) The bandwidth is 2B Hz, and the signal-to-noise ratio is worse than that obtained in a baseband
transmission
S S
= ηAM
N AM N b
with 2
a
P Ma CM2 P MM P MM
ηAM = = 2 = 2
1 + PM a 1 + Ca2 P MM CM
+ P MM
M a2
b) The bandwidth is 2B Hz, and the signal-to-noise ratio is the same as that obtained in a
baseband transmission
S S
=η
N DBL N b
c) The bandwidth is B Hz, and the signal-to-noise ratio is the same as that obtained in a baseband
transmission
S S
=η
N BLU N b
d) The bandwidth is B + ∆B Hz, where ∆B is the excess of bandwidth due to the vestige, and
the signal-to-noise ratio is slightly lower than that obtained in a baseband transmission
S S
= ηBLV
N BLV N b
with
B
ηBLV =
B + ∆B
e) The bandwidth is approximately 8B Hz, and the signal-to-noise ratio is better than that ob-
tained in a baseband transmission, the gain being proportional to the square of the modulation
index
S PM 2 S
≈ 3 2 βf
N FM CM N b
Exercise 2.6 Solution
cos(ωc t + ϕ)
In a coherent receiver, the phase of the carrier at the receiver must coincide with the phase of
the carrier that was used in the modulator to generate the signal, c(t) = Ac cos(ωc t + ϕc ), that
is, it must be true that ϕ = ϕc . Normally, to achieve this synchronization of the phases, a pilot
signal can be transmitted at the frequency and phase of the carrier, or a phase-locked loop, or
PLL, will be used, in the receiver.
Conventional AM modulation, which can also make use of a rocket receiver, can use a simpler
receiver, in this case an envelope detector, which can be implemented with a rectifier diode
and a low-pass RC filter, as shown in the figure.
r(t) R C d(t)
c) In the case of double sideband modulation, the effect of having a different phase in the carriers
of the transmitter (ϕc ) and the receiver (ϕ), will be that the ideally demodulated signal (without
noise or distortions) will have the expression
Ac
d(t) = m(t) cos(ϕc − ϕ),
2
with which the phase difference between carriers will generate an attenuation in the received
signal depending on the phase difference. It can be verified, for example, that a phase shift of
90º would be critical, since in this case the received signal would be completely cancelled.
In the case of single sideband modulation (and vestigial sideband, which is very similar), the
problem is even greater, since the expression of the signal at the output of the demodulator is
Ac Ac
d(t) = m(t) cos(ϕc − ϕ) ± m̂(t) sin(ϕc − ϕ),
2 2
whereby a phase difference will not only attenuate the term associated with the signal m(t),
but will make the term not null proportional to the Hilbert transform of the modulating signal,
m̂(t), which is an interference on the wanted signal m(t).
a)
(
0, |ω| > ωc
SS (jω) = 2
Ac [SM (jω − jωc ) + SM (jω + jωc )], |ω| < ωc
SS (jω)
A2c AM
ω (rad/s)
−ωc −ωc + W 0 ωc − W ωc
b) In terms of spectral efficiency, the SSB modulation is the most efficient of all, since the band-
width required to transmit a modulating signal of bandwidth B Hz is B Hz, while both the
conventional AM and the double side band AM require twice this bandwidth, 2B Hz, while the
FM modulation requires a larger bandwidth, dependent on the modulation index, according
to the modulation index, according to the rule, according to 2(β + 1)B Hz.
a) The bandwidth is
BWAM = 2B = 10 kHz.
The modulated signal in the time domain will have the form
1.5 ................ ......... ....
... ........ ..
............ ...... .............. . .
............ ...
... ..... ..... .. ... . .. ... .. ...
... ..... ......... ......... ....... ........ ....... ........... ......... ....... ....
1 .. . .. ... ... .
.. .. .. . .. . . .. .. . .
.. .. .. .
... .. ... ..... .......
. . . .. . .......... ........ ... ..... ... .... ... .... ........ ............ .......... ........ ... ..... ...
... . ... . .. .... ....... . . . . . . .
. . ..... .... .. .. .. .. .. .. . .. .. .. ..... ..... ..... .... .. .. .. .. ..
0.5 .... ... .... .. .... .. ... ............................... .. .... .. .... .. .... .. .... ... .... .. .... .. ... ................................ .. .... .. .... .. .... ..
... .. ... .. ... .. .... .. ... ..... .. ... .. ... .. ... .. ... .. ... .. ... .. ... .. .... .. ... ..... .. ... .. ... .. ... .. ... ..
... ... ... ... ... ... ... ... ... ... .... ... .... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .... ... ... ... ... ... ... ... ... ...
.
0 .... .. .... .. .... .. .... .. .... .. .... .. .... .. .... .. .... .. .... .. .... .. .... .. .... .. .... .. .... .. .... .. .... .. .... .. .... .. .... ..
... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ...
... .. ... .. ... .. ... ... ..... ..... ... .. ... .. ... .. ... .. ... .. ... .. ... .. ... ... ..... ..... ... .. ... .. ... .. ... ..
. .
-0.5 .... ... .... .. ... .. ...... ... .. ...... .... .. .... .. .... .. .... ... .... .. ... .. ...... ... .. ...... .... .. .... .. .... ..
... .. ... ... ...... ... ... ..... ... .. ... .. .. . ... .. ..... ... ... ..... ... .. ... ..
... .. ..... ...... ..... ... .. ... .. ... .. ... .. ..... ..... ... .. ... ..
. . .. . .. .
. .
. . .
. .
. .
. ... .
.
. ... ...... ......
-1 ..... .... ............... s(t) . ... .... ... ....
. .. .... ...
.
..... .... ..
.. ... .. ... .
.... . ............... 1 + a · m(t) .. ...... ....... .. ... .......
.. . ...
-1.5 .
0 0.5 1 1.5 2 2.5 3 3.5 4
t(×10−4 )
b) Amplitude modulations that do not include carrier transmission are
The bandwidth of the DSB modulation is the same as that of conventional AM, while that
of SSB is half, and that of VSB is approximately half (neglecting the vestige of the removed
sideband).
c) The theoretical bandwidth is infinite, since the Fourier transform can be written as an infinite
sum of sinusoids at frequencies ωc ± n ωm rad/s. The effective bandwidth is the bandwidth
that contains 98% of the signal power, and in this case it is
BF M = 2 (βf + 1) fm = 80 kHz.
d) The main advantage of angular modulations is their greater signal-to-noise ratio (noise immu-
nity), which increases with respect to the ratio in baseband proportionally to the square of
the modulation index. The main drawback is that the width band increases proportionally to
twice the rate of modulation plus one, which is significantly higher than that required by linear
modulations.
a) Amplitude modulations differ from angle modulations in that in the former what varies with
the information or modulating signal is the amplitude of the signal, while in the latter the
amplitude remains constant, but the angular information (instantaneous frequency or phase
term) changes. Therefore, the amplitude modulations are those of figures B and C. The
analytical expressions for conventional AM modulation and double sideband modulation are,
respectively
sAM (t) = [1 + a × ma (t)] × c(t), sDBL (t) = m(t) × c(t).
In the first, the envelope of the signal is [1 + a × ma (t)], which for a < 1, as is the case, is
always positive, while in the second, the modulating signal is simply multiplied by the carrier,
which makes the signal invert relative to the carrier for negative values of m(t). It is trivial to
see that
Regarding the angular modulations, the PM and FM modulations have the following analytical
expression
(
kp × m(t), for PM
s(t) = Ac cos(ωc t + ϕ(t)) with ϕ(t) = Rt .
2πkf × −∞ m(τ ) dτ, for FM
Therefore, the phase term ϕ(t) is proportional in one case to the modulating signal, and in
another case to its integral. In any case, when ϕ(t) takes positive values, the sinusoid argument
is increased, which means that the modulated signal “lead ” relative to the carrier, while when
it takes negative values, the sinusoid argument is decreased, which means that the modulated
signal “lag” relative to the carrier. Therefore, in a PM modulation, the signal will lead the
carrier for positive values of m(t), and it will lag for negative values, which happens in signal A.
In an FM modulation, whether it leads or lags the carrier is related to the sign of the integral
of m(t), which happens in figure D. Therefore
• Signal A: PM modulation.
• Signal D: FM modulation.
i) In this case it is best to use an angular modulation with a high modulation index, since
these have a better signal-to-noise ratio than angular modulations, a ratio that is propor-
tional to β 2 . Among the two variants, for the same value of β (same bandwidth), FM
modulation has a better signal-to-noise ratio (3 times that of PM).
ii) The simplest receiver that can be used in analog modulation is an envelope detector, which
can be implemented with a diode, and an RC filter. The modulation that this receiver can
use is conventional AM modulation. All other amplitude modulations must use a coherent
receiver, and angular modulations use more complex receivers.
iii) In this case, the smaller the bandwidth of the modulated signal, the more signals can be
multiplexed, and the modulation that has a smaller bandwidth is single-sideband mod-
ulation, which requires the same bandwidth as the modulating signal. The rest of the
modulations use a greater bandwidth.
• A double sideband signal is generated (double amplitude, in the notation that was followed
in the subject, but this amplitude factor is not really relevant) multiplying the modulator
by the carrier signal.
• This double-sideband signal is filtered with a vestigial sideband filter, a filter that has to
meet certain conditions (next section).
b) The condition that a vestigial sideband filter must satisfy is that its frequency response has
an odd symmetry with respect to the carrier frequency, ωc , so that the contribution of the
filter response shifted ωc to the left plus the same response shifted ωc to the right (effect of a
demodulator) is constant in the bandwidth of the signal
The only filter that satisfies the condition is filter A. In this case, it would be a filter for
an upper sideband vestigial sideband modulation, since the frequency band above the carrier
frequency (plus the corresponding lower sideband vestige) is allowed to pass.
SM (jω)
2AM
−W 0 +W ω
S PX 2 × 10−14
(dB) = 10 log10 = 10 log10 = 10 log10 1000 = 30 dB
N PZ 2 × 10−17
A2c
SS (jω) = [SM (jω − jωc ) + SM (jω + jωc )] ,
4
that is, two replicates of the power spectral density shifted ±ωc , with a scale factor in the
2
amplitude A4c , giving rise to the density shown in the figure
SS (jω)
AM
2
The signal-to-noise ratio using this type of modulation is equal to that obtained in a
baseband transmission, which was calculated in the previous section, that is, 30 dB.
that is, the part of the corresponding sideband of two replicates power spectral density
shifted ±ωc as shown in the figure
2AM
The signal to noise ratio using this type of modulation is equal to that obtained in a
baseband transmission, which was calculated in the first section, 30 dB.
Exercise 2.14 Solution
The power spectral density of the modulating signal is that of the figure
SM (jω)
AM
−W 0 +W ω
SS (jω) π
2
AM
16
SS (jω)
AM
4
that is, the part of the corresponding sideband of two replicates power spectral density shifted
±ωc as shown in the figure
SS (jω)
AM
ωini = ωc − ∆W
ωf in = ωc + ∆W
ωini ωc ωf in ωc + W
ii) The power spectral density of the modulated signal, using the example filter, would be
that of the figure. The dashed line superimposes the shape of the power spectral density
for a single sideband modulation. Only positive frequencies are represented.
ωini = ωc − ∆W
ωf in = ωc + ∆W
ωini ωc ωf in ωc + W
i) For narrow band modulations, the power spectral density is approximately equal to that
of a conventional AM modulation but with modulation index a = 1, which would be the
same as part (a) but with an amplitude A4M for the replicas of the spectrum.
ii) In this case the bandwidth is approximated by Carson’s Rule, which says that
a) The power spectral density of the modulating signal is that of the figure
SM (jω)
AM
−10 0 +10 ω
2π (kHz)
B: Double sideband amplitude modulation (without carrier) with carrier frequency fc = 100
kHz.
– The form of the D.E.P. of the modulator centered on the carrier frequency. Analyti-
cally
A2c
SS (jω) = [SM (jω − jωc ) + SM (jω + jωc )] ,
4
C: Conventional amplitude modulation (double sideband with carrier), with carrier frequency
fc = 90 kHz.
– The shape of the baseband power spectral density appears, scaled taking into ac-
count the modulation index a, centered on the carrier frequency, where the deltas
corresponding to the spectrum of the carrier itself appear. Analytically
A2c A2c a2
SS (jω) = π [δ(ω − ωc ) + δ(ω − ωc )] + 2
[SM (jω − jωc ) + SM (jω + ȷωc )]
2 4 CM
D: Upper sideband vestigial sideband amplitude modulation with carrier frequency fc = 100
kHz.
– The upper sideband of the D.E.P. appears. of the modulator (corresponding to the
positive frequencies) centered on the carrier frequency, without a trace in its initial
part, and a trace of the lower sideband remains, so that the spectrum has an odd
symmetry around the carrier frequency (complementing the trace that remains of the
lower sideband with the one that remains in the upper sideband).
+1 m(t)
t
−1
m(t)
s(t) = Ac [1 + a mn (t)] cos(ωc t), with mn (t) =
cM
where cM is the dynamic range of the modulating signal (cM = max(|m(t)|)).
π
SS (jω) 2
AM
16
AM
32
ω
990 995 1000 1005 1010 2π (kHz)
b) The power spectral density of double sideband modulation is as shown in the figure
AM SS (jω)
4
AM
8
ω
990 995 1000 1005 1010 2π (kHz)
SS (jω)
AM
AM
2
ω
990 995 1000 1005 1010 2π (kHz)
d) The vestigial sideband filter has to have odd symmetry about the carrier frequency, to satisfy
that
|HBLV (jω − jωc ) + HBLV (jω + jω)| = C, in |ω| ≤ 2πB
In this case, it must have odd symmetry with respect to fc = 1 MHz, and considering that it is
a lower sideband, the response must be constant between 990 and 999 kHz, with the transition
from 1 to 0 with odd symmetry from 999 to 1001 kHz. A possible example would be
HBLV (jω)
1
ω
990 995 1000 1005 1010 2π (kHz)
BF M ≈= 80 kHz.
S
(dB) = 34.31 dB.
N FM
ii) PM modulation with modulation index β = 5
BP M ≈ 120 kHz.
S
(dB) = 33.98 dB.
N PM