Escolar Documentos
Profissional Documentos
Cultura Documentos
amplitude
6
-10
phase
1
0
-4
/4
f
-10
-4
-/4
-/2
f
10
phase
3/8
x2
amplitude
10
-3/8
am coefficients
1
0
0
1
0
1 1
0
1
0
11
00
1
0
1
0
0
00
1
0
0
00 0
11
0 1
1
1 0
0
0 1
1
1 1
0
011
1
00 1
11
0
0 1
1
4
2
5
6
8 9
3
0 1
7
bm coefficients
11
00
00
11
11
00
0
00
11
0
11
00
00
11
00
11
1
2
1
0
0
1
1
0
00
11
11
00
00
00
11
0 11
1
00
11
00
11
11
00
00
00 11
00
11
0011
11
00
11
4
5
6
8
3
7
1
0
0
1
0
1
9
Question 4 (1 point)
Energy and power
A voltage signal is given by: x(t) = 2 (r(t + 2) + r(t) 2r(t + 1) + r(t 6) r(t 4)).
Which of the following statements is correct?
a) x(t) is a power signal, its power equals 16 Watt
b) x(t) is an energy signal, its energy equals 12 Joule
c) x(t) is neither an energy nor a power signal
d) x(t) is a power signal, its power equals 8 Watt
Question 5 (1 point)
Systems, time domain
2
t1
R
x()d.
A system H is given by: y(t) d y2 3y(t) =
dt
-10
Xs (f )
spectrum I
200
-50
200
200
-30
-10
200
200
30
100
-30
-10
spectrum III
-30
50
Xs (f )
100
-50
10
spectrum II
-50
10
100
10
50
30
50
Xs (f )
100
-10
30
10
Question 8 (2 points)
Sampling
Given a signal x(t) = 10 cos(10t + /7) + 4 sin(30t + /8).
We sample x(t) at 11 Hz. When we consider the Fourier transform Xs (f ) of the (impulse-train)
sampled signal xs (t), and only consider the positive frequencies of Xs (f ), starting at f = 0,
then the following frequencies below 10 Hz are present. That is, we see peaks at:
a) 4 Hz, 5 Hz and 6 Hz
b) 5 Hz
c) 5 Hz and 6 Hz
d) 4 Hz, 5 Hz, 6 Hz and 7 Hz
Question 9 (1 point)
Quantization
We would like to quantize an analog voltage signal in the range from [0.0, 6.4] Volt, with a 0.2
Volt resolution. Using a binary representation, what is the required word-length? That is, how
many bits do we need for each sample?
a) 6
b) 5
c) 4
d) 3
Question 10 (2 points)
Signal detection - probabilities
For a signal detection problem it is given that observable y (on the received signal) is distributed
as y N(1, = 2) for a binary 0, and as y N(1, = 2) for a binary 1.
Using the best possible decision rule, what is the false alarm probability when the probability
of missed detection is 0.5636?
a) None of the other answers is correct
b) 0.0500
c) 0.2005
d) 0.1230
Question 11 (2 points)
Bit Error Probability
The navigation message of GPS is assumed taken from the carrier. A typical GPS receiver
signal power Px (when used outdoor) equals 158 dB W, and typical noise power Pe equals
187 dB W. The bit error probability P (E) can be computed with:
P (E) = 1
r
x
mP
Pe ,
with m the number of samples per bit. The receiver is used both outdoor as well as indoor.
When used indoor the signal attenuation factor is 250 (that is, the signal power is multiplied
1
with 250
); the noise power Pe remains unchanged.
We take one sample per bit and ask three distinguished professors to compute P (E) in both
situations. In the table below we summarize the end-result of their calculations.
calculation of professor I
calculation of professor II
calculation of professor III
One of the possible data links for ADS-B is the so-called Mode-S Extended Squitter, which
operates on a frequency of 1090 MHz.
In this question we consider the budget for an aircraft to ground station link (see figure above).
The transmitter on-board the aircraft has 300 W of transmission power available. The aircraft
is 120 nautical miles from the ground station antenna (1 nmi = 1852 m). Both the aircraft and
ground station have omni-directional antennae (isotropic).
For the aircraft transmitter, it is assumed that 10 percent of the signal power gets lost in the
cabling. It is also assumed that, once the signal gets at the ground station, 10 percent of the
signal power gets lost in the cabling of the ground station receiver.
The speed of light c can be approximated with 3 108 m/s.
The signal power that the ground station receiver gets from its antenna, expressed in dB, equals:
a) 135, 36 dB W
b) None of the other answers is correct
c) 116, 27 dB W
d) 125.12 dB W
Open Questions
Note For each question, please carefully show how you arrived at your result. Include comments
whenever possible, or needed, as if you are explaining what you are doing to a fellow student.
Question 13 (8 points)
Fourier Series
Consider the periodical signal x(t), defined by:
x(t) = 1 + 2 cos(2t) + 4 sin(6t /6).
a) What is the period T0 of this signal, in seconds? Explain your answer.
(2 points)
(2 points)
(4 points)
Question 14 (8 points)
Fourier transform
Consider the unit-pulse signal b(t) =
t
a) Prove that B(f ) = sinc( f ), see the formula sheet (Table 4-2, pair # 1).
Now, the triangular signal x(t) =
t
(1 point)
(3 points)
c) Prove this Fourier transform using the differentiation theorem (Table 4-1, property #6).(4 points)
10
Question 15 (8 points)
Consider polar signalling, with signal level A Volt and the bit duration Tb seconds. We assume
that, on average, the 0-s and 1-s each occur 50% of the time, and that all bits are independent.
a) Show that the average power of this polar baseband signal x(t) equals A2 .
(1 point)
A receiver receives this polar baseband signal, and samples each bit m times. The m measurements are collected in a vector y. Assume that the receiver and transmitter are perfectly
synchronized and that there is no attenuation of the signal power. Hence, the signal is received
as is, plus noise. The latter is reflected in the measurements standard deviation .
The probability of incorrect detection P (E) is given by:
, with = ||x x ||.
P (E) = 1 2
a
0
A m
.
(2 points)
We have to decide between a 0 (zero transmitted) for which the expected mean of y would be
x0 , and a 1 (one transmitted), for which the expected mean of y would be xa . We assume all
observables to be normally distributed and uncorrelated.
The SLR test will yield a decision rule that says:
decide that it is a 1 if
cT (y x0 )
, and a 0 otherwise, with c =
> 2
xa x0
.
c) Work out this decision rule for the polar baseband signal. Explain the result, i.e., does this
decision rule make sense?
(2 points)
It is often more realistic to assume that the transmitter puts the polar baseband signal x(t)
first onto a carrier wave c(t) = cos(2fc t), yielding the BPSK bandpass signal v(t): v(t) =
x(t) c(t).
d) Prove that the power of the BPSK bandpass signal equals A2 /2.
Hints: use the modulation theorem, assume that fc T1b when computing the PSD and then use
the definite integral on top of page 6 of the formula sheet.
(3 points)