6. • Exercises
– They are introduced during the contact hours.
– Exercises are typical for exam problems
• Assignments
– Groups will be formed before in week 1.6
– Group assignment will be issued in week 1.6
– Due dates are menNoned in the assignment text
– Assignment reports do yield credits
• Maneesh Verma deals with all assignments and
exercises for satellite communicaNons, for
satellite tracking Bart Root is your guide.
06/10/17 TU Del3 class ae3535-16 6
7. • Preference for MATLAB during exercises/assignments
• Structure of a report
• Explain the problem and the soluNon
• Include MATLAB results
• Explain task distribuNon in your group (if applicable)
• Explain how you validated your results?
• HandwriSen reports are fine as long as we can read them
• Copied and pasNng from lecture notes, the internet and other
sources is not allowed
• Convert your report to an unsigned PDF file
• Submit your reports to brightspace, there are individual
and groups folders for assignment or exercise, consult
your TA.
06/10/17 TU Del3 class ae3535-16 7
8. 06/10/17 TU Del3 class ae3535-16 8
Week Topics
1.1 IntroducNon to radio technology in spaceflight, Nme and
frequency domain, Fourier transformaNon, FFT
1.2 ProperNes of the Fourier transform, schemaNcs transmiSer
and receiver, superheterodyne receiver, intermodulaNon,
so3ware defined receiver (SDR)
1.3 LC networks, admiSance of inductors and capacitors, RC
constant, LC circuit, π filters, LCR circuits, Q factor,
transmission lines, antenna’s
1.4 Impedance matching, modulaNon, propagaNon, signal and
noise, link margin
1.5 Digital modulaNon, radio astronomy, GNSS
1.6 PracNcal
1.7 PracNcal
10. • Satellites, rockets, parts of rockets, etc,
all is nowadays tracked by radio
• Visual contact:
– maybe for several kilometer, and only for
verificaNon purposes
– In the past we relied on opNcal tracking
• Radio-range depends on:
– Frequencies used
– Antenna characterisNcs
– Transmit power and receiver sensisiNvity
– Antenna horizon and field of view
• CommunicaNon and navigaNon are
closely related, in fact, there is hardly
any difference
06/10/17 TU Del3 class ae3535-16 10
Baker-Nunn camera
S-band tracking system
17. Normally signals appear in the time domain:
v(t) with t ∈ [0,T]
where T is the length of a record. If we assume that:
v(t +T) = v(t)
then the function is said to be periodic. Furthermore
if v(t) has a finite number of oscillations in [0,T] then
we can develop v(t) in a series:
v(t) = Ai
i=0
N/2
∑ cos(ωit)+ Bi sin(ωit)
which is known as a Fourier series and where Ai and Bi
denote the Euler coefficients.
06/10/17 TU Del3 class ae3535-16 17
22. • EssenNally y=FFT(x) carries out a Fourier transform
• FFT algorithm input
– Real vector x(0..N-1) with N datasamples
– The record starts at 0 and is filled to N-1
• FFT algorithm output
– Euler coefficients are stored in the form of complex numbers
– Stored in y(i) are:
y(0) = A0 + I.B0 , y(1) = A1+ I.B1 , …, y(N/2) = AN/2 + I.BN/2
– I is a complex number:
– Be careful with scaling factors, check this always with a test funcNon of
which you now the Euler coefficients in advance
– Suitable test funcNons are for instance linear sin and cos expressions
• FFT algorithms exploit symmetries of sin and cos funcNons, please
use MATLAB because it is thoroughly debugged.
• Be aware that indices in MATLAB vectors start at 1 (and not 0)
I = −1
06/10/17 TU Del3 class ae3535-16 22
43. A F A F O
antenna
informaNon
A: amplifiers, F: band pass filters, O: local oscillator
“InformaNon” changes either the frequency, the phase
or the amplitude of the oscillator
The band pass filters are required to eliminate,
unwanted, parasiNc frequencies such as overtones
which are taboo on the output
Never operate a transmiSer without an antenna,
because it is likely to kill the output amplifier
The use of transmiSers is bound to regula2ons, only
certain frequencies, power, and bandwidth can be used
for designated applicaNons.
06/10/17 TU Del3 class ae3-535 43
Transmission
line
44. antenna
L L C
A A
SPKR
D
A: amplifier
D: demodulator
C
D
Input Output
The voltage level
a3er demodulaNon
proporNonal to amplitude
input signal
06/10/17 TU Del3 class ae3-535 44
72. R
C
SPST
VC +VR = 0 with C =
Q
V
and I =
dQ
dt
Q(t)
C
+ I(t)R = 0 ⇒
Q(t)
C
+
dQ(t)
dt
R = 0
Q(t) = −RC
dQ
dt
⇒ Q(t) = Q(t0 )e−t/τ
with τ=RC so that V(t) =V(t0 )e−t/τ
When a complex notation is used
R.ejωt
+
ejωt
jωC
= 0 ⇒ ω =
j
RC
06/10/17 TU Del3 class ae3-535 72
73. ZC =
1
jωC
and ZL = jωL
ZLC = ( 1
ZL
+ 1
ZC
)−1
= ∞ ⇒
ω =
1
LC
C L
I
known
parallel resistance
consequence
The only thing you need to know is the impedance of inductors and capacitors,
The parallel resistance should be infinite for a LC circuit without dissipaNon
In that case ω should be as given, it is maintained when there are no losses
06/10/17 TU Del3 class ae3-535 73
76. C
L R
Energy loss in a LCR circuit
Z−1
= ZL
−1
+ ZC
−1
+ R−1
To obtain the dissipation P:
V = Hejωt
so that T =
2π
ω
P =
1
T
V2
Z0
T
∫ dt =
H2
R
1+
(ω2
LC −1)2
R2
ω2
L2
P0 = P(ω0 ) =
H2
R
and ω0 =
1
LC
Search now the ω where P(ω0 ) =
1
2
2 P(ω)
since this results in the bandwidth
(ω2
LC −1)2
R2
=ω2
L2
⇔ (ω −ω0 )(ω +ω0 ) =
ω
RC
≈ ω0Δω
By definition Q =
ω
Δω
=ω0RC =
RC
LC
= R
C
L
(Quality factor)
06/10/17 TU Del3 class ae3-535 76
83. 06/10/17 TU Del3 class ae3-535 83
In a transmission line signal propagation is described by
the so-called telegraphers equations (Oliver Heaviside, 1880)
∂V(x,t)
∂x
= −L
∂I(x,t)
∂t
− R.I(x,t)
∂I(x,t)
∂x
= −C
∂V(x,t)
∂t
−G.V(x,t)
For an ideal line we can set R = 0 and G = 0, we find:
∂V(x,t)
∂x
= −L
∂I(x,t)
∂t
and
∂I(x,t)
∂x
= −C
∂V(x,t)
∂t
so that
∂2
V(x,t)
∂t2
−u2 ∂2
V(x,t)
∂x2
= 0 and
∂2
I(x,t)
∂t2
−u2 ∂2
I(x,t)
∂x2
= 0
where u =1 LC is the propagation speed in the transmission line
85. Standing wave soluNon
Let us try separation of variables
V(x,t) =V(x)exp( jωt) and I(x,t) = I(x)exp( jωt)
in this case one case show that:
d2
V(x)
dx2
+ ω2
LC.V(x) = 0 and
d2
I(x)
dx2
+ ω2
LC.I(x) = 0
we define wave number k =ω LC =
ω
u
d2
V(x)
dx2
+ k2
V(x) = 0 and
d2
I(x)
dx2
+ k2
I(x) = 0
Result: one dimensional Helmholtz wave equation
86. General soluNon standing waves
06/10/17 TU Del3 class ae3-535 86
V(x) =V1 e− j.k.x
+V2 ej.k.x
I(x) =
V1
z0
e− j.k.x
−
V2
z0
ej.k.x
z0 =
L
C
V
x
(characterisNc impedance)
Coaxial cables usually come with an impedance between 50Ω
and 100Ω, twin lines 300 to 450Ω. Coax: asymmetric, Twin:
symmetric
93. • The decibel is defined as:
• Normally Pref = 1
– dBW is used for power relaNve to 1 WaS
– dBm is used for power relaNve to 10-3 WaS
– dBi is the gain relaNve to an isotropic antenna
– dBd is the gain relaNve to a dipole
– dBc is takes relaNve to a carrier
• MathemaNcs:
– Adding dB’s is the same as mulNplicaNon
– SubtracNng dB’s : divide values
dB=10log10
P
Pref
!
"
#
$
%
&
107. • Dipole
• Half dipole, quarter dipole, etc
• MulN-element yagi antennas
• Helical antenna’s
• Patch antenna’s
• Parabolic antenna’s
• AcNve antenna’s (means that a remote receiving
antenna in a noise free environment comes with
an LNA to counteract transmission line losses)
06/10/17 TU Del3 class ae3-535 107
108. • Antenna gain reveals that the antenna is somehow
direcNonal.
• Your ears and your voice are direcNonal antennas
• Rather than that all radiated power goes out in every
direcNon, a direcNonal antenna takes care of radiaNng
the transmiSed power along a paSern.
• Antenna gains are expressed as dBi and these are
relaNve to a theoreNcal isotropic antenna that has
unit gain.
• DirecNonal antennas has a front-back raNo in dB
• When you receive a signal a similar thing happens, the
incoming energy is focused onto one point
06/10/17 TU Del3 class ae3-535 108
119. This lecture
• Impedance matching
• IntroducNon to modulaNon
• Signal propagaNon
• Signal and noise
• Link margin calculaNon
• Reference material:
– Electronics, A system approach, 6th ediNon Neil Storey
– Week 1.3: Chapters 1-9
– Week 1.4 and 1.5: Chapter 29
121. Impedance matching
• Impedances:
– Transceiver : design of power amplifier (50 Ω)
– Transmission line : type of line that is used (50 Ω)
– Antenna : type of antenna that is used (we hope it is 50 Ω)
– In the ideal case all impedances should be the same
• The reality is:
– The antenna does not match to the transmission line
– TransmiSed signal is reflected back to the transmiSer,
– Reflected signal could damage the transmiSer
• For RF circuits we measure a standing wave raNo (SWR)
– We opNmize the SWR and it should become 1
– There are various ways to accomplish an impedance match
06/10/17 TU Del3 class ae3-535 121
134. FM beRer than AM?
• Yes:
– FM less suscepNble to variaNons in recepNon strength
– FM is therefore more resilient to noise
– FM does not require highly linear amplifiers
• No:
– Requires more bandwidth, FM is typically used at
frequencies >100 MHz where there is enough
bandwidth
– DemodulaNon circuits are more difficult (PLL’s etc) (is
not a real issue nowadays)
06/10/17 TU Del3 class ae3-535 134
147. Physics of noise at the receiver
06/10/17 TU Del3 class ae3-535 147
In reality we deal with electronic components which
have a certain temperature T, all components emit
electromagnetic radiation:
P = kT B with k =1.38064854×10−23
J K−1
and B bandwidth
The Bolzmann constant comes from
k =
R
N
where R is a gas constant and N Avogadro's number
For semiconductors etc the thermal voltage is:
VT =
kT
q
where q is the charge of an electron
P is the Power contained in the noise
148. Physics of noise: spectral density
06/10/17 TU Del3 class ae3-535 148
Spectral density is specified in dBm/Hz, but how?
For T=290K (room temperature) we find
Pdbm =10log10 (
kT
10−3
) = −174 dBm/Hz
for a bandwidth of 1 MHz we get -114dBm,
note that the dBm calculation refers to 1mW
and that we can add to multiply or subtract
to divide due to the definition of the decibel.
152. Link margin calcula2on
06/10/17 TU Del3 class ae3-535 152
TransmiSer Receiver
PT PR
Free space loss
PR = PT − L +GT +G R with L =
4πd
λ
⎛
⎝
⎜
⎞
⎠
⎟
2
or in dB
L = −20log10 (λ)+ 20log10 (d)+ 21.98 where 21.98 =10log10 ((4π)2
)
M = PR − Nr is the so called link margin, preferable it is > 0
Nr = kT B + Nd
Nd : You got it from the noise floor measurement
153. Example link margin calcula2on
• Transmit power : 10mW which is 10dBm
• Receiver noise: -100dBm (noise floor measurement)
• Distance = 1000m, Frequency=5.8GHz, FSL = 107.7dB
• Gains transmiSer: 0dB, receiver: 0dB (ant./line etc)
• Margin: 10-107.7-(-100) = 2.3dB > 0 dB (ok);
– Obstacles (wet leaves, etc) would aSentuate the signal
– They introduce noise or even completely block the signal
– Any reflector in the Fressnel ellipse affects the quality
• Various opNons to improve the situaNon:
– Increase the antenna gain
– Choose beSer antenna posiNons (no trees, hills, buildings)
06/10/17 TU Del3 class ae3-535 153
155. Signal, noise, bandwidth
06/10/17 TU Del3 class ae3-535 155
dB
f
Signal 1 Signal 2
noise
Clearly signal 1 is below the noise level, and signal 2 is above it, and intuiNvely
you would think that signal 1 could not be received while signal 2 could, however,
this is not per se true. In week 1.5 we will conNnue with this topic
SNR
177. We start with
S
N
=100.1×SNRdb
C ≤ Blog2 1+
S
N
⎛
⎝
⎜
⎞
⎠
⎟ and
S
N
B =
Eb
No
C
C
B
= log2 1+
Eb
No
C
B
⎛
⎝
⎜
⎞
⎠
⎟ ⇒ 2C B
−1=
Eb
No
C
B
η =
C
B
⇒
2η
−1
η
=
Eb
No
⇒
η→0
lim
2η
−1
η
⎛
⎝
⎜
⎞
⎠
⎟ = ln(2) = 0.693...
As a result we find that min
Eb
No
⎛
⎝
⎜
⎞
⎠
⎟
dB
= -1.592... dB
179. Digital receiver sensi2vity model
The relation between carrier (signal) to noise ratio and the
spectral density of energy is B
C
N
=
Eb
N0
fb where fB is the
bit rate and B the bandwidth,
EB
N0
is the energy per symbol
to noise ratio spectral density. As a result the receiver sensitivity
sensitivity σ becomes:
σ =10log10 (B)+10log10
C
N
⎛
⎝
⎜
⎞
⎠
⎟−174dBm + NF
which is equivalent to:
σ =−174dBm + NF +
Eb
N0
⎛
⎝
⎜
⎞
⎠
⎟
dB
+10log10 ( fB )
Where NF is the noise floor of the receiver, it says something
about the quality of the receiver.
189. GPS signal structure
• What informaNon is relevant for the user?
– L1 : contains C/A codes and the P (or Y) codes
– L2 : contains P (or Y) codes
– C/A codes bandwidth is 1.023 MHz, data rate = 50 bps
– P(Y) code bandwidth is 10.23 MHz, data rate = 50 bps
• Codes are unique for each GPS space vehicle (S/V)
– Code repeNNon length C/A = 1023, or 1 milliseconds at 1.023 mbps
– Code length P(Y) = 6.19 x 1012, or 7 days at 10.23 mbps,
– The full P(Y)-code cycle is length is however much longer
• New GPS signals are planned in the near future, the
above overview is not meant to represent the new
situaNon
• Also: GNSS = GPS + Galileo + Glonass + Beidou
06/10/17 TU Del3 class ae3-535 189
190. PRN or Gold code
XOR 0 1
0 0 1
1 1 0
D
Ck
Q
Flip
Flop
FF1 FF4 FF5 FF6 FF7 FF8 FF2 FF3
preset
clock
PRN code
XOR
A
B
C
01001110101…
XOR truth table
Q=D a3er
acNve flank
of Ck
hSps://www.maximintegrated.com/en/app-notes/index.mvp/id/1890
193. GPS Receiver
• EssenNally it is a superheterodyne receiver except that
you would not be able to hear anything because:
– All S/V PRN signals are on top of one another
– The signals are below the noise level of the receiver
• So how would this work? SNR < 1 seems a bit strange.
• Digital code correlaNon solves this problem:
– Generate PRN code replica’s (C/A and NAV are always
accessible, P-codes were open, Y-codes are classified)
– Degrees of freedom during correlaNon are the code phase
offset and the frequency of the signal.
– Conclusion, there are at least two tracking loops in a GPS
receiver, one aligns the codes, the other aligns the
frequency
06/10/17 TU Del3 class ae3-535 193
196. 06/10/17 TU Del3 class ae3-535 196
By correlating the received signal with a replica code
you introduce a so-called processing gain:
db(Gain) = 10 log10
bandwidth of the PRN code
bandwidth of the data rate
!
"
#
$
%
&
db(Gain) =10 log10
2MHz
100Hz
!
"
#
$
%
& = 43db
which lifts the signal above the noise level
Processing gain due to correla2on
For more details: D. Doberstein, Fundamentals of GPS receivers, appendix A, Springer Verlag 2012
197. GPS naviga2on
• Transmit Nme of GPS signal is known (atom clock)
• Received from the GPS S/V’s are (X,Y,Z,T)satellite
• Observed: Code-phase difference between the local
oscillator (iniNally not synchronized) and the
transmiSed code
• This informaNon is called pseudo-range informaNon
(comes with a 1msec ambiguity)
• Pseudo range = c . (Treceived – Tsend) + Biasreceiverclock
• Phase range à integrate the Doppler effect (This is
what scienNsts/engineers call the carrier phase)
06/10/17 TU Del3 class ae3-535 197
200. Exercise on digital modula2on
• Problem descripNon
– Assume a 2.4 GHz QPSK signal,
– Channel spacing is 40 MHz for 802.11n wifi
– Assume SNRs between -15 and +15dB
– BER = 1/2*erfc(sqrt(2*Eb/No)*sin(pi/4))
• Exercise:
– The channel capacity is a funcNon of the SNR, How is
the BER is affected by the SNR? Make a graph.
– At what SNR is your wifi connecNon sNll comfortable?
– EsNmate the spectral efficiency parameter, and the
minimum Eb/No that is possible.
06/10/17 TU Del3 class ae3-535 200
208. What do you get?
• Moteino board, plugged onto a breadboard
• Jumper wires and USB cables
• Sign-up with name and student ID and return the
hardware a3er period Q2
• You can leave your experiment on the 9th floor, e.g. in
my office,
• You can take the boards with you as long as safe
transportaNon is demonstrated
• Preferred way of transportaNon: a carton box
• Any boards that leaves the faculty needs to be
registered
06/10/17 TU Del3 class ae3-535 208
219. • Experiment
– Find an area that is mostly free of obstrucNon
– Measure the RSSI offset close to the beacon.
– Measure the RSSI values at a distance up to 100 meter
from the beacon
– Evaluate the free space loss term in the link budget.
• ReporNng
– Answer all quesNons related to the FSL experiment
– Include a plot the measured free space loss
– Match the plot with a FSL equaNon
• Maximum: 3 points out of 10
06/10/17 TU Del3 class ae3-535 219
222. Experiment to simulate a satellite to groundstaNon link
• Experiment
– You need two set-ups of the Arduino IDE
– No baSery powered experiments are allowed, only laptop and USB
– One moteino is the satellite, the other moteino is the ground staNon
– The satellite performs measurements of the NTC and/or LDRs.
– Ground staNon collects the measurements
– LEDs are the actuators on the satellite
– Make a funcNonal diagram first, consult a supervisor to check the design
– Get it to work and demonstrate the end-result to the TA
• ReporNng
– Include the schemaNc, calculate the current by pin in the moteino board
– The full experiment should be described in the report.
• Maximum: 2 points out of 10
223. • What we always expect
– A group report should be submiSed as one unsigned PDF file
– No separate MATLAB or Python code files or plots
– Clearly state which problem (1 to 4) you are reporNng
– All names and study numbers should be included
– DistribuNon of tasks should be included in the report
– Deadline: November the 14th 2017.
• Mandatory
– Free space loss experiment: 3 points
– Exercises previous weeks: 3 points
• OpNonal
– ModulaNon performance : 2 points
– SimulaNon ground staNon satellite: 2 points