EE521 Analog and Digital Communications
Download
Report
Transcript EE521 Analog and Digital Communications
EE521 Analog and
Digital
Communications
James K. Beard, Ph. D.
Tuesday, January 18, 2005
[email protected]
7/16/2015
Week 1
1
Essentials
Text: Bernard Sklar, Digital
Communications, Second Edition
Prerequisites
Consent
of Instructor (Dr. Silage)
SystemView (CD-ROM with text)
Office
E&A 709
Hours
7/16/2015
TBA, Tuesday afternoons planned
Week 1
2
Today’s Topics
Course overview
Baseband signals and modulation
Introduction to SystemView
Feedback
What
is your background?
What do you expect from EE320?
Discussion (as time permits)
7/16/2015
Week 1
3
ECE Department’s Thumbnail
EE521 Digital Communications
Prerequisite: Instructor's Permission (Dr. Silage)
Modulation
Coherent
detection
Signal-to-noise ratio
Probability of bit and symbol error
Modulations – PSK, FSK, ASK, QAM, M-ary FSK, Mary PSK
Was EE400
7/16/2015
Week 1
4
Course Summary
Fourteen weeks of classes
Two in-progress exams, one final exam
In-progress on 5th and 12th weeks, 20%
Final on fifteenth week, 20% of grade
of grade each
Individually assigned project
Assigned
in fifth week
Execute your project in SystemView
40% of grade
Deductions from final grade
0.5% for each unexcused absence
1% for each missed 10 minute Pop
7/16/2015
Week 1
Quiz response
5
Two Semesters in the Plan
EE521 is coordinated with a second
course next semester
EE521 covers approximately half of the
text
The follow-on covers the second half of
the text
Course number TBA
7/16/2015
Week 1
6
Topics (1 of 3)
Formatting and baseband modulation
Encoding
digital, text and analog data
Sources of corruption: noise, channel issues
Pulse code modulation
Uniform and non-uniform quantization levels
Baseband modulation
Baseband modulation and detection
Digital
signals in noise
Detection of encoded data from digital signals
7/16/2015
Week 1
7
Topics (2 of 3)
Bandpass modulation and demodulation
– advantages and disadvantages
Modulation and detection of digital signals
Coherent and non-coherent detection
Error performance for M-ary encoding
Rationale
Channel coding
Waveform
coding and error control
Sequences and error detection and correction
Block codes
7/16/2015
Week 1
8
Topics (3 of 3)
Modulation and coding trade space
Goals
and requirements
Sampling and aliasing
Channel capacity
Designs for maximum data rate and minimum
errors
7/16/2015
Week 1
9
Next Semester
Communication link analysis
Channel coding
Synchronization
Multiplexing and multiple access
Encryption
Channel issues
Supplementary topics
7/16/2015
Week 1
10
Essential Technologies
Analog and digital signal processing
Real
and complex signals
Vector representation of phase
Sampling capabilities and limitations
Probability and Statistics
Behavior
of noise in linear systems
Effect of noise on decoding operations
Coding, modulation, and demodulation
7/16/2015
Week 1
11
History
1864 – Maxwell predicted radio waves
1887 – Hertz demonstrated radio waves
1897 – Lodge demonstrated wireless communications
1901 – Marconi demonstrated transatlantic communications
1903 – DeForest demonstrated first vacuum tube amplifier
1906 – Fessenden started first AM radio station
1927 – First TV broadcasts
1947 – Microwave relay from Boston to NYC
1947 – Bell Labs announced the transistor
1955 – TI announced production silicon transistors
1958 – First satellite voice channel
1981 – First cell phone system, in Scandinavia
1988 – First digital cell phone system in Europe
7/16/2015
Week 1
12
Formatting and Baseband
Modulation
Topics from Text Chapter 2
2.1,
Baseband signals
2.3, Messages, characters and symbols
2.4, Formatting analog information
Baseband signals
Messages, characters, and symbols
Formatting data
7/16/2015
Week 1
13
Sklar Chapter 2
7/16/2015
Week 1
14
Baseband Signals
Definition: Data source, the signal to be
presented to data sink at receiver end
Principal types
Digital
Text
or data
Analog
7/16/2015
Week 1
15
Message, Characters and
Symbols
Definitions
– the data to be presented to the
data sink or used to generate analog
output at the receiver
Character – base message unit, such as a
letter of the alphabet or an 8 or 16 bit word
produced by an ADC
Symbol – a grouping of bits for encoding
Message
Characters and symbols are often
different sizes
7/16/2015
Week 1
16
Formatting Analog
Information
Operations are
Sampling – uniformly spaced captures of a
Digitization – reduction of a data sample to
waveform
a set of
discrete levels
Limitations include
Aliasing;
the digitized forms of two signal differing
from one another by the sample rate are identical
Quantization; may be uniform or non-uniform, and
presents a noise floor
Clipping; there is a maximum value that may be
represented
7/16/2015
Week 1
17
Sampling
Can be characterized as
Switch
a waveform into a follower
At the sample time
Switch a holding capacitor into the holding circuit
Switch off the waveform input
The
follower becomes an integrator with no input
and holds the signal
An ADC operates on the held signal
Sampling is an analog operation and has a
frequency response
7/16/2015
Week 1
18
Aliasing
Definition:
A tone
signal with a frequency higher than the
sample rate will produce a data format identical to
that of a tone with a frequency lower than sample
frequency
The difference between the two frequencies is an
integer multiple of the sample frequency
Usual ambiguity range is from minus half the
sample frequency to plus half the sample
frequency
7/16/2015
Week 1
19
Issues With Sampling
Signal center frequency is often much larger
than the bandwidth
Two techniques are used to keep data rate to
twice the signal bandwidth
Quadrature
demodulation, the use of two LO’s in
quadrature with two mixers to produce two baseband
signals in quadrature
Digital quadrature demodulation, undersampling at IF
followed by digital quadrature demodulation
7/16/2015
Week 1
20
Real vs. Complex Sampling
For real sampling at frequency fs > 2.B
The
ambiguity range is 0 to fs
Negative frequencies are ambiguous with
positive frequencies
For quadrature demodulation and output
samples at frequency fqs > B
ambiguity range is –fqs/2 to +fqs/2
Positive and negative frequencies are
unambiguous
The
7/16/2015
Week 1
21
Digital Quadrature
Demodulation
Undersample
Signal
at frequency f0 with bandwidth B
Sample at rate fqs
4 f0
f qs
B
2k 1
Result
Signal
aliases to plus or minus half the sample rate
FIR filter and decimate to produce the final result
Signal is then ready for FFT or other processing
7/16/2015
Week 1
22
Digitization Levels
Uniform and non-uniform commonly used
Uniform sampling levels – used for high quality
systems
Requires
Music
Radar
more bits, typically 10 to 16 bits
and sonar
Non-uniform sampling levels – used where
bandwidth drives the channel utility
Cell
phones
Some desk phones and VOIP
Typically 8 bits
7/16/2015
Week 1
23
Effects
Both types have a “hard” ceiling
Extra ADC circuitry offers hard limiting in place
of end-around overflow
Noise floor is different
levels – usually characterized as an additive
noise with an RMS value of 1/12 the LSB (see next
slide for real-world considerations)
Nonuniform levels – COMDAC levels or fine at low
amplitudes, coarse at high levels
Quantization noise floor higher for large signals
Uniform
7/16/2015
Week 1
24
Real-World
Sampling and ADC
Sampling is
The last mixer in the receiver chain
Produces signal at baseband from signal at IF
Sample clock phase noise is part of the signal chain error sources
Dynamic range
Dynamic range is less than theoretical with nearly all ADCs
Effective number of bits (ENOB) and SNR are two synonymous
ways of specifying ADC dynamic range
ENOB is typically 1.5 bits less than the word length of an ADC
Spurious spectral lines
Sample aperture time and duration jitter can cause low level tonal
components in ADC output
Can come from inside the ADC chip or from the sample clock
Some resulting spurious lines are signal-dependent
7/16/2015
Week 1
25
Trades in Quadrature
Demodulation
Trades for analog quadrature demodulator
Matching
of channels for in-phase and
quadrature channels limits negative frequency
discrimination
Negative frequency ghosting typical 40 dB to
60 dB
Traces for digital quadrature demodulator
Ghosting
is determined by sampling and FIR
filtering/decimation scheme
7/16/2015
Week 1
26
Why Digital Quadrature
Demodulation
Advantages
Only
one ADC, no multiplexing or matching
Performance is determined by
Antialiasing filtering prior to ADC
Specifications of FIR decimation filters
Disadvantages
Sampling
is done at IF
Jitter and aperture time specs are tougher
Exchanged difficulties – from analog channel
matching to problems in sampling at IF
7/16/2015
Week 1
27
SystemView Overview
Simulations of digital signal processing
operations
Analog,
through oversampling and emulation of
analog operations
Digital, through direct emulation
Analysis
Bode
plots, root locus, other
Visualization
Time
domain plots
Virtual spectrum analyzer
7/16/2015
Week 1
28
Why SystemView
Dual-gate MOSFET RF converter, 144 MHz to 10 MHz. Uses the
RCA 3N140 for its overload and mixing capabilities. Circa 1968.
7/16/2015
Week 1
29
7/16/2015
Week 1
30
Summary
Course overview: we will learn how to
implement critical technologies for digital
communications
Course is integrated with a follow-on next
semester
Baseband signals are where we do
modulation == formatting for the channel
SystemView lets us have most of the
benefits of a lab with minimum time
7/16/2015
Week 1
31
Feedback
Around the room
Please let me know
What
your background is
EE
Design
SystemView
Communications
What
7/16/2015
you expect from this course
Week 1
32
Text and Assignment
Text
Benard Sklar, Digital Communicatinons ISBN 0-13-084788-7
SystemView User's Manual, Elanix, Inc
Assignment: Read Text
http://www.elanix.com/
http://www.elanix.com/pdf/SVUGuide.pdf
Chapter 2, 2.1, 2.3, 2.4, 2.5, 2.6, 2.7, 2.8
Load SystemView and examine the samples and demos
Browse appendices of text for review and
supplementary material
Look at TUARC
K3TU, websites
7/16/2015
http://www.temple.edu/ece/tuarc.htm
http://www.temple.edu/k3tu
Week 1
33