What is digital signal processing?
Basic DSP system:
Analog
Signal
Sampling
Digital
Signal
DSP
Digital
Signal
Reconstruction
Analog
Signal
Advantages:
• DSP hardware is flexible and programmable
• DSP chips are relatively cheap (easily mass-produced)
• Digital storage is cheap
• Digital information can be encrypted, coded, and compressed
Disadvantages:
• Sampling leads to loss of information
• High-resolution ultra-fast A/D and D/A may be expensive
• Digital processing cannot always be done in real-time
Applications of DSP
• Digital Audio and Video
− compact discs
− Dolby recording
− HDTV
− JPEG, MPEG
• Speech systems
− recognition
− synthesis
• Industrial
− noise cancellation
− manufacturing control
• Biomedical
− ultrasound
− hearing aids
• Telecommunications
− wireless phones
− DSL, cable modems
• Science
− speech pathology
− gene research
• Sonar and radar
− surveillance systems
− oceanography
Long range propagation experiments
Acoustic Thermometry of Ocean Climate (ATOC) 1995-1996
North Pacific Acoustic Laboratory (NPAL) 1998-1999
SPICE04/LOAPEX 2004-2005
• ATOC
Long range acoustics experiments
60
– 75 Hz source
– 28/84 Hz source
50
• SPICE04/LOAPEX
– Shallow/deep
250 Hz sources
– Fixed/towed 75 Hz
sources
Latitude
• NPAL
– 75 Hz source
Pioneer
Seamount
40
Pt. Sur
30
20
Kauai
Hawaii
10
0
−10
−180
Kiritimati
−160
−140
ATOC
NPAL
SPICE04/LOAPEX
−120 −100
Longitude
−80
−60
Spectrogram analysis of tomographic signals
Spectrogram shows frequency content as a function of time
− Similar to output of bank of bandpass filters
Pseudo-random sequence signal:
Linear FM signal:
Spectrogram of linear FM signal
800
800
700
700
600
600
Frequency (Hz)
Frequency (Hz)
Spectrogram of M−sequence signal
500
400
300
500
400
300
200
200
100
100
0
0
20
40
60
80
Time (seconds)
100
120
140
0
0
20
40
60
80
Time (seconds)
100
120
140
Foundations of DSP
What do you have to know to start working on DSP?
• Fluent in mathematics
• Thorough understanding of sampling and its consequences
• Knowledge and appreciation of signal transforms
• Basic spectral analysis tools: DFT/FFT
• Ability to implement simulations in software such as Matlab
• Understanding of the science, mathematics, and/or engineering
behind a particular application
What will ECE 410 cover? ⇒
ECE 410
• Syllabus
−
−
−
−
−
Review of DT signals & systems
DT processing of CT signals
Filter design
Discrete Fourier series & discrete Fourier transform
Introduction to spectral analysis
• Course information sheet
−
−
−
−
−
Active and collaborative format
Website
Required textbooks and software
Grading and policies
Office hours
⇒ Questions?
Sample RAT: O/S/B Sections 2.0-2.2
1. Name
2. T/F A discrete-time sinusoid with frequency ω = 0 is
indistinguishable from a discrete-time sinusoid with
frequency ω = π.
3. T/F Causal systems are nonanticipative.
4. T/F Discrete-time complex exponentials are not always
periodic.
5. T/F BIBO stands for bounded-input, bounded-output.
Sample in-class problem
Consider the system defined by y[n] = x[n] + 3u[n + 1].
1. Is it linear?
2. Is it time-invariant?
3. Is it stable?