EE123: Digital Signal Processing, Spring 2024
EE123: Digital Signal Processing, Spring 2024
The purpose of this course website is to link to course documents, like lecture notes, homeworks, and labs, and to provide information about the course to the public. Always remember to check Ed for announcements and communications with the course staff.
Catalog Description: (4 units) Discrete time signals and systems: Fourier and Z transforms, DFT, 2-dimensional versions. Digital signal processing topics: flow graphs, realizations, FFT, quantization effects, linear prediction. Digital filter design methods: windowing, frequency sampling, S-to-Z methods, frequency-transformation methods, optimization methods, 2-dimensional filter design.
Prerequisites: EECS 120, or instructor permission.
Course objectives: To develop skills for analyzing and synthesizing algorithms and systems that process discrete time signals, with emphasis on realization and implementation.
Why should you care? Digital signal processing is one of the most important and useful tools an electrical engineer could have. It impacts all modern aspects of life and sciences; from communication, entertainment to health and economics.
Gopala Anumanchipalli / gopala@berkeley.edu
OH: Monday 3 PM- 4 PM, Cory 490A
Louis Liu/ louis_liu@berkeley.edu
OH: Wednesday 10 AM - 11 AM, Cory 111
Lecture, homework, and general questions
Kaylo Littlejohn / kaylo_littlejohn@berkeley.edu
OH: Wednesday 1 PM - 2 PM, Cory 111
Lab questions
Jason S. Kim, jasonskim007@berkeley.edu
Tuesday/ Thursday, 5:00 pm - 6:30 pm, Etcheverry 3108
Friday: 1:00 pm - 2:00 pm, Social Sciences Building 20
Wednesday 11 AM - 12 PM, Cory 111
Wednesday 12 PM - 1 PM, Cory 111
Wednesday 2 PM - 4 PM, Cory 111
Wednesday 9 AM - 11 AM, Cory 111
A list of the topics that will be covered is given here, in the order that they will be covered. This may change based on time.
Intro to DSP, discrete signals, LTI systems
DTFT, properties, frequency response
The z transform, properties, ROC, inversion, DFT
Fast convolution, circular vs. linear convolution, overlap add/overlap save
FFT, decimation in time/frequency, spectral analysis using the DFT
Spectral analysis using the DFT, windowing, resolution tradeoffs, zero-padding
STFT, T-F tiling, uncertainty principle
Wavelets in continuous time, Haar expansions, Discrete Wavelet Transform using Haar basis
T-F tilings of wavelet bases, orthonormal bases
Wavelets as multi-resolution ladder of spaces, connection to filters and filter banks
Sampling, Aliasing, Reconstruction, DT processing of CT signals
Resampling, CT processing of DT signals, interpretation of non-integer delays
Filter design, FIR filters, windowing, intro to minimax optimal FIR filters (Parks-McLellan)
Multirate identities, sampling rate conversions, polyphase decompositions and filter banks
PRFBs, vector-space view of filter banks as orthonormal basis expansions
Filter-banks and wavelets: Mallat’s Algorithm
Transform analysis of LTI systems, Allpass, Min-Phase and Generalized Linear Phase
Optimal FIR filters, Parks-McLellan Algorithm, Equiripple filters
Oversampled ADC, noise shaping and quantization noise analysis
Basics of image compression
Sampling below the Nyquist Rate: finite rate of innovation sampling
Compressed Sensing
2D-DFT
Tomography
Homework (Weekly): 10%
Labs: 15%
Midterm 1: 30% (TBD)
Midterm 2: 30% (TBD)
Project: 15%
Weekly assignments consist of problem sets. In addition, there will be about 4-6 lab assignments consisting of programming using Jupyter notebook.
Homework will be assigned each Friday and due the next Friday at 11:59pm with a grace period of 2 days.
Homework submission will be in digital form through Gradescope. Here's a LaTeX template Miki_Lustig_hw01_sol.tex that produces this output after compilation. If you don't want to typeset, you can use a tablet computer or scan handwritten homework using a document scanning app for your smartphone.
No late homework without prior consent from the instructor.
Homework will be self graded. Self grading will be due the Monday after the solutions are released.
Homework slip policy: the homework with the lowest grade will be dropped.
01/19/24 HW1
01/26/24 HW2
02/02/24 HW3
02/09/24 HW4
02/16/24 HW5
02/23/24 HW6
03/08/24 HW8
03/15/24 HW9
03/22/24 HW10
Midterm 1: Thursday, March 7th
Midterm 2: Thursday, April 11th
01/16/24 Lecture 1 Intro, DT Signals, Systems & Properties
01/18/24 Lecture 2 LTI systems and DTFT, Lecture notes
01/23/24 Lecture 3 Z-Transforms
01/25/24 Lecture 4 Convergence, Z-Transforms Lecture notes
01/30/24 Lecture 5 Transform Analysis of LTI Systems
02/01/24 Lecture 6 Complete AllPass, MinPhase Systems
02/06/24 Lecture 7 Discrete Fourier Transform (Chapter 8 O & S) Lecture Notes
02/08/24 Lecture 8 Fast Convolutions, FFT (Chapter 8 & 9 from O & S), Lecture Notes
02/13/24 Lecture 9 Spectral Analysis using DFT (Chapter 10 from O & S), Lecture Notes
02/15/24 Lecture 10 Spectral Analysis contd, Short Time Fourier Transform, Gatspar Notes
02/20/24 Lecture 11 Introduction to Wavelets
02/22/24 Lecture 12 Discrete Wavelets, IWT, Haar Wavelet
02/27/24 Lecture 13 Intro to Sampling (Ch 4 from O&S)
02/29/24 Lecture 14 Resampling
03/05/24 Lecture 15 Polyphase decomposition
03/07/24 Midterm 1
03/12/24 Lecture 16 Sampling Review
03/14/24 Lecture 17 FilterBanks, Perfect Reconstruction Filter Banks
03/21/24 Lecture 18 Practical ADC, Notes
04/02/24 Lecture 19 Practical ADC Quantization
04/04/24 Lecture 20 Generalized Linear Phase
01/19/24 Disc1, Music Generation
02/09/24 Disc4_pre, Disc4_sol, DFT&DTFT
02/16/24 Disc5_pre, Disc5_sol, STFT
03/01/24 Midterm 1 review, z-transform demo
03/08/24 Disc7_pre, Circular&Linear Conv Explain, Disc7_sol
04/05/24 Midterm 2 review, Midterm 2 review sol
04/16/24 DDSP