Risto Wichman Course data From 2009 onward: lectures in English only Lectures start on September 9 th , 2+2 hours during 6 weeks Exercises start on September 10 th Teaching asistant: Ejaz Aqib Course book: S.K. Mitra: Digital Signal Processing. A Computer-based Approach, McGraw-Hill, 3rd Edition, 2006/ 4 th Edition, 2011 Chapters 13-14 + some things from Chapters 7 and 10 Course material is in Noppa. 6 credits based on lectures (exam) and mandatory homeworks ! Basics of digital signal processing (eg., Aalto course T-61.3010/ELEC-C5230) or equal knowledge (Signals and systems etc.). ! Basic Matlab knowledge will help in solving the Matlab- homework problems. Simulink is also acceptable " Freeware Octave, www.octave.org, can be used as well
Course prerequisities To pass the course you must pass the exam and mandatory homework assignments. Alternative way to pass: Make Simulink + Pro Audio development kit C6727-PADK work together Lectures: background and theory, Matlab examples Concentrate on few concepts and learn them well (instead of superficially touching dozens of topics)! Exercises: Application of concepts presented in the lectures Exam: No additional material (calculators, tables of formulas etc.) is allowed Questions do not stay the same, old exams in web Course arrangement
! Final grade of the course comes from the following formula: Final grade = min[1 ; E ; H] * round{max[E ; ((0.4 + BIAS)*H + 0.6*E)]} where " E is the grade of the exam (0, 1, 2, 3, 4 or 5), " H is the grade of the homework (0, 1, 2, 3, 4 or 5) and " BIAS = 0.02. " About 50% of points are needed to pass the course ! This formula will give the final grades as illustrated in the table below.
E = 0 E = 1 E = 2 E = 3 E = 4 E = 5 H = 0 0 0 0 0 0 0 H = 1 0 1 2 3 4 5 H = 2 0 1 2 3 4 5 H = 3 0 2 2 3 4 5 H = 4 0 2 3 3 4 5 H = 5 0 3 3 4 5 5 Course grading Exercise system Exercise paper + hint session on Tuesday. Q&A session on Friday, new and old exercises Deadline to return solutions to the box in G4 on Monday Course contents Assumed background knowledge Discrete-Time Signals and Systems Discrete-Time Signals Typical Sequences and Sequence Representation The Sampling Process Discrete-Time Systems Time-Domain Characterization of LTI Discrete-Time Systems Simple Interconnection Schemes Finite-Dimensional LTI Discrete-Time Systems Classification of LTI Discrete-Time Systems Correlation of Signals Discrete-Time Fourier Transform The Continuous-Time Fourier Transform The Discrete-Time Fourier Transform Discrete-Time Fourier Transform Theorems The Frequency Response of an LTI Discrete-Time System Phase and Group Delays From ELEC-C5230 Assumed background knowledge Digital Processing of Continous-Time Signals Sampling of Continuous-Time Signals Analog Lowpass Filter Design Anti-Aliasing Filter Design Reconstruction Filter Design Effect of Sample-and-Hold Operation Finite-Length Discrete Transforms Orthogonal Transforms The Discrete Fourier Transform Relation Between the Fourier Transform and the DFT, and Their Inverses Operations on Finite-Length Sequences Classification of Finite-Length Sequences Linear Convolution Using DFT Assumed background knowledge z-Transform Definition and Properties Rational z-Transforms Region of Convergence of a Rational z-Transform The Inverse z-Transform z-Transform Properties Computation of the Convolution Sum of Finite-Length Sequences The z-Transform Function Discrete-Time Systems in the Transform Domain Transfer Function Classification Based on Magnitude Characteristics Transfer Function Classification Based on Phase Characteristics Types of Linear-Phase Transfer Functions Simple Digital Filters Assumed background knowledge Digital Filter Structures Block Diagram Reprensentation Basic FIR Digital Filter Structures Basic IIR Filter Structures IIR Digital Filter Design Bilinear Transform Method of IIR Filter Design Design of Lowpass IIR Digital Filters Design of Highpass, Bandpass, and Bandstop IIR Digital Filters Spectral Transformations of IIR Filters IIR Filter Design Using Matlab Computer-Aided Design of IIR Digital Filters FIR Digital Filter Design FIR Filter Design Based on Windowed Fourier Series Computer-Aided Design of Equiripple Linear-Phase FIR Filters FIR Digital Filter Design Using Matlab ! A key characteristic of multirate algorithms is their high computational efficiency facilitating an efficient implementation ! Altering the sampling rate at some point of the process creates several opportunities and enhancements for the filter design. ! The techniques are mostly related to aliasing due to the sampling rate change. ! Elementary theory and simple building blocks lead to sophisticated filter and system designs. " Combinations abound. No use to learn solutions by heart but learn the principles
Course contents Course contents Classical treatment of digital signal processing Filter (frequency response) design but no filtering Filters with fixed coefficients i.e. no adaptive filters Noise attenuation is not considered Delay from filtering is not considered Real-coefficient filters, not complex ones (except in uniform filter banks) Basic building blocks Up-sampling and down-sampling Discrete Fourier Transform (DFT) Polyphase decomposition Everything in the course can be derived and understood based on these three basic things
! Up-sampling and down-sampling " Aliasing " Linear but time-variant operations " Noble identities " Fractional (rational) sampling rate change " Sampling of band-pass signals ! Properties of linear-phase FIR filters " Magnitude response, amplitude/zero-phase response ! Window-based design of FIR filters ! Sparse impulse response ! Multistage filters ! No IIR filters in this course Basic concepts Basic concepts ! Polyphase decomposition, DFT ! Interpolation " Interpolation of samples vs. filter coefficients ! Nyquist filters ! Modulated filter banks ! QMF banks " Perfect reconstruction " Power complementary filters " Power symmetric filters ! L-channel QMF banks ! Multilevel filter banks ! Filter banks and wavelets ! Multistage Filters " Cascaded integrator comb filters (CIC) " Interpolated FIR filters (IFIR) " Complementary filters ! Interpolation, sampling rate change, fractional delay " Direct implementation of H(z) " Polyphase implementation of H(z) " Lagrange interpolation (no splines) " Farrow filters " Polynomial approximation of polyphase structure Filter structures ! Nyquist filters " Half-band filters " Lth-band filters " Raised cosine filters (type of Lth band filters) " Ideal band-limited filter ! Modulated filter banks " Discrete Fourier transfrom (DFT) banks " Non-critically sampled DFT banks " OFDM and filter banks
Filter structures Filter structures ! L-channel filter banks ! Modulated filter banks ! Direct inverse of modulation matrix H (m) (z) ! Polyphase representation of L-channel QMF bank ! Lattice L-channel QMF bank ! Multilevel, octave filter banks " Mother wavelet, scaling function Some Applications Analog low-pass filter must compensate the in-band distortion of S&H, provide -80 dB attenuation in the stopband and have a (24.1-20) KHz transition band Compact-disk 4-to-1 Oversampling Due to oversampling, the analog low-pass filter may have a transition band of (176.4-40) KHz, and the system becomes more cost- efficient Direct conversion transmitter for OFDM Orthogonal frequency division multiplexing (OFDM) modulation OFDM is a simple transmultiplexer - filterbank The same technique as in the previous slide to relax the specifications of the analog Tx filter Dirty RF Subband systems and transmultiplexers Filterbank Transmultiplexer Material Course material Lecture slides 13.1 -13.6 and 14.1-14.5 from Mitra 4 th
edition All things in the slides are not covered: spline interpolation and IIR filters skipped Supplementary slides Bandpass sampling Farrow filters Commutator representation of interpolation and decimation (Root) raise cosine Nyquist filters, OFDM Non-critically sampled filter banks, interpolation Lattice QMF structure Tiling patterns of tree-structured filter banks Course material Matlab examples Further illustrate the concepts in lecture slides Naming system rw_program_X.m: Modified from the programs that came with Mitras book example_X.m: Examples in the book but not among the programs accompanying the book Some_name.m: Other things that have come up the lectures ! S.K. Mitra: Digital Signal Processing. A Computer-based Approach, McGraw-Hill, 3rd Ed., 2006, 4 th Ed. 2011 " Course book 2006- " Matlab examples, didactic treatment " No communications applications ! f.j. harris: Multirate Signal Processing for Communication Systems, Prentice-Hall, 2005 " Practical, real world applications " Theory casually explained ! N. Fliege: Multirate Signal Processing " Course book -2005 " Lots of stuff Books on the subject ! P. Diniz, E. da Silva, S. Netto, Digital Signal Processing - System Analysis and Design, 2nd Edition, Cambridge University Press, 2010 ! Contains more advanced concept that Mitras book ! P.P. Vaidyanathan: Multirate Systems and Filter Banks, Prentice-Hall, 1993 " The book on multirate systems " Theoretical, advanced ! A.N. Akansu, R.A. Haddad: Multiresolution Signal Decomposition, Academic Press, 2nd Edition, 2001 " Emphasizes source coding ! R.E. Crochiere, L.R. Rabiner: Multirate Digital Signal Processing, Prentice-Hall 1983 ! M. Vetterli, J. Kovacevic: Wavelets and Subband Coding, Prentice Hall, 1995 ! H.S. Malvar: Signal Processing with Lapped Transforms, Artech House, 1992 ! O. Jahromi: Multirate Statistical Signal Processing, Springer 2007 ! Chapters on multirate filtering in dozens of DSP text books.