734 Course Description

Time and Place:	Lec. 1, 11:00 - 11:50 AM, MWF,3444 Engr. Hall
Instructors:	Yu Hen Hu, 3625 Engr. Hall, Tel. 262-6724 E-mail: hu AT engr dot wisc dot edu
Credits:	3
Prerequisite:	ECE 431 Digital Signal Processing, ECE/CS 552 Computer Architecture, or consent of instructor. General knowledge of numerical linear algebra, signal processing algorithms, and computer architecture desirable.
Goals:	This course presents design methodologies of multimedia and communication algorithms over embedded micro-architecture and platforms. The emphasis is on joint optimization of application algorithm design and implementation hardware/software design space exploration. Current focus are on standard oriented multimedia applications including JPEG2000, H.264 and wireless communication applications such as WiFi, WiMax and software defined radio. Implementation platforms including sub-word parallel architecture, graphic processing unit, multi-core and system on chip (network on chip) will be discussed.
Topics:	Design and Implementations of multimedia and communication standards and algorithms Basic signal processing algorithm review: linear transformations, filtering Algorithm representation and parallelism: dependence graph, signal flow graph, synchronous data flow Non-recurrent algorithm transformation: Pipelining and vector pocessing, retiming Recurrent algorithm transformation: iteration bounds, folding, unfoling, look-ahead transform Subword parallel ISA platform implementation strategies Multi-core implementation strategies System-on-chip platform implementation strategies Re-configurable computing using field programmable gate array (FPGA) Signal processing arithmetic units: distributed arithmetic, CORDIC Implementation of video coding standards MPEG, and JPEG: DCT and DWT architecture, motion estimation architecture, entropy coder architecture Implementation of communication algorithms: digital modulation systems
Textbook:	K. K. Parhi, VLSI Digital Signal Processing Systems, Design and Implementation, John Wiley & Sons, Inc, NY, NY. Notes will be available on line.
Computer Usage:	Matlab may be used to demonstrate some algorithms. Students who work on specific class projects may need to use other software.
Homework:	Two to three homework assignments will be given. In addition to problems, homework may involve hand-coding of algorithms
Grading Policy: (tentative)	30% 4 homework assignments; 40% individual class project 30% take-home final exam

COURSE DESCRIPTION