Ram Zamir' Homepage

Prof. Ram Zamir

Department of Electrical Engineering-Systems
Tel Aviv University
Ramat Aviv 69978
ISRAEL

Phone: 972-3-6406273

Fax: 972-3-6407095

Mail: zamir@eng.tau.ac.il

line

General Research Interests:

Analog coding & good frames

Statistical communication

Network information theory, multiterminal source coding, multiple descriptions

Rate-distortion theory, quantization and universal compression

Use of side-information in lossy compression, and in channel coding and decoding

Information theoretic bounds on estimation and coding

Book: Lattice Coding for Signals and Networks

A Structured Coding Approach to Quantization, Modulation and Multiuser Information Theory.

** Cambridge University Press, August 2014**

See http://www.amazon.com/Lattice-Coding-Signals-Networks-Quantization/dp/0521766982#reader_0521766982.

The Book

Why to write a book (on lattice codes)? a seminar for the celebration of the publication, Oct. 20, 2014

Current Projects & recent talks:

Analog coding and good frames

Nested codes for information embedding, distributed compression and interference cancelation

How to exploit quality information using variable-partition quantizers?

Adaptive ``shaping'' of quantization-codebooks via the mechanism of ``natural type selection''

Minimum entropy noise compensation, for channels with side-information at the transmitter

Diversity techniques in signal compression for lossy packet networks

Tree-codes for universal lossy compression, and the "Lempel-Ziv" algorithm

Coding and modulation for non-Gaussian multiple-access channels ("non Gaussian" = e.g., impulsive or fading)

Tutorial Talks About Lattices:

Lattice Coding for Signals and Networks: - Application and Design (Tutorial talk at ISIT 2012, MIT)

and Parts B, and C.

Can Structure Beat Random? - the Story of Lattice Codes (Plenary talk at ISIT 2010, Austin)

and (IT Newsletter, March 2011)

Lattices are Everywhere (ITA 2009, UCSD)

Slides: Lattices are Everywhere!

Slides: Lattices for Source Coding with Side-Information.

Slides: Lattices for Channel Coding with Side Information.

Slides: Lattices for Modulation and Bandwidth Conversion.

Information Theory 1 (graduate level): information measures; entropy and the lossless source coding theorem; capacity and the channel coding theorem; extension to continues signals and channels; rate-distortion function and the lossy source coding theorem; advance topics (multi-terminal systems, universal coding).

Digital Transmission of Signals (under-graduate level): PCM, DPCM and vector quantization; equalization and maximum likelihood sequence estimation; adaptive filtering methods in communication.

Information Theory 2 (graduate level): network information theory (separate encoding of correlated sources, multi access channels, broadcast channels, coding with side information, multiple descriptions, successive refinement, and more); the method of types (error exponents, universal compression, mismatch encoding, large deviation); information theoretic inequalities; complexity measures for sequences. (more.)

Data and Signal Compression (graduate level): lossless source coding; universal source coding; Source optimized quantization (scalar and vector), high resolution analysis, lattice quantization and entropy coding, Shannon theory for lossy compression, the "water filling" principle and tranform coding, predictive coding; speech, picture and video compression algorithms, joint source/channel coding, compression for lossy packet networks.

Projects and Bibliography (1999/2000)

Random Signals and Noise (under-graduate level): PART I - Random variables and vectors. PART II - Generation of a random process, Joint statistics, Auto-correlation, Gaussian process, Stationarity, WSS, Auto-regressive process, Asymptotic Stationarity, Poisson and Wiener Processes, empirical statisitics and LLN, Markov chains, Ergodicity, Power spectrum; Random signals and linear time-invariant systems; Linear mean-square estimation of a noisy process (Wiener Filter); other estimation / filtering criteria.