IEEE Navigational Bar IEEE Home Search IEEE Join IEEE About IEEE
- - -

Plenary talks and Tutorials of Ninth Annual symposium

- -

Plenary -1: "Broadband Access Using DSL Technologies" by Prof V Umapathi Reddy, IEEE Fellow, IIsc

Prof V Umapathi Reddy

The voice band modems operate over a bandwidth of approximately 3.4Khz. But the twisted pair can support much larger bandwidth. However the frequency response of the pair may vary over 100 dB over a frequency range of 1 MHz. The capacity of an ideal channel with a bandwidth of 1 MHz and SNR of 30 dB at the receiver is approximately 10 Mbps, while for a practical channel of the same bandwidth and the same total average transmitted power, but with non-uniform frequency response, the capacity is approximately 6.6Mbps. It is well known that a practical modulation scheme, such as PAM and QAM, requires about 9 dB higher SNR than that given by the capacity theorem to transmit at a data rate equal to the channel capacity and receive it at the receiver reliably (BER of 1E-6), assuming that ISI is removed completely. As the algorithms requried are computationally intensive, Multi Carrier Modulation has been proposed as the alternative to achieve the capacity. In MCM, the channel is partitioned into a number of parallel channels and different number of bits are transmitted in different subchannels depending on the SNR of the subchannel.

Discrete Multitone (DMT), a DFT based multi-carrier technique was the standard for ADSL applications. A zipper multi-carrier duplex scheme involving frequency division duplexing (FDD) was recently proposed for Very high speed digital subscriber applications (VDSL). A modified version of this scheme, using cyclic prefix and cyclic suffix, can eliminate the near end echo and NEXT, if the subcarriers are nearly non-overlapping.

Plenary Talk -2: "Internet for Net economy" by Dr Uday Shukla, IBM Global
Dr Uday Shukla With the current trend of the World Wide Web shrinking the univese to a few keystrokes, it is imperative that most businesses across the world would adapt, over a period of time, to a new and progressive way of collaborative living on the web. Dr Shukla grouped the technologies under the following
-Pervasive
-Knowledge managment
-Collaborative Technology
-Deep computing
-Internet media
Some of the insights from his presentation related to the future of Internet are
What we don't know
-How fast BW demand will grow?
-What new applications will be created?
-What new business models will be suitable?
-What governments will do?
What we know
-Growth estimate will be underestimated
-Applications expand to fill bandwidth
-Competition will be fierce
-E business will transform all sectors
-Learn a lot by looking at key users
-Kids and University researchers are trend setters

Tutorial-1: "Electronic markets" by Prof Y Narahari, CS&A, IISc Prof Y Narahari
Electronic markets are scalable web based platforms for buyers, sellers and brokers to carry out business transactions. The tutorial explained the different building blocks of a digital market using an example e-market KASBAH created at MIT. Excellence in E-markets is achieved only through superior core competence in a wide variety of disciplines: Internet technologies, distributed objects, software architecture, cryptography and security technologies, supply chain management, optimisation and game theory, theory of auctions and negotiations, scalable web server architectures, information retrieval,web databases and the like. Involving such a rich and diverse set of disciplines, E-markets are certainly an exciting and challenging delight for researchers, academiicians, software architects, designers and business professionals. You can download the presentation from the website.

His student Mr Rajgopal explained the design using UML and implementation of Intelligent Market with Agent based Integrated negotiations (Imagine). where the aim is to achieve social welfare of the e-market, by achieving the right balance between different producers and determining the prices based on the demand. The model can also account for several preferences of the customer through a value function.

Tutorial-2:"Recent trends in mixed signal processing for telecom applications" by R Krishnan, Texas Instruments.
Signal processing ICs are the key enablers fo broadband communication through ADSL or cable modems. Apart from a very high sampling rate for ADC and DAC, low noise and distortion requirements make the design of mixed signal chips a challenging task. Optimum design of the signal processing channel in a mixed signal chip involves the careful choice of the oversampling ratio to reduce the silicon area and power consumption. Varying the oversampling ratio calls for architectural tradeoffs like sigma-delta converter to a Nyquist rate converter and analog filtering vs digital filtering, partition between DSP chip and mixed signal chip. These are examined with an example design of a mixed signal chip for Home phoneline networking alliance.


Report on 9th Annual symposium
---


Section Home
-
-
- -
© Copyright 2000, IEEE.   Terms & Conditions.  Privacy & Security Small IEEE Logo
(Arjuna Rao, Chavala)
URL: https://www.ewh.ieee.org/r10/bangalore/ebcast/ebcast00/sridhar.html
(Modified:18-Jun-2000)