Click on the link below for more details

Monday, June 6, 2011
12:00 PM – 01:20 PM ∙ Rooms 307-308

Software Defined Radios - Facts and Fantasies

Chairs/ Moderators:
Lawrence Kushner, Intersil Corp.
Timothy Hancock, MIT Lincoln Laboratory

Panelists:
Geoff Dawe, BWS Consulting
James Kimery, National Instruments
Larry Larson, UCSD
Kamal Sahota, Qualcomm
Bruce Fette, DARPA
Vanu Bose, Vanu, Inc.

Sponsor:    RFIC

Abstract:

The concept of Software Defined Radios (SDR) originated decades ago in the defense sector, culminating in the development of a number of successful SDR demonstrations and deployments.  The flexibility of SDR, being able to serve a wide variety of changing radio protocols offers the military interoperability and maintainability not achievable with conventional radios.

Research in Software Defined Radios has accelerated during the past two decades, with work in universities, industry, and government advancing the concepts.  Simultaneously, the relentless march of Moore’s Law has made digital processing almost free, shifting more and more of the radio processing into the digital domain.  The “Holy Grail” of SDR, an antenna followed by an Analog-to-Digital Converter (ADC) and high-performance Digital Signal Processor (DSP) now seems within reach.  Similarly, one can conceive of the transmit path consisting of a DSP followed by a DAC and a power amplifier.  Do these architectures make sense?  Is SDR the best solution in terms of size, weight, power, cost, and cost-of-ownership or is reconfigurable conventional RF hardware with a standard software interface a better solution?

Our panel of experts will discuss and debate the current state-of-the-art of radio design, and how SDR fits in.  We will discuss what is a Software-Defined Radio, what applications are best suited for SDR, and where future SDR research is heading.  The audience will be encouraged to participate as well, submitting questions for the panel, engaging in the discussion, and voting, in real-time, “fact” or “fantasy” after each topic of debate using an rf audience response system.

Tuesday, June 7, 2011
12:00 PM – 01:20 PM ∙ Rooms 307-308

What is the limit of multi-radio integration… or rather, is it ‘disintegration’?

Panel Organizer:
R. Bogdan Staszewski, Delft University of Technology

Moderator:
Oren Eliezer, Xtendwave

Panelists:
R. Bogdan Staszewski,  Associate Professor, Delft University of Technology
Andre Hanke,   Senior Principal RF System Engineer, Intel
Walid Ali-Ahmad, Technical Director, MediaTek
Keith Carter,   WLAN Senior Manager, Broadcom
Fred Schindler,   Director, RFMD
Thomas Kazior,   Principal Engineering Fellow, Raytheon

Abstract:

Only a decade ago, single-chip RF-SoC integration was universally thought to be impossible or at least uneconomical. Nowadays, the pioneering days of single-chip radios are largely over and the innovation efforts are applied to integrating multiple radio cores on the same silicon die. This effort has already resulted in commercial offerings of multi-core wireless connectivity and cellular radios from a few companies, but has revealed some interesting RF co-existence issues: Integrating additional radio cores appears to exponentially increase the overall design and productization complexity, more so than in the case of isolated radios.  Why is that and what can be done to address that?  What is the ultimate limit of multi-core radio integration?
As new wireless standards continue to emerge, it becomes necessary to support additional frequency bands and wider modulation bandwidths, while maintaining backwards compatibility with the existing standards. This puts enormous pressure on the complexity and quality of RF front-end components (PA’s, T/R switches, band-pass filters and duplexers) to the point that they predominate in both cost and occupied space, which might suggest the reversal of the integration trend.  Can the ever multiplying antenna-interfacing components still be integrated?  Do they follow a different integration path from that of RF-SoCs?  Does the optimal system partitioning suggest the RF-SoC ‘disintegration’?