Spectrum Sharing: System Design Perspective

Speaker: Danijela CabricUniversity of California, BerkeleyA major shift in wireless communications is now emerging with the development of cognitive radios, which attempt to share spectrum in a fundamentally new way. Cognitive radios address the problem of poor spectrum utilization exhibited in many frequency bands. On a conceptual level, cognitive radio networks sense the spectral environment and adapt transmission parameters to dynamically reuse available spectrum. The novelty of this approach requires us to re-architect the mechanisms for using radio frequencies and find a way for multiple systems to co-exist through sharing rather than fixed allocations.This talk addresses fundamental questions in cognitive radios system design and investigates its feasibility by bridging the theoretical and practical aspects of the physical and network layers. We begin with spectrum sensing, the key enabling functionality for cognitive radios, which requires detection of very weak signals of different types in a minimum time with high reliability. We show that the biggest barrier for sensing in very low signal to noise ratio regimes is the variability in the noise and interference that cannot be perfectly calibrated during sensing time. Robustness can be increased by differentiating signals from noise by detecting signal features or by exploiting channel diversity with network cooperation. Through physical implementation and experiments of proposed sensing methods, we identify the minimum possible sensing times and detectable signal levels. Moreover, we include the sensitivities exhibited by these methods to radio and channel impairments. Next, we consider radio architecture for wideband spectrum sharing radios and show that large dynamic range requirements present major challenge in their implementation. This calls for the development of novel mixed signal techniques. By exploiting spatial dimension for selective processing of desired signals through antenna array architectures, strong interferers can be adaptively suppressed. We conclude with a discussion of the requirements for the signaling and protocol designs that support dynamic spectrum access and spectrum sensing coordination.Brief Biography:
Danijela Cabric is a Ph.D. candidate in Electrical Engineering at the University of California, Berkeley in the group of Prof. Robert Brodersen. She received the M.S. in Wireless Communications and VLSI System Design in 2001 from the University of California at Los Angeles and the Diploma in Electrical Engineering from the University of Belgrade, Serbia, in 1998.

Location: Hedco Neurosciences Building (HNB) - 100

Audiences: Everyone Is Invited

Contact: Ericka Lieberknecht