Systems Science Seminar|
A Non-cooperative Perspective of Cooperative Communications
University of Southern California - Department of Electrical Engineering
Thursday, January 27, 2011|
4:00pm - 5:00pm
About the Event
Wireless networks are going through a phase of explosive growth. New devices and applications have increased manifolds the demands of capacity, reliability and quality placed on such networks. However, building of new infrastructure alone is not going to satiate these demands. There is a fundamental limitation, namely, the amount of spectrum, and more must be done with the same. Among other things, cognitive radio technologies will have to be deployed to make more efficient use of currently available spectrum, as well as cooperative communication schemes will have to be employed. In the first part of the talk, we look at cognitive radio systems in licensed bands. Most of the work on such systems has focused on cognitive sensing, and it is generally presumed that primary licensees of wireless spectrum will willingly share spectrum with secondary wireless devices. This cannot be taken for granted unless the primary spectrum owners have an incentive to do so. This could happen, if for example, they got compensated for sharing spectrum and suffering interference. We, thus, look at incentive mechanisms for sharing spectrum while using cooperative communication schemes. We give formats of incentive-compatible contracts that can be used under both complete and incomplete information, that are robust to the moral hazard problem. We show that incomplete information (i.e., not knowing channel coefficients of the secondary user) does not present a significant impediment in implementing socially optimal contracts. In the second part of the talk, we look at the problem of (non-cooperative) relaying in wireless networks. Relaying is seen as one of the means of increasing capacity of wireless networks. Thus, various cooperative communication schemes have been proposed that achieve close to the information-theoretic network capacity. The premise behind such schemes is the cooperative behavior of users. However, given limited resources such as battery power, this cannot be taken for granted, and users may be selfish and care only about their own rates. We study the incentive problem in the setting of a generalized Gaussian relay (MAC-GF) channel, and consider that users use a cooperative communication scheme proposed by Sendonaris, et al. We show that under certain channel conditions, operating on a part of the Pareto-optimal boundary of the achievable rate region with this scheme is a Nash equilibrium, and in fact these are the only Nash equilibria.
Rahul Jain is an Assistant Professor in the EE & ISE Departments at the University of Southern California, Los Angeles, CA. He received his B.Tech in EE from the Indian Institute of Technology, Kanpur, an MS in ECE from Rice University, and an MA in Statistics and a PhD in EECS from the University of California, Berkeley. Prior to joining USC in Fall 2008, he spent two years at the IBM T J Watson Research Center, Yorktown Heights, NY. He has received numerous awards including the NSF CAREER award in 2010, an IBM Faculty award in 2010, a Best paper award at The Value Tools Conference 2009, and a James H. Zumberge Faculty Research and Innovation Award in 2009. His interests span communication network analysis with focus on network economics and game theory, stochastic models, optimization and learning.
Contact: Ann Pace
Sponsor: University of Michigan
Open to: Public