Home  »   Program Information  »  Eigen-inference analysis methods for cognitive radio News Sponsors
European Wireless 2010 - LUCCA  
Registration:
On-line Registration Form


Important Deadlines:
Paper Submission
November 8, 2009
Extended Deadline

Notification of Acceptance
January 25, 2010

Camera-Ready Version Due
February 28, 2010

Proposals for Tutorials
December 31, 2009

Notification of Acceptance for Tutorial Proposals
February 15, 2010

Eigen−inference analysis methods for cognitive radio

Tutorial summary

It is acknowledged today that the success of future generation wireless networks, expected to reliably transmit gigantic information content, will require a major revolution in the way data are exchanged. As such, cognitive radios are foreseen as a promising field of research which intends to take over the interference limitation and resource misuse of current wireless systems by making devices smarter. However, for a device to be smart, i.e. for it to be aware of its complex (multi−cellular, multi−user, multi−antenna etc.) environment, it has to gather, process and feed back a lot of information. It is therefore essential to identify the key parameters that characterize such complex wireless networks so to minimize processing and feedback burdens, hence making cognitive radios more efficient and more appealing to industrial applications. It has been recently recognized that the theory of large random matrices is able to (i) analyse the achievable rate performance of complex large wireless networks, (ii) reduce the parameters of such systems to their simplest granularity (by exhibiting those key parameters), (iii) provide computationally inexpensive methods to infer information on the surrounding environment in a blind manner and (iv) provide capacity−optimal feedback−based solutions for decentralized networks to be capable of self−organization. The purpose of this tutorial is to introduce the main tools of random matrix theory and show how they apply to cognitive radio problems in large wireless networks. The objectives of this tutorial are therefore for the attendees to acquire a rigorous methodology to treat large dimensional problems and to bridge both Random Matrix Theory and Cognitive Radio into a novel heuristic framework. The original features of this tutorial encompass:
  1. A rigorous introduction of random matrix theory and the recent breakthroughs achieved by this field of research in both the theoretical analysis and the practical solutions for large wireless networks.
  2. A discussion on the definition of a smart wireless device, based on the information−theoretical principle of maximum entropy. Based on this definition, optimal solutions for channel inference and signal sensing will be proposed; however, this definition will be shown to encounter major difficulties both in terms of mathematically possible system analysis and in the extent of its realisable practical applications.
  3. An answer to the difficulties pointed out in the previous section, based on the tools of large random matrices. We will show how the study of such matrices allows one to alleviate the aforementioned problems and we will specifically provide the example of self−organizing wireless networks that are shown to achieve optimal transmission rates at the expense of limited information feedback.

Speakers Resumes