Decision and Cognition

Instructor.

Motoya Machida, Tennessee Technological University.
E-mail: mmachida tntech.edu

Alexander Shibakov, Tennessee Technological University.
E-mail: ashibakov tntech.edu

Website.

math.tntech.edu/machida/cognition (includes a link to this syllabus in Japanese)

Description of the course. The aim of this short course is to provide an introduction to the mathematical aspects of an exciting new frontier in telecommunications: cognitive radio. Enabled by the latest developments in digital and analog hardware, as well as new spectrum regulations considered or already implemented by various countries, the cognitive (or AI enhanced) radio has the goal of dramatically improving spectrum utilization by allowing each radio transceiver to make its own decisions as to when to occupy an unused portion of unlicensed spectrum.

This radically new technology presents unique problems to the designers, ranging from purely hardware related to regulatory challenges to the problem of accurately modeling and predicting spectrum use. This course will concentrate on the mathematical challenges of spectrum modeling and prediction.

To provide a meaningful introduction for a mathematics oriented listener, a portion of the course will be devoted to an overview of the traditional telecommunication infrastructure. Some history and technical aspects will be mentioned as well. The classical apparatus of integral transforms and information theory will be presented and the core concepts, such as noise and bandwidth, will be discussed.

The remainder of the course will be spent on the survey of the mathematical challenges of cognitive radio. We will discuss the major directions in the existing body of research that aims to achieve the leap in spectrum optimization promised by the cognitive radio framework, starting with the proposed ways of modeling the `background spectrum'. These approaches range from the conservative ones based on the Poisson process to the more ambitious strategies based on hidden Markov models, and game theoretic approach.

The course concludes with a brief discussion of the full multiuser environment a typical cognitive transceiver will likely find itself in. The enormous mathematical challenge of designing a decision algorithm for such a device is outlined and some directions for future research will be mentioned.

Organization, assignment and grading. The lecture will be given in English. Attendance is required and reviewed for the grade, and a short report will be asked and completed at the end of the course. An excused absence must be arranged with the instructor in advance in order to avoid the adverse affect for grading.