Nemo Semret, PhD candidate, COMET group
http://www.ctr.columbia.edu/~nemo

Open Spectrum Access

The problem of spectrum allocation lies at the intersection of network control and telecommunications policy. By taking an economic (market based and game theoretic) approach in the engineering of the system, we seek solutions where the intelligence and decision making is distributed and thus scalable, and the policy objective of a more efficient and fair utilization of a public resource results from the induced market dynamics.

Use of Tesbed

The Open Spectrum Access system is deployed on a platform composed of the Java programming language and the Internet's World Wide Web, a combination which is emerging as a standard for distributed and open systems.

Deliverables

  • TREX: a prototype Open Spectrum Access system in the form of a dynamic auctioning game which users anywhere on the Internet can instantly access and easily participate in.
  • Wireless network performance analysis based on realistic/real traffic demands.
  • Analysis of the economic efficiencies and the optimality of the spectrum allocation resulting from the system.
  • Summary

    The allocation of resources can be viewed as an exchange of commodities. The seeds of this idea are already in place in the world of radio spectrum allocation with the FCC beginning to use auctions to allocate licenses. We are now suggesting that auctions are too limiting because of the static ownership-like nature of the result, and that it be taken a step further, to the concept of Open Spectrum Access, which is essentially dynamic auctioning, where spectrum is allocated on a continual basis. The result will be open access with a market based congestion fee.

    In the prototype under development, the exchange of spectrum for a given geographic area takes place in a virtual clearing house, which is implemented as a server program. Participants are located anywhere on the internet, and bidding is performed by a software agent which is a Java applet (a program embedded in an HTML document) that users obtain simply by going to a specific page on the WWW. The applet is run locally and seamlessly by their browser, and can be controlled manually or automatically by an intelligent bidding algorithm, and monitored via a graphical user interface.

    The behaviour of the software agent and the economic analysis of the system are based on state-of-the-art game theory. From the policy perspective, this dynamic and open appproach to spectrum allocation will be evaluated in comparison to the current licensing and static auctioning of spectrum.


    Participants: Aurel Lazar, Eli Noam, Nemo Semret, S. Munni Ullah, Alex Wolfson, Di Zhang.