|
|
algorithms which converge to stable allocations which minimize network
interference.
Game theory provides the analysis tools needed to ensure your cognitive radio
network behaves in a predictable, optimal manner instead of producing networked
chaos. While game theory is a topic not encountered in traditional engineering
curricula, we believe it is critical to successfully implementing cognitive radio as
reflected in our long history developing the application of game theory to the design
of cognitive radio networks. Researchers at CRT wrote the first paper which
proposed the use of game theory to analyze cognitive radio networks, wrote the first
paper which showed how game models can be used to predict convergence and
stability properties of networks.
Our long academic involvement with game theory led to the creation of several
short game theory tutorials which we can leverage to build a short course tailored
to the needs of your organization. Topics covered by CRT in past game theory
tutorials at DySPAN07 and the Wireless@VT Summer School include:
- Introduction to Game theory (basic game models, review of fundamental
mathematical concepts, relationship between game theory and cognitive
radio networks)
- Equilibrium concepts (Nash, mixed-strategy equilibria, bargaining games,
Shapley values, Nash bargaining solutions)
- Equilibrium evaluation (Pareto efficiency, notions of fairness)
- The Notion of Time and Imperfections in Games and Networks (extensive
form games, repeated games, asynchronous repeated games, trembling
hand models, noisy observations)
- Designing Cognitive Radio Networks to Yield Desired Behavior
(Punishment and Reward in Networks, Potential Games and Implicit
Cooperation, supermodular games)
- Directions for future research and regulation
Recognizing that different companies have different backgrounds, we tailor each
tutorial to the specific needs of each customer. For example, we adjusted the
preceding topics to produce for L3-CSW an expanded 2-day tutorial with the
following outline.
More information about the tutorials and engineering services offered by CRT is
available on our Services page.
|
|
|
|
|
|
|
Unlike traditional radios, when cognitive radios are
deployed in a network, the radios' conflicting goals lead to
contention for networking resources. If not properly
accounted for, this contention can lead to network instability,
convergence to catastrophic resource allocations, or
unpredictable behavior. However, as CRT has shown with
its patents pending "zero-overhead" algorithms, this same
distributed behavior can be managed to yield low-complexity
|
|
|
|
|
