Optimal Transit Price Negotiation: The Distributed Learning Perspective

We present a distributed learning algorithm for optimizing transit pricesin the inter-domain routing framework. We present a combined game theoretical anddistributed algorithmic analysis, where the notion of Nash equilibrium with the firstapproach meets the notion of stability in the second. We show that providers canlearn how to strategically set their prices according to a Nash equilibrium; even whenassuming incomplete information. We validate our theoretical model by simulationsconfirming the expected outcome. Moreover, we observe that some unilateral deviationsfrom the proposed rule do not seem to affect the dynamic of the system.