End-to-end DWDM optical link power-control via a Stackelberg revenue-maximizing model

This paper deals with a Stackelberg formulation for the power control problem in optical networks. The new model adds an extra dimension to the recent OSNR game model and gives flexibility in optimizing the network performance (e.g. OSNR) and regulate the network conditions (e.g. power capacity). We 'engineer' the Stackelberg player to be a revenue-maximizing agent who designs pricing policies with complete information of the followers to optimize his own profit. We investigate both the unconstrained Stackelberg model and the one subject to capacity constraints, and characterize their solutions analytically. Finally, we use geometric programming and coordinate descent method to design a distributed and iterative algorithm that is suited to the Stackelberg model implementation in optical networks.