Performance evaluation and game theoretic pricing of optimal routing and flow assignment in optical networks

Transport service providers need control and optimization strategies for wavelength management, network provisioning, restoration, and protection allowing them to define and deploy new services offers. In this thesis, we apply microeconomic models to investigate the behavior of optical networks in different market environments, and the effect of different pricing schemes in the profitability of optical networks. The routing and wavelengths allocation were modeled using network flows optimization. Network pricing decisions were modeled by applying a game theoretic approach, where the game payoff is the profit of operating the network. We have developed a customer/supplier interaction protocol for network competition control. Two models of competition were introduced. The first model is called "active price competition", allows the customer (wavelength buyer) to interact in order to lower the prices. The second model, is called "passive price competition", assumes that the customer is passive and does not interact with the competitors; the customer accepts the lowest offered price or does not buy even from the lowest bidder if the price is above the buyer's maximum set cost value. In this model, the competitors monitor their own profit and the customer requests success rate to determine their pricing strategy. In both models, we examined different case studies, comparisons and performance evaluations. A mixed integer optimization is employed to determine the wavelengths allocation and flow assignment of the requested customer demand.