Real option and game theoretic approach to telecommunication network optimization.

Abstract

This dissertation introduces a new method for telecommunications network optimization based on real options and game theory. The model is applied to routing optimization, capacity expansion, and resource allocation problems. First, we study network routing and capacity expansion in a telecommunications network. The real option theory is used to model network delay/blocking under the correlated network demands. This gives an analytical representation for network costs. Based on the costs, we solve optimal network routing by using a global optimization method. And we suggest a new network capacity expansion method that depends on the optimal routing. Our model implies that if the network point-to-point demands have high volatility and negative or low correlations then there is a significant benefit from a complex network structure with multiple nodes, because it has many routing options that have high value in this case. Further, due to the option value in this case there is a high need for network capacity expansion. We also show that the average blocking/delay cost and capacity expansions are concave functions of the demand volatilities. Second, we study wireless telecommunications network systems. We formulate game models for resource allocation in Time Division Multiple Access (TDMA) and Code Division Multiple Access (CDMA) systems by using a mixed strategy equilibrium. In the game models, each user minimizes his/her power and connection costs. The users' activities are analyzed by using simulations. The results imply that the TDMA game model improves the systems' performance significantly regardless of the power cost. However, due to the near-far problem in the CDMA system, the power cost control is critical to maintain its efficient resource allocation.