Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)

First Advisor

Manjunath Hegde


We use shadow prices to measure the rate of change of performance functions of a communications network with respect to incremental changes in one or more of the design parameters such as capacity, anticipated demand, reservation levels, routing proportions, etc. The ability to evaluate such shadow prices allows us to introduce designs that optimize network performance according to network-wide criteria. We present models for the evaluation of shadow prices for different cases of networks: single-rate circuit-switched networks, wireless networks, multi-rate circuit-switched networks. For these cases, we demonstrate the wide applicability of shadow price methods. We formulate the new and very useful notion of sum capacity that can be used to compare different adaptive routing schemes, we can determine the actual pricing of calls to reflect their effect on the entire network, we trade off new call blocking with the forced termination of calls due to handoff drops in wireless networks, we dimension networks to achieve a prescribed set of new call and handoff blocking probabilities, we determine the tradeoffs between calls of different rates in multi-rate networks. We also point out problems of current interest which can be addressed by the shadow price framework and outline a brief methodology.