Methods to evaluate the performance of multiple uncooperative users for green cognitive radio networks

Abstract

In this paper, we propose a method to calculate the throughput of multiple uncooperative cognitive radio users. We use a bank of Markov models to derive the fraction of time that each user spends in successful channel access. The mutual interference among the users is modeled into the transitional probabilities of the Markov models, which makes it tractable to calculate multiple user throughput in large cognitive radio networks (CRN). In addition, to evaluate the level of greenness of the multiple user CRN we develop a method to optimize the ideal capacity of the fully cooperative CRN. This method relies on solving the sum-of-ratios linear fractional programming for optimization. Simulations are conducted to show the big gap between the uncooperative CRN performance and the ideal performance.