The Allocation in Cognitive Radio Network: Combined Genetic Algorithm and ON/OFF Primary User Activity Models

摘要

Cognitive radio (CR) has appeared as a promising solution to the problem of spectrum underutilization. Cognitive radio user (CU) is an intelligent equipment who scent the spectrum which is licensed to primary radio users (PUs) when it is idle and use it with other CUs for their communication. Thus by modeling PUs activity, CUs can predict the future state ON or OFF (busy or idle) of PUs by learning from the history of their spectrum utilization. In this manner, CUs can select the best available spectrum bands. On this point, many PU ON/OFF activity models have been proposed in the literature. Among this models, Continuous Time Markov chain, Discrete Time Markov chain, Bernoulli and Exponential models. In this paper, we firstly compare these four models in term of better numbers of OFF slots to deduce which model give best performance of available resources. Then, the activity history patterns generated from each model are combined with the genetic algorithm as sensing vectors to select the best available channel in terms of quality and least PU arrivals.