In Cognitive Radio technology, reusing of dedicated frequency bands in time and frequency domain (and even in special domain) is possible. To work efficiently, cognitive networks like all other communication networks needs some optimizations. One of the spectrum access models for cognitive network users is an interweave model. In this model, users of cognitive networks (secondary users) are not allowed to transmit data simultaneously with users of incumbent network. Indeed, secondary users sense the spectrum before sending any data, finding the channel idle, they start transmitting data. In co-existence between primary and secondary network through an interweave model, interference imposed to primary users due to secondary users’ transmissions is inevitable and should be hold in a certain level. On one hand, the mentioned interference emanates from spectrum sensing error and on the other hand, it is due to primary users’ returns during secondary transmissions. The amount of interference in the first case depends on sensing parameters and it depends on primary user traffic in the second case. Furthermore, with the advent of green communication in recent years, energy efficiency in communication networks has been considered widely as an important criterion. Hence, both energy efficiency and collision probability, as important criteria in cognitive network, are affected by the two kinds of interference. Therefore, mathematical modeling of secondary users’ spectrum access regarding to interference times for the sake of optimization goals is important. This dissertation analyzes secondary users’ spectrum access process regarding to primary and secondary traffic in single user and multi-user scenarios and formulates energy efficiency and collision probability. Subsequently, the effect of different parameters such as transmission time, primary users’ data traffic and spectrum sensing parameters in the mentioned secondary network criteria is investigated. At first, by modeling of the primary user’s traffic as two-state continuous time Markov chain, single-user secondary network with sporadic and heavy traffic is modeled and analyzed and then single-user cognitive network with general traffic is investigated. Afterwards, with the same model for primary user’s traffic, analyses of single-user secondary network are developed for multi-user network. Consequently, analyses of multi-user cognitive network based on common control are performed. Iiring the extracted methods and formulas, primary user’s traffic model is changed from two-state continuous time Markov chain to more general model, alternating phase type renewal process, and single-user secondary network with sporadic traffic is analyzed under.
