A generalized order user selection scheme for cognitive networks with imperfect channel estimation and multiple primary users

In this paper, we study the performance of multiuser cognitive networks with multiple primary users and imperfect channel estimation. Two scenarios are considered: primary users with the same spectrum band and primary users with orthogonal spectrum bands. The utilized generalized order user selection scheme is efficient in situations where a user other than the best user (user of best channel) is erroneously selected by the scheduling unit for conducting its communication as an imperfect channel estimation or an outdated channel information condition. In this scheme, the secondary user with the second or even the Nth best signal-to-noise ratio (SNR) is assigned the system resources by a scheduler. In our paper, closed-form expressions are derived for the outage probability, average symbol error probability (ASEP), and ergodic channel capacity of the secondary system assuming imperfectly estimated Rayleigh fading channels. Furthermore, closed-form expressions are derived for the outage probability, considering the interference from a primary transmitter. Also, to get more insights about the system performance, the system is studied at the high SNR regime where the diversity order and coding gain are derived and analyzed. The achieved analytical and asymptotic results are verified by Monte-Carlo simulations. The main results illustrate that the number of primary users affects the performance of the secondary system through affecting only the coding gain and not the diversity order. Results show that in the case where the primary users use the same spectrum band, the system performance enhances as the number of primary users decreases and degrades as the number of primary users increases. In contrast, when the primary users use orthogonal spectrum bands, findings illustrate that the system performance degrades as the number of primary users decreases and enhances as the number of primary users increases. Also, results show that a zero diversity gain is achieved by the system and a noise floor appears in the results when the secondary users channels are imperfectly estimated.