الفهرس 
Cognitive CycleOnly 14 pages are availabe for public view
 |  | from | 122 |  |  |
| | | from | 122 | | |
Abstract
Cognitive Radio Networks (CRNs) are classified into three main categories according to
the way in which a secondary network accesses the spectrum of a primary network. Firstly,
the Opportunistic Spectrum Access (Overlay), in which secondary users (SUs) sense the
presence of primary users (PUs) and communicate only over the vacant bands. Secondly,
the Spectrum Sharing (Underlay), in which SUs share all the spectrum band with PUs.
Finally, Sensing-Based Spectrum Sharing (Interweave), which is a combination of the
previous two categories.
In this thesis, the problem of maximizing the secondary throughput under the transmit
power and interference power constraints for the three types of CRNs is considered.
Firstly, a proposed optimal power allocation scheme using genetic algorithm (GA) is
presented for a multiple-input-multiple-output (MIMO) system in CRN. This scheme is
used to maximize the secondary throughput under interference constraints in a system model
of multiple SU pairs coexisting with multiple PU pairs in an underlay spectrum sharing
network. For the sake of comparison, the minimal throughput among all SUs is compared
with the other power allocation schemes, namely, maximum-minimum-throughput-based
power assignment (MMTPA) and equal power assignment (EPA).
Secondly, an optimal power allocation strategy is used to maximize the secondary
throughput under power constraints in overlay cooperative CRNs. The effect of using
multiple antennas at the relay is studied and the results are compared with using single
antenna. Throughput maximization is addressed by simply seeking an optimum value of the
relay amplification factor.
Finally, sensing-based spectrum sharing is considered for cooperative CRNs. The
secondary throughput is maximized under power constraints by obtaining an optimum value
of the relay amplification factor. Comparison between using single antenna, double antennas
at the relay node, and double antennas at all the nodes of the cognitive relay network is
performed. Furthermore, the sensing-based spectrum sharing is compared with the
opportunistic spectrum access in the case of using double antennas at all the nodes of the
cognitive relay network.