Dispersed spectrum sensing and receiver diversity techniques in cognitive radio networks

In cognitive radio (CR) systems, due to the inefficiency in spectrum usage, some approaches are available proposing the usage of frequency band dynamically instead of allocating static carriers. They all take an important part in wireless communication standards developed in the recent years. Using dynamic spectrum access methods is a new generation radio communication technology. The CR system monitors the whole frequency spectrum and is aware about the spectrum usage based on the specific time, location and frequency and it also defines intelligent policies to exploit the existing spectrum resources. In a major part of this thesis, the performance analysis of CR networks over additive white Gaussian noise (AWGN) and fading channels is studied. Besides, the system models are investigated to increase the optimal and more efficient usage of the resources in the radio frequency spectrum. In this thesis non-cooperative and cooperative communication systems are analysed. In addition, it is indicated that the system performance can be increased using the receiver antenna diversity techniques both in non-cooperative and cooperative models. In this context, the thesis consists of two main parts. In the first part of the thesis, the performance of the detection probability in the dispersed frequency spectrum over AWGN and Rayleigh, Rician, Log-Normally distributed fading channels for the non-cooperative communication systems using energy detector (ED) method in CR networks is investigated. The increase in the utilization of the dispersed frequency spectrum and in the detection probability for non-cooperative spectrum sensing is realized with the chirp-z transform (CzT) based algorithms. Besides, it is denoted that, the ED based spectrum sensing performance is increased in CR networks over Rayleigh fading channels with multiple antennas in single-input multi-output (SIMO) system model using selection combining (SC), equal gain combining (EGC), and maximal ratio combining (MRC) techniques in the receiver side. Then general closed form expressions for the optimal detection threshold values, minimizing the total error rate and increasing the system performance over AWGN and Rayleigh, Rician, Log-Normally distributed fading channels are derived. In the simulations, it is shown that the proposed dispersed chirp-z transform (DCzT) algorithm has considerably better performance than the conventional FFT-based approach especially for low signal-to-noise ratio (SNR) values. Then, the simulation results show that the MRC scheme has considerably better performance than the others. The accuracy of the results based on the theoretical analysis is verified by means of the simulation results. As a last item, the proposed system model with DCzT algorithm may be associated with the 4th generation (4G) wireless communication standard called long term evolution (LTE) is presented in the first part of this thesis. The investigation of the spectrum utilization effects in the cooperative communication systems constitutes the second part of this thesis. In the cooperative sytem model, using SC and MRC receiver diversity, the general closed form expressions for the optimal detection threshold values over Rayleigh fading channels are derived. In the proposed single-relay cooperative communication scheme, the theoretically derived expressions for the probability density function (pdf) and the detection probability (Pd) in SC method are presented. In the simulations, it is shown that the spectrum sensing performance for cooperative communication is much better than the non-cooperative one for different SNR values. Detection performance is increased while SC, EGC, and MRC diversity techniques are used in the receiver side as compared to the no diversity case. The simulation results are presented for different SNR values using the optimal detection thresholds. Finally, we verify our theoretical analysis by means of simulation results. Our analysis reveals that the theoretical results confirm a good agreement between the simulation results, validating the accuracy of the derived analytical derivations.