الفهرس 
Conclusions and Suggested Future Works
Suggested Future WorksOnly 14 pages are availabe for public view
 |  | from | 135 |  |  |
| | | from | 135 | | |
Abstract
The 5G vision and its associated key requirements such as high data rate and massive capacity lead to a huge demand for the radio spectrum, the scarcity of frequency resources become a challenging problem in wireless communications. To face this problem, the intelligent use of available spectrum using cognitive radio is proposed and the regulatory bodies have allowed unlicensed users to use radio channel on the licensed users’ spectrum without interfering with the licensed users. Hence, new physical layer (PHY) designs and waveforms are being researched, which can fill in the TV white spaces (TVWS) in an opportunistic manner. To deal with fragmented spectrum, the modulation of the multicarrier waveform should support this characteristic. The requirements of the potential applications of 5G including machine-to-machine (M2M) communication and IoT cannot be supported by Orthogonal Frequency Division Multiplexing (OFDM) used in LTE which lead to design new waveforms. Generalized Frequency Division Multiplexing (GFDM) is one of waveforms candidate for 5G. GFDM is well suited for cognitive radio; it has a low out of band radiation. Spectrum sensing is the most important task in cognitive radio which detects the non-utilized frequency bands. Cyclostationary sensing is considered as the robust detector in low Signal to Noise ratio. It achieves a high probability of detection.
This thesis proposes the cyclostationary sensing of GFDM in cognitive radio transmission. The detection of narrowband channels using time smoothing algorithms, strip spectral correlation algorithm is time efficient algorithm when calculating the spectral correlation function of GFDM; it achieves the optimal detection of the signal in low SNR. The work is extended to detect GFDM in wideband using Sub-Nyquist sampling techniques. Considering the cyclostationarity properties of modulated signals, we propose an optimized recovery method for GFDM signal in the wideband regime. By exploiting the signal sparsity, we are able to recover the spectral correlation function (SCF) of GFDM from digital samples of GFDM taken at Sub-Nyquist rate to reduce the sampling time. Furthermore, the generalized likelihood ratio test is applied to the recovered function to detect multiple signal sources and identify the spectrum occupancy. Numerical results show that our method achieves a high probability of detection at low SNR and also the robustness of this method to the rate reduction in wireless networks. The performance of detection of GFDM outperforms the detection of OFDM in which indicates that GFDM is a strong contender for 5G and Cognitive radios.