الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 150 |  |  |
| | | from | 150 | | |
Abstract
Cognitive radio (CR) is one of the most important emerging technologies that is developed to accommodate high data rate expected in future wireless networks. Devolution of efficient spectrum sensing (SS) techniques is required for effective implementation of CR networks. In this thesis, a multi-dimensional small-scale cooperative SS approach based on employing of multiple antenna elements (MAE), matched filtering (MF), and direction of arrival (DOA) estimation is proposed. The multi-dimensional sensing is performed by identifying both the primary user (PU) signal and its DOA, which corresponds to the temporal domain and spatial domain, respectively. The utilization of MAE provides many versions of the received PU signal, which considered as the core of many signal processing algorithms such as DOA estimation of PU signals. DOA estimation depends on calculating time difference of arrival (TDOA) between the output correlation signals that come from fast convolution blocks. The received PU signal versions are also exploited as inputs for separate MF detectors to obtain different decisions taking the advantage of the spatial diversity at the CR receiving antennas. Furthermore, the estimated DOA of the PU signal is involved to make a combination of the signal versions generating from each antenna branch, applying the delay and sum beamforming to produce an impregnable signal with high signal to noise ratio (SNR). This strong signal is also used as input to a separate MF to obtain a precise decision. Finally, the concept of cooperation is achieved by aggregation these various decisions to make a reliable decision. Several comparison scenarios are carried out to verify the supremacy of the proposed SS approach compared to the conventional MAE based SS technique under extremely low SNR and high interference. Besides, the simulations and derived theoretical expressions are highly congruous. IV Over the last decades, digital beamforming (DBF) is used for performance enhancement of uniform linear array (ULA) based spectrum sensing techniques. In general, a ULA suffers from high side lobe level (SLL) and low realized array gain. Therefore, the ULA-based CR frontend exhibited a low detection performance. Moreover, the attempts to increase the realized array gain are still not satisfying to increase the detection performance, because the problem of high SLL is still a challenging issue. In this thesis, a highly efficient technique is proposed for performance enhancement of ULA-based SS techniques using side lobe level reduction (SLLR) beamforming that is implemented on the receiving ULA. It is worthy mentioned that, a maximum SLLR combined with keeping the same half power beamwidth (HPBW) as the original array, results in a synthesized array with a dramatically improved gain. Hence, both the received SNR and signal-to-interference plus noise ratio (SINR) of the combined received signal are considerably enhanced at the CR terminal, which greatly increase the detection ability of the proposed CR system. The proposed approach is developed by integrating the SLLR-based antenna array with the generalized likelihood ratio test/direction of arrival (GLRT/DOA)-based SS technique that is denoted as GD technique. In contrary of the ULA, the excitation coefficients of the synthesized array are non-uniform, required a special signal processing to include it in the received system model and derive closed-form expressions of the probability of detection, probability of false alarm, and their associated decision threshold, all of which are already discussed in this work. Several simulation scenarios are done to approve the validity of the proposed SS approach over the classical single antenna element based SS techniques and other state-of-the-art beamforming based SS techniques.