Due to the scarcity of frequency spectrum in view of the evolution of wireless communication technologies, the cognitive radio (CR) concept has been introduced to efficiently exploit the available spectrum. This concept consists in introducing unlicensed/secondary users (SU’s) in existing networks to share the spectrum of licensed/primary users (PU’s) without harming primary communications hence the name of “spectrum sharing” technique. We study in this dissertation, the capacity and the achievable rate of the secondary user within various communication settings. We, firstly, investigate the capacity of the (SU’s) at low power regime for Nakagami fading channels and present closed form of the capacity under various types of interference and/or power constraints. We explicitly characterize two regimes where either the interference constraint or the power constraint dictates the optimal power profile. Our framework also highlights the effects of different fading parameters on the secondary link ergodic capacity. Interestingly, we show that the low power regime analysis provides a specific insight on the capacity behavior of CR that has not been reported by previous studies. Next, we determine the spectral efficiency gain of an uplink CR Multi-Input Multi- Output (MIMO) system in which the SU is allowed to share the spectrum with the PU using a specific precoding scheme to communicate with a common receiver. Applied to Rayleigh fading channels, we show, through numerical results, that our proposed scheme enhances considerably the cognitive achievable rate. For instance, in case of a perfect detection of the PU signal, after applying Successive Interference Cancellation (SIC), the CR rate remains non-zero for high Signal to Noise Ratio (SNR) which is usually impossible when we only use space alignment technique. In addition, we show that the rate gain is proportional to the allowed interference threshold by providing a fixed rate even in the high SNR range. Finally, we study the impact of the broadcast approach and multi-layer coding on the throughput of CR systems for general fading channels. And we found that at the absence of the channel state information(CSI), we show that this improvement could be almost reached by 2-Layers coding. Then, we introduce a quantized CSI policy and highlight its improvement in terms of throughput before we study the rate when BA with quantized CSI is adopted. Numerical results show that the improvement of the additional layers is decreasing as the number of quantized regions increases.

In a MIMO system, a transmitter with perfect knowledge of the underlying channel state information (CSI) can achieve a higher channel capacity compared to transmission without CSI. When reciprocity of the wireless channel does not hold, the identification and utilization of partial CSI at the transmitter are important issues. This thesis is focused on partial CSI acquisition and utilization techniques for MIMO channels. We propose a feedback algorithm for tracking the dominant channel subspaces for MIMO systems in a continuously time-varying environment. We exploit the correlation between channel states of adjacent time instants and quantize the variation of channel states. Specifically, we model a subspace as one point in a Grassmann manifold, treat the variations in principal right singular subspaces of the channel matrices as a piecewise-geodesic process in the Grassmann manifold, and quantize the velocity matrix of the geodesic. We design a complexity-constrained MIMO OFDM system where the transmitter has knowledge of channel correlations. The transmitter is constrained to perform at most one inverse Discrete Fourier Transform per OFDM symbol on the average. We show that in the MISO case, time domain beamforming can be used to do two-dimensional eigen-beamforming. For the MIMO case, we derive design criteria for the transmitter beamforming and receiver combining weighting vectors and show some suboptimal solutions. The feedback channel may have uncertainties such as unexpected delay or error. We consider channel mean feedback with an unknown delay and propose a broadcast approach that is able to adapt to the quality of the feedback. Having considered CSI feedback problems where the receiver tries to convey its attained CSI to the transmitter, we turn to noncoherent coding design for fast fading channels, where the receiver does not have reliable CSI. We propose a data-dependent superimposed training scheme to improve the performance of training based codes. The transmitter is equipped with multiple training sequences and dynamically selects a training sequence for each data sequence to minimize channel estimation error. The set of training sequences are optimized to minimize pairwise error probability between codewords.