Multiantenna Interference Mitigation Schemes and Resource Allocation for Cognitive Radio

Maximum and efficient utilization of available resources has been a central theme of research on various areas of science and engineering. Wireless communication is not an exception to this. With the rapid growth of wireless communication applications, radio frequency spectrum has become a valuable commodity. Supporting very high demands for data rate and throughput has become a challenging problem which requires innovative solutions. Dynamic spectrum sharing (DSS) based cognitive radio (CR) is envisioned as a promising technology for future wireless communication systems, such as fifth generation (5G) further development and sixth generation (6G). Extensive research has been done in the areas of CRs and it is considered to mitigate the spectral crowding problem by introducing the notion of opportunistic spectrum usage. Spectrum sensing, which enables CRs to identify spectral holes, is a critical component in CR technology. Furthermore, improving the efficiency of the radio spectrum use through spectrum sensing and dynamic spectrum access (DSA) is one of the emerging trends. In the first part of this thesis, we focus on enhancing the spectrum usage of CR’s using interference cancellation methods that provides considerable performance gains with realistic computational complexity, especially, in the context of the widely used multicarrier waveforms. The primary focus is on interference rejection combining (IRC) methods, applied to the black-space cognitive radio (BS-CR). Earlier studies on the BS-CR in the literature were focused on using CRs as repeaters for the primary transmitter to guarantee that the CR is not causing significant interference to nearby primary users’ receivers. This kind of approaches are transmitter-centric in nature. In this thesis, receiver-centric approaches such as multi-antenna diversity combining, especially enhanced IRC methods, are considered and evaluated. IRC methods have been widely studied and adopted in several practical wireless communication systems. We focus on developing such BS-CR schemes under strong interference conditions, which has not been studied in the CR literature so far. Spatial covariance matrix estimation under mobility and high carrier frequencies is found to be the most critical part of such scheme. Algorithms and methods to mitigate these effects are developed in this thesis and they are evaluated under realistic BS-CR receiver operating conditions. We use sample covariance estimation approach with silent gaps in the CR transmisison. Covariance interpolation between silent gaps improves greatly the robustness with time-varying channels. Good link performance can be reached with low mobility at carrier frequency considered for the TV white-spaced case. The proposed BS-CR scheme could be feasible at below 6 GHz frequencies with pedestrian mobilities. The second part of this thesis investigates the effect of radio frequency (RF) impairments on the performance of the cognitive wireless communication. There are various unavoidable imperfections, mainly due to the limitations of analog high-frequency transmitter and receiver circuits. These imperfections include power amplifier (PA) non-linearities, receiver nonlinearities, and carrier frequency offset (CFO), which are considered in this study. These effects lead to significant signal distortion and, as a result of this, the wireless link quality may deteriorate. In multicarrier communications such signal distortions may lead to additional interference, and it is important to evaluate their effects on spectrum sensing quality and on the performance of the proposed BS-CR scheme. This part of the thesis provides critical analysis and insights into such issues caused by RF imperfections and demonstrates the need for designing proper compensation techniques required to avoid/reduce such degradations. It is found that the transmitter’s PA nonlinearities affect in the same way as in basic OFDM systems and BS-CR receiver’s linearity requirements are similar to those for advanced DSP-intensive software defined radios. The CR receiver’s CFO with respect to the PU has the most critical effect. However, synchronizing the CR with the needed high accuracy is considered achievable due to the PU signal’s high-power level. The final part of the thesis briefly looks at alternate waveforms and techniques that can be used in CRs. The filter bank multicarrier (FBMC) waveforms are considered as an alternative to the widely used OFDM schemes. Here the core idea is interference avoidance, targeting to reduce the interference leakage between CRs and the primary systems, by means of using a waveform with good spectrum localization properties. FBMC system’s performance is compared with OFDM based system in the context of CRs. The performance is compared from a combined spectrum sensing and resource allocation point of view through simulations. It is found that well-localized CR waveforms improve the CR link capacity, but with poorly localized primary signals, these possibilities are rather limited