Resource Allocation and Precoding in Cognitive Radio and Relay Networks

Abstract

In this dissertation, two promising technologies which improve the spectrum efficiency in wireless networks are investigated: cognitive radio (CR) and relay technology. Cognitive radio, as a promising technology to dynamically access spectrum resources, has drawn great research attention recently. It provides a way to further improve the spectrum efficiency by allowing unlicensed radio devices to learn from the radio environment and adapt their transmit and receive parameters. This dissertation addresses the design of unlicensed networks, including their transmission scheme, resource allocation and precoding/beamforming design when multiple antennas are deployed. Another topic in this dissertation is that of relay networks. By introducing relay stations into the network, multiple benefits can be obtained, such as extended network coverage, improved throughput and higher spectrum efficiency. Also, a relay network makes it possible to enable two-way (even multi-way) transmission among multiple users within the network. In this dissertation, precoding design in multiuser relay networks is discussed. Also, networks based on combined cognitive radio and relay technologies are considered to leverage higher performance, in terms of spectrum efficiency and network throughput. This dissertation is organized into six chapters. Chapter 1 introduces the background of cognitive radio and cooperative relay networks. In Chapter 2, a soft-decision spectrum sensing concept is proposed and based on which, a joint spectrum sensing, access and power allocation problem is addressed for multi-band cognitive radio networks by means of convex optimization. Chapters 3 and 4 deal with the precoding design in multi-user cognitive relay networks. Chapter 3 considers the multi-way transmission among multiple users and adopts minimum mean square error (MMSE) as the design objective while Chapter 4 considers a two-way relay network and a joint signal and interference alignment algorithm is proposed. In Chapter 5, a signal group based interference alignment algorithm is proposed for a generalized MIMO Y channel where each user sends messages to all the other users. In Chapter 6, conclusions are drawn and possible future research avenues are highlighted. Some interesting ideas about how to improve the spectrum efficiency in wireless networks have been proposed and analyzed in this dissertation. The proposed soft-decision spectrum sensing method allows more flexibility in designing the radio access strategies in cognitive radio systems and achieves significantly higher throughput compared with a traditional hard decision spectrum sensing based algorithm. Furthermore, the proposed precoding/beamforming algorithms in the latter part of this dissertation enable concurrent transmission of multiple users within the same frequency band, which can significantly reduce or completely remove the inter-user interference. These technologies make it possible to utilize the limited radio resources more efficiently and therefore can support the ever-increasing demand arising from various wireless devices and applications. iii