Research On Application Of Multi-Antenna Technology In Cognitive Radio Systems

The most remarkable character of wireless communication which differs from other communication methods is the openness of wireless channel. Wireless communication systems operating over the same frequency will interfere with each other. In order to avoid this co-channel interference, different wireless communication systems are usually deployed over different frequency ranges. Therefore, radio spectrum is invaluable resource. However, as the development of theories and technologies, new wireless communication systems are keeping emerging, and wireless services are expanding. Hence, existing radio spectrum resources is going into short supply day by day.Cognitive radio can significantly improve the utilization of radio spectrum, radically solve the problem of spectrum insufficiency, and therefore has attracted the attentions of scholars and researchers all over the world. On other hand, multi-antenna, or equivalently multiple input multiple output (MIMO) technology, which can improve the link capacity in multiples without any extra bandwidth by spatial multiplexing, has been the research focus of wireless communication recently. Both cognitive radio and multi-antenna are effective technologies for spectrum efficiency improvement. And using multi-antenna technology in cognitive radio systems can significantly improve spectrum utilization further.Key technologies in cognitive radio mainly include three aspects: quick and accurate spectrum sensing, adaptive transmission and spectrum management. This dissertation has systematically studied the multi-antenna technology in cognitive radio systems, which belongs to the category of adaptive transmission. By means of multi-antenna technology, the interference to primary network induced by secondary network can be completely avoided or reduced by keeping primary network and secondary network orthogonal or semi-orthogonal in space dimension. In this way, both primary network and secondary network can operate over the same frequency spectrum simultaneously. Hence, spectrum efficiency can be improved substantially. In this dissertation, cognitive radio systems are divided into centralized systems and distributed systems two types and are studied separately. Main innovation points can be summarized as follows:Firstly, in centralized cognitive radio networks, a secondary network shares a spectrum with a primary network. In order to make primary and secondary network work simultaneously, secondary base station (BS) is equipped with multiple antennas. In addition to time and frequency dimensions, space dimension is introduced. Our goal is to maximize the capacity of secondary network on the condition of limiting the interference to primary network below the threshold of interference temperature. By joint consideration of beamforming, secondary user scheduling and power allocation three steps, a complete set of solution for downlink of secondary network has been presented. The computational complexity of the solution is analyzed and its performance is evaluated by simulation. And the simulation results indicate that the proposed solution is very robust to power allocation.Secondly, centralized cognitive radio network is considered again. What's different, our problem becomes how to maximize the capacity of secondary network when interference to primary network needs to be completely avoided. Similarly, beamforming, secondary user selection and power allocation are taken into account. In beamforming, based on subspace theory, "projection method" and "orthogonal basis method" have been proposed. For orthogonal basis method, two strategies for realization, which are respectively termed SVD-based scheme and QRD-based scheme, have been presented. In secondary user selection, "selecting one by one" and "removing one by one", two methods are proposed and their applications are pointed out. Simulation results reveal that even though simple but non-optimal equal power allocation scheme is adopted, only a little performance loss is introduced.Thirdly, in distributed cognitive radio networks, a pair of secondary users share a spectrum with a pair of primary users. By exploiting the correlation of radio channel in time domain, adaptive beamforming scheme has been proposed, which can adaptively adjust the direction of beam according to primary users'working condition. If perfect channel estimation is assumed, the interference to primary user can be completely avoided. Even though channel estimation error exists, the interference to primary user can still be effectively limited by power control at secondary user.