Wideband Spectrum Sensing Algorithm Research In Cognitive Radio Networks

With the explosive growth in new wireless communication technologies, numbers of wireless applications begin to come out, therefore, wireless networks are facing great challenges. In order to satisfy the various requirements of wireless service and provide a better user perceived quality, more frequency resource should be allocated to existing wireless networks. However, the current technology can not provide an available solution to take advantage of UHF and frequency above. Thus there is an obvious contradictory between limited frequency resource and growing demand of frequency. Cognitive radio networks (CRN) can perceive the environment and adjust its parameters and then provide good performance of end-to-end transmission. Considering the low utilization of spectrum under current fixed spectrum allocation policy, CRN can opportunistically access the network, with the interference caused to the authorized network under control, and is becoming a promising method for easing the frequency embarrassment.This thesis focuses on the obtaining information of CRNs via spectrum sensing and cognitive database. Spectrum sensing is used for actively detecting the radio environment for vacant spectrum before accessing wireless network. However, considering the bottleneck of hardware technology for A/D conversion of tremendous data, compressed sensing for wideband which can greatly reduce the number of samples serves as a good solution. For cognitive terminals equipped with spectrum sensing module, channel fading and hidden terminal problems are obstacles for achieving good detection performance. With cognitive database incorporated, hidden terminal problem is solved. To sum up, the thesis is constituted of the following aspects.First of all, we make a brief description of cognitive radio and cognitive network which are effective technology for easing the frequency embarrassment. Next, give the definition of cognitive network, and mainly introduces the two methods for obtaining information, spectrum sensing and database accessing. Then detailed explanation of recent developments of spectrum sensing and database is given and concludes to the combination of wideband spectrum sensing and database for better wireless network access.For wideband detection algorithm part, firstly the relevant challenges are discussed. Considering the limitation of RF front-end, compressed sensing is introduced. By comparing with the existing compressed sensing methods, the compressed ration allocation based collaborative wideband spectrum sensing is proposed as the main thought of the thesis. By guaranteeing the system detection performance, the proposed method reduces the samplings tremendously for signal recovery, and decreases the overhead at low performance degradation.For collaborative detection of spectrum sensing and cognitive database accessing, the benefit of incorporating database is analyzed at the beginning. Then in the second chapter, elaborate illustration of multi-dimensional cognitive database is given. Based on the detailed introduction of information stored in database, an interactive mechanism between spectrum sensing and database is brought up in chapter four. And detailed description of each module and interface parameters are presented. The algorithm proposed can solve the hidden terminal problem for wideband spectrum sensing, and simplifies the access procedure. By base station based detection with strong data processing ability, users are offered better chances for network access.Finally, the conclusion is given in the end of this thesis, there are still some limits in the proposed schemes and the author also points out the research filed in the near future.