Research On Efficiency And Robustness Of Cooperative Spectrum Sensing In Cognitive Radio Networks

Recently, Dynamic Spectrum Access (DSA) has attracted a lot of attention. It allows secondary user to operate in licensed spectrum bands without harmful inter-ference to primary users so that the utilization of radio spectrum resource can be greatly improved. The Cognitive Radio (CR) is seen as the enabling technology for DSA, since it can sense the spectrum environment and make use of idle spectrum bands.To make use of idle spectrum efficiently and avoid interfering with primary users, a CR needs to carry out spectrum sensing accurately and quickly. However, due to the limitation of hardware, fading/shadowing, etc., it is not easy for a single CR to achieve the above goals. Recent research results indicate that cooperation can improve the performance of spectrum sensing significantly, which makes a more reliable final spectrum sensing decision by taking advantage of diversity gains.However, due to its technical features, there are still some issues to be resolved with the application of cooperative spectrum sensing in the cognitive radio networks, which are shown as follows:Cooperation also introduces additional overhead, because the local sensing re-sults need to be reported to the fusion center. This is required cooperative sensing to be conducted in a reliable and efficient way.Spectrum sensing in cognitive radio typically assumes that the variation of spec-trum status happens only before the beginning of the sensing block. However, actually, the spectrum status may vary at any sample of a CR user frame, which bring challenges to cooperative spectrum sensing.In the process of cooperative spectrum sensing, malicious users could send false local sensing results to the fusion center, which can cause severe distortion to the final sensing decision. To deal with this problem, the robustness of cooperative sensing must be improved.To answer the above problems, this thesis addresses on the cooperative spectrum sensing technologies in the cognitive radio network, and the contents of this thesis include the following issues:(1) Research on pipelined cooperative spectrum sensing in cognitive radio networks. (2) Research on multi-process cooperative spectrum sensing in cognitive radio networks. (3) Research on robust and efficient cooperative spectrum sensing scheme in cognitive radio networks The work in this thesis has been supported by the National Science Foundation of China "Research on Key Issues on Cognitive Radio Networks" (No.60602029), "Research on Wireless Sensor Networks based on Network Utility Maximization" (No.60772088), and the Chenguang Project for Youth of Wuhan City under Grant "Research on Cognitive Radio Networks" (No.200750731261) respectively.The contributions of this thesis include:1. Research on pipelined cooperative spectrum sensing in cognitive radio networks:Since most existing works assume a typical CR user frame structure in which spectrum observing and sensing results reporting are conducted sepa-rately in different time duration, we propose a novel pipelined cooperative sensing framework. In this framework, secondary users conduct spectrum sensing and results reporting in a pipelined way, and the time consumed by reporting is also utilized for sensing. Through pipelining, the new framework provides a much wider observing window than the general frame structure, and will react to any spectrum status change within the observing window, which results in a better performance for cooperative sensing.2. Research on multi-process cooperative spectrum sensing in cognitive radio networks:In order to improve the response speed to the variation of spectrum, we propose a multi-threaded SPRT method. In the MTSPRT method, there are multi SPRT processes working concurrently. Therefore, whenever a spectrum status variation happens in the observing window, there is always a SPRT process corresponding to it, which can arrive at a decision without any delay.3. Research on robust and efficient cooperative spectrum sensing scheme in cognitive radio networks:To improve the immunity of cooperative spec-trum sensing against false sensing information, we propose a robust cooperative spectrum sensing scheme which is based on weighted sequential probability ratio test (WSPRT) technology. Through reputation maintenance, weight allocation and multi-threaded detection, the proposed scheme can effectively deal with dif-ferent kinds of attack patterns. The proposed robust scheme can identify and weed out malicious users accurately and efficiently.