Research On Spectrum Sensing Technology In Cognitive Vehicular Ad Hoc Networks

With the rapid development of Intelligent Transportation System(ITS),people's requirements for road safety services and in-vehicle entertainment services are getting higher and higher.The explosive growth of various in-vehicle communication services has resulted in the lack of radio spectrum resources for vehicular ad hoc network(VANET)communications.In order to solve this problem,cognitive radio technology is introduced into the vehicular ad hoc networks,which can opportunistically access the underutilized parts of the licensed frequency band to provide additional spectrum resources for vehicular communication.Spectrum sensing technology is the first step to realize the reuse of licensed spectrum resources in cognitive radio system.Firstly,this thesis introduces the related knowledge of spectrum sensing technology in Cognitive Radio enabled Vehicular Ad-hoc Networks(CR-VANETs),and puts forward the limitations and challenges of existing spectrum sensing technology.In view of the shortcomings of traditional spectrum sensing technology in the environment of VANET,a spectrum sensing method suitable for vehicular communication environment is proposed.The details are as follows:1.A dynamic double threshold energy detection algorithm based on FCM is proposed.The mobility of cognitive vehicles leads to dynamic changes in the vehicle communication environment,and the noise power received by different road segments also changes.Existing energy detection algorithms require some priori information such as signal-to-noise ratio to obtain threshold values,which can not adapt to the rapidly changing vehicle communication environment.In order to solve this problem,a dynamic double-threshold energy detection algorithm based on FCM is proposed by using the local history sensing data of cognitive vehicles.The FCM algorithm adaptively acquires the detection threshold of energy detection in a dynamically changing network environment.Aiming at the adverse effect of noise uncertainty on single threshold detection,a double threshold detection scheme is introduced.The detection statistics of uncertain regions are also judged to retain more sensing information.Simulation results show that the proposed algorithm has good adaptability and better detection performance in vehicle communication scenarios.2.A cooperative sensing algorithm based on vehicle location and correlation is proposed.Due to the limitations of single-node spectrum sensing technology,cooperative sensing is introduced to improve the detection performance of the system.Cooperative sensing needs to consider both sensing performance and system overhead to achieve a good compromise between them.In this thesis,cooperative nodes are selected based on vehicle location and correlation.Considering the influence of path loss and shadow effect,the algorithm guarantees the sensing performance while selecting fewer sensing nodes to participate in the cooperative sensing;Then,a data fusion method between hard decision fusion and soft decision fusion is proposed.Cognitive vehicles participating in the cooperative sensing upload 2bit local sensing information,and RSU make a linear weighted fusion decision.In this thesis,a simulation experiment is carried out on the proposed algorithm in the scene of vehicle movement.Compared with the existing spectrum sensing algorithm,the proposed algorithm is improved and the balance between the sensing performance and overhead is achieved.