Research On Spectrum Sensing Algorithm Of Cognitive Internet Of Vehicles Based On Collaboration

With the rapid development of Internet of Vehicles,we have been inseparable from the vehicle.The spectrum sensing technology of the cognitive vehicular ad-hoc network has attracted the attention of the majority of researchers to reduce traffic congestion,energy conservation,green commuting,and improve band utilization.In order to solve the above problems,this paper studies the performance of spectrum sensing under the framework of the cognitive vehicular ad-hoc network system,and establishes the spectrum sensing model of vehicular communication network.A Doppler frequency shift occurred in the vehicle moves at a high speed,which causes the signal dispersion and signal frequency shift,and dramatically reduce the detection accuracy and performance of the system.This paper proposes a Doppler segmented signal model with using the Doppler frequency shift estimation algorithm based on the radar-based Keystone transform to compensate the Doppler frequency offset of high-speed moving vehicles.The model takes into account the following situations that the signal-to-noise ratio is low,the channel is complex and the signal is weak.Therefore,the double-threshold energy detection algorithm and cyclic feature detection algorithm are combined to detect twice the signal,and the signal that cannot be determined between the thresholds enters the second Sub-cycle feature detection.In the spectrum sensing algorithm,the determination of threshold will affect the detection performance.In this paper,the sequential unconstrained minimization technique is used to determine the threshold,which improves the detection performance.The above-mentioned spectrum sensing detection results for a single cognitive vehicle need to be collected in the fusion center for judgment,which also has certain limitations.Centralized cooperative spectrum sensing is adopted in this paper and upload the judgment result of a single cognitive vehicle to the fusion center and make a judgment.K-rank fusion is adopted in this paper.Authorized users are determined to exist when more than half of the results determine the existence of authorized users;otherwise,primary users do not exist.Finally,the improved algorithm is simulated and verified on the Matlab platform by comparing energy detection algorithm and cyclic feature detection algorithm.The improved spectrum sensing detection algorithm of cognitive vehicular ad-hoc network is proposed in this paper to improve the detection performance of the system and verify the correctness of the theory.