The Research On Wideband Spectrum Sensing Technology

With the popularity of radio,spectrum resources have become increasingly scarce,which has resulted in the very limited frequency bands that unauthorized users can legally use,and the frequency bands being used are not high in the total frequency band.The emergence of cognitive radio technology is to solve the defects of frequency band resource congestion and uneven distribution without affecting the main user band resources.In cognitive radio technology,broadband spectrum sensing is a very important component,and collecting broadband signals and detecting broadband signals are the focus and difficulty of broadband spectrum sensing.Especially when the collected signal is up to GHz,when using the traditional Nyquist sampling theorem for wideband signal sampling,there are few AD devices that can support the Nyquist sampling theorem,and the cost is very high.When detecting broadband signals,most of them belong to blind detection,and have little prior knowledge of the signal.Existing detection methods may not be applicable to low noise ratio situations(energy detection,etc.),or computational complexity High(cyclic stationary detection,etc.).On this basis,this paper has conducted a detailed study on the undersampling-based broadband spectrum sensing technology.The specific research content and results are as follows:1.Under the basic model of spectrum sensing,the most basic three kinds of classic spectrum sensing algorithms are analyzed: an energy algorithm that uses signal energy for detection,and a matched filter that maximizes the signal-to-noise ratio of the signal output under specific conditions.The detection algorithm and the detection algorithm of the cyclic stationary feature using the inherent periodicity of the signal.In order to make these three algorithms more systematic,in Chapter 3,not only the characteristics of the three algorithms are introduced,but also the test process of the three algorithms is re-derived,and the advantages and disadvantages of the three detection algorithms and the scope of application are explained.It has laid a theoretical foundation for the full text research work.2.A spectrum sensing method based on nested sampling is proposed,which is a method to break the Nyquist theorem to analyze modern signals.Using this method,you can use a sampling rate much lower than the Nyquist rate while undersampling thesignal and recovering its spectrum.When performing spectrum analysis on the broadband signals involved in the process of broadband spectrum sensing,the results of nested sampling can be directly spectrum analyzed.In this paper,in nested sampling and spectrum sensing,a spectrum sensing algorithm based on nested sampling is proposed,compared with As far as other algorithms are concerned,the algorithm has a lower sampling rate,and has higher detection performance under the same sampling rate.3.A spectrum sensing algorithm for non-circular signals based on nested sampling is proposed.In the actual communication system,the non-circular signal is a very common signal.For acyclic circular signals,the second-order statistical characteristics include conjugate covariance in addition to covariance,but existing perceptual algorithms ignore their conjugate covariance.Based on this,for non-circular signals,a new non-circular signal detection algorithm is designed,and the test statistics are constructed using the covariance and conjugate covariance through nested sampling,and the theoretical threshold is derived.In order to verify the detection performance of the proposed method,the performance of the proposed method was studied from the perspective of the characteristics of the test statistics,sampling sequence,sampling length and false alarm probability.The results show that this method has better perceptual performance than other perceptual algorithms.In addition,the proposed method does not require any known channel,noise and prior information of the main user,and can achieve better performance even in scenarios with a small number of sampling points and low SNR,and can be widely used in spectrum detection In practical applications.