Research On Communication Signal Spectrum Sensing Technology In Alpha-stable Distribution Noise

With the rapid trend of wireless communication technology,people's demands for various wireless services is rising,which makes relatively scarce spectrum resoures more tense.The emergence of cognitive radio technology has alleviated the scarcity of spectrum resources,while also improving the efficiency of spectrum utilization.This paper takes the spectrum sensing technology in cognitive radio as the main research content.Based on the commonly used digital modulation signal under white Gaussian noise and alpha-stable distribution noise,the applicable spectrum sensing algorithms are studied respectively.The specific problems for algorithms under white Gaussian noise are addressed,and the statistical order limiting problem of alpha-stable distribution noise has been improved.This paper first briefly describes the classification and theory of classical spectrum sensing algorithms.Three mathematical models of non-Gaussian noise are described.Selecting alpha-stable distribution noise to represent the impulse components of real noise properly,the statistical properties and fractional lower order moments(FLOM)of the noise are briefly introduced.Secondly,under the environment of white Gaussian noise,theoretical derivation and simulation compare the performance of the energy detector algorithm implemented in time domain and frequency domain.On the basis of cyclostationary signal theory,more reliable cyclostationary feature detection is achieved with the cycle frequency.In view of the problem caused by the noise uncertainty and the unsatisfactory perception of some existing eigenvalue-based detectors,the maximum eigenvalue to signal energy(MEE)detector is proposed to achieve the immunity of variety of noise power.Based on the nonlinear fractal dimension theory,the box dimensions of different communication signals are depicted,with which feature the effective detector is implemented.The advantages and disadvantages of the algorithms in this chapter and adaptability are summarized.Finally,under the alpha-stable distribution noise,theoretical analysis and simulation verify the applicability of the traditional energy detector with two nonlinear function processing algorithms.Based on the fractional low-order cyclic autocorrelation function of communication signals,simulation results of the relationship between the detection probability and the sensing environment under different parameters with ? and p are analyzed,besides using multiple cyclic frequencies.Based on the fractional low-order moment sampling covariance matrix,the difference between the maximum-minimum eigenvalue(DMM)algorithm under white Gaussian noise is improved.The statistical mean and variance value of the detector is re-derived using the theory of FLOM.In addition,theoretical derivation and analysis of simulation data suggest that using fractional low-order moment matrix determinants can also achieve effective spectrum sensing with lower computational complexity.The effectiveness of box-dimensional spectrum sensing algorithm under impulsive noise is analyzed.In addition,the sensing effect of the box dimension spectrum sensing algorithm under impulsive noise is analyzed,and the computational complexity and advantages and disadvantages of the above-mentioned algorithms are compared.