Research On Spectrum Sensing Algorithm In Cognitive Radio

The static spectrum allocation strategy leads to severe scarcity and tension of spectrum resources.Cognitive radio technology can improve spectrum utilization efficiency and alleviate spectrum resource tension.Spectrum sensing is one of the key technologies of cognitive radio.The energy detection method and spectrum sensing algorithm based on power spectrum are simple and easy to implement,and have been applied in practical applications.Based on these two types of algorithms,spectrum sensing algorithms are deeply studies.Firstly,two single user spectrum sensing algorithms are studied.Aiming at the problem that traditional dual threshold energy detection algorithms need to determine thresholds based on known noise variances and have poor detection performance,a spectrum sensing algorithm based on power spectrum and dual thresholds(PSADT)was studied.The algorithm uses the power spectrum to calculate signal energy and perform detection outside the dual threshold.The ratio of the average difference between the maximum and minimum values of the power spectrum and the minimum value of the power spectrum is used as a detection statistic within the two thresholds.The power spectrum near the fundamental frequency is used to estimate the maximum value,and the previous part of the power spectrum after ascending order is used to estimate the minimum value to reduce noise impact.The false alarm probability of the algorithm is derived and the theoretical threshold is obtained.The simulation results show that the PSADT algorithm has a strong ability to resist noise power uncertainty in additive Gaussian white noise channels and Rayleigh fading channels.When the frequency offset coefficient increases from 0.02 to 0.04,the detection performance of the PSADT algorithm deteriorates less than that of the contrast algorithm,and it has a stronger ability to resist frequency offset.Aiming at the problem of the large error of the estimate of the noise variance in the detection statistics of the contrast algorithm(PSEGAR),which resulted in poor detection performance,a spectrum sensing fusion algorithm based on power spectrum(SSFOPS)was studied.The noise variance is estimated by using the geometric mean of the power spectrum and the arithmetic mean of the partial minimum values after the power spectrum is sorted,respectively,and the detection statistics is constructed by using these two estimates and the maximum and minimum values of the power spectrum.The false alarm probability of the algorithm is derived and the theoretical threshold is obtained.Moreover,the selection of parameters is analyzed by simulation.The simulation results under additive Gaussian white noise channels and Rayleigh fading channels show that the detection performance of the SSFOPS algorithm is superior to PSEGAR,maintaining a strong ability to resist noise power uncertainty while improving the ability to resist frequency offset.Then,a centralized cooperative spectrum sensing algorithm is studied.In order to fully utilize the feature of the energy vector characteristics of each user,a cooperative spectrum sensing algorithm based on multi head self attention mechanism was proposed(MACSS).Each user independently samples data,and the fusion center collects energy data from each user,and constitutes an energy vector as input to the self attention network.A network model based on multi head self attention mechanism is designed.It effectively extracts to local features of the energy vectors of signals and noise,effectively extract local features,and achieves intelligent collaborative spectrum sensing.Simulation results show that the performance of the proposed MACSS algorithm is far superior to the K rank criterion algorithm(KRCA)and the deep cooperative spectrum sensing algorithm based on dimension reduction and clustering(DRAC).The false alarm probability is 0.01.When the signal-to-noise ratio is-16 d B,the detection probability of the MACSS algorithm is 22%and 30% higher than that of the DRAC method and the KRCA method,respectively.When the signal-to-noise ratio is-13 d B,the detection probability of the MACSS algorithm is 91%,while the DRAC method is 78%,and the KRCA method is 73%.Finally,the performance of the PSADT and SSFOPS algorithms in this paper is tested.In order to verify the practical feasibility of the algorithm,a spectrum sensing algorithm performance testing system has been built by using a PC and HDU-SDR software radio development platform which consists of an AD9361 RF chip and a ZYNQ series XC7Z020 processor.The PSADT and SSFOPS algorithms are programmed and implemented.The communication signals are transmitted and received by using the software radio platform,and the detection performance of PSADT and SSFOPS on actual signals is tested.The performance test results show that the PSADT algorithm and the SSFOPS algorithm are superior to the comparison algorithm,respectively,and have good frequency offset resistance.The performance curve rules are consistent with the computer simulation results,verifying the correctness and practicality of the PSADT algorithm and the SSFOPS algorithm.