Research On Spectrum Sensing Algorithm Based On TD-LTE System

With the increasing demand for mobile Internet,spectrum resources are becoming more and more scarce.As one of the key technologies of cognitive radio,spectrum sensing can dynamically discover idle spectrum resources and improve spectrum utilization.The current spectrum sensing algorithms have a series of problems,such as limited sensing accuracy and insufficient flexibility in complex communication environment.In this paper,a spectrum sensing algorithm based on TD-LTE system is studied.By constructing the sensing regional model,the space-time-frequency multi-dimensional spectrum sensing is realized and the spectrum utilization is improved.A sensing regional model of high frequency spectrum utilization is constructed.Based on the base station distribution of TD-LTE system,this paper deeply analyzes the channel loss model and the interference source of the primary user,constructs the sensing regional model.The simulation results show that the proposed sensing regional model has higher spectrum efficiency.In the region with high channel occupation and primary user activity range greater than 2km,the spectrum efficiency can be improved by more than 11.12%.A space-time-frequency multi-dimensional spectrum sensing algorithm with adjustable performance is designed based on sensing regional model.Based on the time frame structure of TD-LTE system,the selection of spectrum sensing time length is completed.And the tunability of spectrum sensing performance is realized through the optimal sampling frequency selection algorithm and dynamic cooperative sensing algorithm.The results show that the average error between the classification model performance predicted by sampling frequency and the actual classification model performance is less than 0.77%,and the maximum error is less than 2.11%.The performance of the proposed spectrum sensing algorithm can dynamically adjust the spectrum sensing performance by changing the sampling frequency and the number of sensing nodes.A spectrum sensing algorithm based on spectrum prediction is studied.Based on the LSTM fitting model,the channel occupancy ratio is predicted,and the channel occupancy prediction results are used to adjust the parameters of the spectrum sensing model to improve the spectrum sensing performance.The simulation results show that the spectrum sensing algorithm based on spectrum prediction can effectively improve the spectrum sensing performance.After the optimization of occupancy ratio prediction results,both single node spectrum sensing and cooperative sensing can achieve higher sensing accuracy.43 Figures,13 Tables,72 References...