Research On Novel Efficient Cooperative Spectrum Sensing Algorithms

With the rapid development of Internet of Things technology,the number of mobile communication services and devices have increased dramatically.In order to meet the demands of massive users and services on speed,delay,and energy consumption,a large amount of spectrum resources and efficient allocation methods of resources are required.However,as a limited resource,spectrum resources are fixedly allocated to specific services by the government,which results that spectrum resources are scarce.Cognitive radio effectively improves spectrum utilization and alleviates the scarcity of spectrum resources through dynamic spectrum sensing technology.This paper conducts research in depth on efficient spectrum sensing algorithms in cognitive radio.The main purpose of the research is to make the proposed algorithm cost lower computational complexity,less bandwidth cost,or less sensing energy.The contribution and innovation are summarized as follows:1.Spectrum sensing algorithms based on eigenvalues are studied.Aimming at the deficiency of eigenvalue-based sensing algorithm with high computational complexity and inaccurate thresholds,two novel algorithms based on the LDLT decomposition are proposed.The largest eigenvalue and the smallest eigenvalue obtained from the LDLT decomposition are selected as the test statistics respectively.Therefore the computational complexity is reduced to one third of the traditional eigenvalue based sensing algorithms.In addition,the theoretical relationship between the threshold and false alarm probability is deduced via the hard decision criterion.Theoretical analyses and simulation experiments show that the two proposed algorithms have better detection performance than traditional eigenvalue decomposition algorithms when the number of collaborative users is small.2.The trade-off between soft-combining and hard-combining in cooperative spectrum sensing is studied.The soft decision based algorithms have high bandwidth consumption in the local data transmission process,however,the fusion center will obtain more accurate primary user signal power and better sensing performance.The local transmission bandwidth of hard decision is only 1 bit,which also degrades the sensing performance.It can be seen that there is a certain relationship between the local transmission bandwidth and the detection performance.A trade-off algorithm betweent bandwidth and performance is designed in this thesis.Based on the expectation of uniform distribution and Gaussian truncation distribution,the local decision results are restored to data.Then a final decision will be made in fusion center.In addition,the theoretical analysis of bandwidth consumption is given.The simulation results show that the proposed algorithm is close to the soft-decision sensing algorithm and the bandwidth cost is only less than 2 bits.3.The assignment and optimization of channels and sensing sensors in heterogeneous networks are studied.Energy harvesting technology is utilized to obtain energy for the device from the external environment,which can ensure the long-term work of the sensors in a heterogeneous network.The information of channel state is an important prior information in spectrum sensing.In this paper,the channel information is considered and energy harvesting is applied to study the energy efficiency of the sensor in heterogeneous networks.At the same time,the frame structure is designed,and the channel state of the next frame is predicted by using prior information.Finally,the mathematical model is established and the parameters under the optimal sensing energy efficiency are obtained via the greedy algorithm.To summarize,the proposed algorithms solve some problems that still exist in different sensing scenarios in order to improve efficency.Theoretical and simulation analyses demonstrate that these proposed algorithms have the value of theoretical significance and practical application.