| In recent years,with the rapid increase of the demand for wireless communication by various intelligent terminals,the scarcity of spectrum resources caused by static spectrum allocation strategy is becoming more and more serious.Without affecting the normal communication of the Primary User(PU),Dynamic Spectrum Access(DSA)technology based on the concept of Cognitive Radio(CR)allows the Secondary User(SU)to dynamically search for free spectrum and use spectrum resources already allocated to PU for wireless communication.However,sensing errors often occur when the SU works under low signal-to-noise ratio conditions,which will result in Imperfect Spectrum Sensing(ISS).Therefore,it is of great significance for the study of improving DSA performance under ISS.The channel statistics such as channel mean period and distribution type observed by SU are usually unreliable due to possible errors in spectrum sensing.However,the existing estimation methods of channel statistics have some shortcomings,such as low accuracy and complex calculation.Therefore,aiming at the problem of spectrum waste caused by sensing errors in traditional spectrum sensing schemes with fixed sensing period and the important role of statistical information in improving the performance of DSA,this thesis respectively studies the adaptive spectrum sensing scheme,the channel mean period estimation method and the channel periods Probability Mass Function(PMF)estimation method are studied.The main research contents of this thesis are as follows:(1)A closed-form expression is derived to accurately estimate the channel mean period from the sensing results,the sensing error probability and the sensing period.A two-state Markov chain is used to model channel state and the probability of channel state change in different cases is solved.Based on the relationship between the state transition probability and the number of periods,and the influence of different types of sensing errors on the number of periods under ISS is analyzed,and the closed-form expression for accurately estimating the channel mean period is derived.The established closed form expression has the advantage of fast computation,without the prior knowledge of the channel and the distribution law of the channel period,and can be widely used in most scenarios.Simulation results show that compared with the closed-form expression in the existing literature,the derived closed-form expression for estimating channel mean period is more accurate,and is not affected by the sensing error probability and the sensing period.(2)A closed-form expression is derived to accurately estimate the PMF of channel periods based on the sensing results under ISS,the mean channel period,the sensing error probability and the sensing period.According to the different length of the original period,the influence of sensing errors within the period and adjacent periods on the sensing results under ISS is analyzed.Simulation results show that the derived closed form expression is better than the existing closed form expression in estimating the accuracy of PMF,and it is not affected by the sensing period and the mean channel period.(3)A spectrum sensing scheme based on the adaptive SU’s working state is designed.The SU adaptively senses the spectrum or dynamic spectrum access according to the sensing results.In this scheme,the SU can maximize throughput or energy efficiency by finding the right sensing time under constraints that do not interfere too much with PU.To solve the problem that the average false alarm probability is difficult to deduce due to the unfixed sensing time,the average false alarm probability based on PMF is derived,and the expressions of the SU’s DSA performance such as spectrum utilization rate,throughput and energy efficiency in this the spectrum sensing scheme designed in this thesis are further derived.Simulation results show that compared with the existing spectrum sensing schemes,the scheme designed in this thesis can effectively improve the DSA performance of the SU. |
PDF Full Text Request