Wideband Spectrum Sensing For Hybrid Satellite-Terrestrial Communication Systems

With the development of information globalization,it is urgent to build a global information network with seamless coverage.As one of the important means to realize the wide-area coverage information network,hybrid satellite-terrestrial communication systems is the foundation to support the emerging applications,including the internet of things.Meanwhile,it will play an important role in dealing with territorial disputes and regional conflicts,natural disasters and man-made disasters.However,facing the design of hybrid satellite-terrestrial communication systems,the problem of system coexistence must be solved,in which the most important thing is to realize the spectrum sharing.Spectrum sensing is considered as the technical support for sharing spectrum resources between satellite and terrestrial networks.At present,wideband spectrum sensing has drawn more attentions.As the range of spectrum increases,more opportunities for spectrum access can be provided.It is an inevitable trend that both satellite communications and terrestrial communications will constantly develop higher spectrum resources,so hybrid satellite-terrestrial communication systems will face the problem of spectrum sharing.In order to achieve the mutual avoidance among multiple users under sudden interference and improve the utilization of spectrum,it is necessary to realize wideband spectrum sensing in a higher and wider frequency range.The wider the band,the higher the sampling costs.The existing ADC is difficult to achieve.Whether the sampling rate can be reduced becomes the bottleneck of restricting of development of wideband spectrum sensing technology.In recent years,the compressed sensing theory can support sub-Nyquist wideband spectrum sensing,which is able to recover the sparse signals from only a few measurements.In this thesis,we consider the wideband spectrum sensing scenario in hybrid satellite-terrestrial communication systems.Aiming at its characteristics of low spectrum utilization rate,signal position and quantity constantly changing with time,sub-Nyquist blind wideband spectrum sensing method is proposed.The proposed method can achieve sub-Nyquist sampling for wideband spectrum and handle multiple RF signals with unknown carrier frequency.It is suitable for application scenarios,in which the number of RF signals is dynamic change.It provides an effective way to improve spectrum utilization in hybrid satellite-terrestrial systems.The main research contents can be specified as follows.Firstly,multiband signal modeling for wideband spectrum sensing.In hybrid satellite-terrestrial communication systems,combining wideband spectrum sensing with compressed sensing theory,the multiband signal model for wideband spectrum sensing has been established.Considering the frequency sparsity of the multiband signal,signals can be sampled at sub-Nyquist rate by using modulated wideband converter.Then,the signal supports can be reconstructed to get the global usage status information of the cognitive spectrum.Secondly,pre-decision method based on modulated wideband converter.The operating frequency in hybrid satellite-terrestrial communication systems usually has low spectrum utilization rate and it is idle in most time and space.When there are no users in cognitive spectrum,false alarm occurs in the results of signal supports reconstruction because noise exists.The pre-detection process without additional sampling burdenis is added before the support reconstruction and the preliminary estimation of the spectrum utilization is given.If there are no other system users in the frequency band,the spectrum sensing process can be quickly ended and the results can be given.Then,the spectrum sensing can be quickly entered into the next round to effectively improve the speed of spectrum sensing.Thirdly,a blind signal support reconstruction method is proposed.In hybrid satellite-terrestrial communication systems,RF signals appear are random.The number of RF signals in the cognitive spectrum is dynamic change at a certain time and it is difficult to obtain the prior knowledge of the number of sub-bands.Since support reconstruction methods in modulated wideband converter rely on such prior information,modulated wideband converter cannot be applied directly to hybrid satellite-terrestrial communication systems.A blind support reconstruction method is proposed by exploiting the joint sparse characteristic of multiple measurement vectors.The proposed method can achieve the signal support reconstruction without the imformation about the location and number of sub-bands.Lastly,a blind wideband spectrum sensing method is proposed.Since RF signals in hybrid satellite-terrestrial communication systems is constantly changing,modulated wideband converter is difficult to be applied in the wideband spectrum sensing scenario.In this thesis,a sub-Nyquist blind wideband spectrum sensing method is proposed.In order to control the sampling process flexibly,single-channel structure is used to improve the analog front-end in traditional modulated wideband converter.Then,a pre-detection method is used to realize the rough detection for the cognitive spectrum.The position of sub-bands can be located by adopting a blind support reconstruction method to complete the fine detection.It provides technical support for spectrum sharing in hybrid satellite-terrestrial communication systems.