The Research Of Channel Estimation For UWB Based On Compressive Sensing

Compared with conventional communication technology,UWB communication has lots of advantages.So it is the trend of the future development of communication and becomes the hot topic of current research.It’s necessary to execute effective channel estimation to ensure that the overall performance of UWB communication system.However,since the bandwidth of UWB system is very large, it requires a very high sampling frequency according to conventional Nyquist sampling theorem,which results that it’s difficult to impletement based on existed hardware technology.Fortunately,the rise of compressive sensing(CS) brings further development to UWB channel estimation.So CS draws researchers’ great attention.The method of UWB channel estimation based on CS is discussed in this thesis,which has a significant influence on the development of UWB communication.The basic framework of CS and the model of UWB communication system are analysed and the property of UWB channel is explored in this thesis.According to its property, Bayesian CS is used in UWB channel estimation.In CS theory, sparse presentation dictionary should make a sparser presentation for the original signal, and the measurement matrix should achieve signal reconstruction based on a few sampling numbers as possible.Therefore,a further study about UWB channel estimation based on Bayesian CS is conducted in this thesis. UWB channel estimation based on Bayesian CS with the joint of eigen-based dictionary and adaptive measurement matrix is discussed, in which, eigen-based dictionary make a sparser signal and adaptive measurement matrix implement the required reconstruction accuracy with fewer measurements, thus, sampling with lower frequency is really achieved.In addition, in terms of the inherent statistical property of UWB channel, it is taken in consideration that each atom in eigen-based dictionary has different effect on signal reconstruction. The scheme weighted-eigen dictionary and segmented Bayesian CS based UWB channel estimation is analysed in this thesis. The statistical structure of UWB channel is fully exploited in this scheme, so the computing complexity is lower and the estimation performance is greatly improved.This thesis includes five parts:The research status of CS and UWB channelThe introduction to CS and UWB channel estimationUWB channel estimation based on Bayesian CS with the joint of eigen-based dictionary and adaptive measurement matrixEnhanced UWB channel estimation based on Bayesian CS Summary and outlook These content are of great reference value for the future of UWB channel estimation.