Ultra-wideband Channel Estimation Based On Compressive Sensing

Ultra-wideband(UWB) technology is a new type of wireless communication technology, with many advantages, such as anti-jamming capability, small energy consumption, and good secrecy. However,because of its high bandwidth,it requires extremely high sampling rate to reconstruct signal. This is a major challenge for hardware implementation, which is an important factor restricts its development. As an emerging theory, compressed sensing provides a new solution to this challenge. This theory suggests that it is possible to well recover a signal without any errors via the sampling rate that is far less than Nyquist rate, which can greatly reduce the sampling rate and the complexity of the calculations. The basic premise of using the compressed sensing is the signal is sparse or it can be sparse represented. UWB signal has a natural sparsity. Therefore, the compressed sensing theory and Ultra-Wideband Channel Estimation can be combined effectively.This thesis focuses on the application of the compressed sensing theory in the UWB systems channel estimation. It describes the development of UWB and compressed sensing theory. Then, we study the application feasibility of the compressed sensing in UWB systems, introduce the common UWB channel model and the traditional channel estimation algorithm. Moreover, it takes further discussion about the four kinds of the common channel models in IEEE802.15.4a and researches the sparsity of the four channel models, providing the theoretical basis for the application of the compressed sensing theory in the channel estimation.Finally, we analyze a compressed sensing based UWB communication system which has an analog filter in front of channel and analyze the design idea of this system. On this basis, this thesis provides two improved scheme. One is an improved UWB communication system model based on compressed sensing, which applying a digital random pre-coding block at transmitter instead of analog schemes. Another is an improved algorithm named extended OMP(eOMP) algorithm. This algorithm improves the Element selection of OMP algorithm. Compared with the OMP algorithm. eOMP algorithm has higher speed and a higher reconstruction accuracy.