Study Of Effective Sparse Representation Method Of Ultra-wide Band Signal

As a kind of novel short range communication techniques, Ultra-Wide Band(UWB) has been well known for these years. Comparing with the traditional wireless communication skills, UWB communication system has many unique advantages such as high data transmission speed, low consumption, low power and good anti-interference. However, the conflict between high bandwidth of UWB signal and Nyquist sampling theorem causes that fast sampling frequency should be offered to receiver of UWB communication system for subsequent signal processing. It is cannot be realized and the reason is that the speed and cost of current high speed AD/DA converter cannot meet requirements, which leads to the slow development of UWB.Sparse representation is a crucial part of Compressed Sensing(CS) theoretical framework and its main purpose is to decompose signal with over-complete dictionary and obtain the representation coefficient which contains the information of signal. The innovation of CS is that it combines sampling with compressing together, which can extract out information of signal directly and reduce the sampling frequency. CS breaks through the limit forced on signal processing by traditional sampling theorem and can be a new method to sample and process UWB signal.This paper studies the application of sparse representation onto UWB communication system. First of all, some basic knowledge and theories are introduced; therefore, it is much easier for readers to understand the following parts. Then, an algorithm for building clustering dictionary based on K-SVD training algorithm is proposed, which can efficiently reduce the large redundancy of dictionaries built by traditional algorithms. This method is helpful to cut down the calculation steps and solves the problem of low computing speed. On top of that, an algorithm of sparse representation based on partition is also proposed in this paper. This algorithm raises the speed of decomposition algorithm with high quality of approximation. Finally, the feasibility and the robustness are tested and the results of this paper are gained by theoretical analysis and experimental simulation. The results of this paper illustrates that CS theoretical framework can be used to improve the UWB wireless communication system and all the work lays a good foundation for the further research.