| Impulse radio ultra-wideband(IR-UWB) is an emerging short-range wireless communications technology. Depending on its advantage of low-power, high penetration, high security, high transmission rate, and strong anti-jamming ability, UWB has caused quite widespread concern and in-depth study. However, there are also series problems in ultra-wideband technology application areas, the time domain duration in nanosecond of UWB pulse decides a high bandwidth and this feature will generating a significant burden at ADC’s ability in digital receiver design, designing the sampling part of receiver in strict accordance with the Nyquist sampling theorem will bring high costs.In recent years, Compressed Sensing(CS) is a hot topic in application field of mathematics. The theory predicts that, if the signal satisfies the characteristics of compressibility and sparse,then it can be reconstructed in a high probability only by the number of observations which is far less than the Nyquist sampling theorem. Meanwhile the natural sparse of UWB signal can meet the requirements of compressed sensing theory, then the compressed sensing theory seems to be the best candidate to break the bottleneck of traditional ideas in the design of IR-UWB digital receiver.This thesis describes the basic theory of CS and the basic architecture of the CS-UWB receiver, then carry out further research on CS-UWB channel estimation direction based on the existing research results. The general form of received signal by ultra-wideband receiver is: g = Yq +n, where Y is the local template matrix, n is the noise vector and q is the channel impulse response to be estimated. The accuracy of IR-UWB channel estimation is depend on both Y and n, this paper discussed the impact of these two aspects and put forward a corresponding method on the basis of existing studies. The traditional OMP(Orthogonal Matching Pursuit) algorithm get channel estimation using iterative between the received signal and the local template with poor noise immunity. In this paper, the effection of noise vector is added to OMP algorithm in the traditional iterative process, proposing the Anti-Noise OMP reconstruction algorithm based on the derivation and proof of the formula, and demonstrates the rationality of derived by simulation and comparison with the original OMP algorithm. For the template improving, this paper first analyzed the data measured by ultra-wideband transceiver, describes the environment and the parameters of the experiment, determine if they are subjected to the impact of the distortion by analysis of the different points of the received signal, creating multi-template MTY which contains the information of distortion according to the analytic result, and proved the performance advantages of multi-template through simulation by comparing with single template. Based on the analysis of characteristics of ultra-wideband channel, we use partial channel priori information(CPI) to get a weighting function and use it to further improve the templates and get the improvement in performance. Finally, this paper combines the proposed method for the improvement of noise and local templates together, forming a systemic CS-UWB channel estimation scheme, enhancing the recovery signal-to-noise ratio(RSNR) in about 1d B under different number of iterations. |
PDF Full Text Request