Reaserch Of Bayesian Compressive Sensing In IR-UWB Communication System

Impulse Radio Ultra Wideband communication technology carries information by sending very short pulse waveform, without carrier modulation, which can achieve high data transmission requirements from short range. However, the period of the IR-UWB signal is very short in time domain, which makes a high bandwidth in the frequency domain, thus has brought great challenge to the sampling unit of the receiver, so that the difficult implement with a low cost of high-speed analog-digital converter becomes the main factor that hinder its development.Compressive Sensing(CS) provides a new solution for signal sampling. The theory says that if a signal is compressibility or has a sparse form, we can reconstruct the original signal with a small amount of random sampling values. Due to the natural sparse characteristics of the IR-UWB channel and signal in time domain, they can be processed in the framework of compressed sensing, in order to reduce the sampling rate of the system, and CS-UWB system is constructed to solve the problem of high-speed ADC can hardly be achieved with low-cost, is a research hotspot in recently years.Considering the IR-UWB multipath channel and modulation signal characteristics have sparse, statistics and certainty prior, this article carries out in-depth research about IR-UWB system under the framework of compressive sensing from the perspective of the Bayesian, specific research contents include: We frist describe compressive sensing reconstruction problem in the CS-UWB with sparse Bayesian model, and BCS-UWB system architecture is proposed.Within this framework, a two-stage channel estimation method is proposed based on the statistical prior analysis of channel characteristics, cluster structure is used to estimate the channel information, obtain channel set of cluster location in the first stage. combining with the cluster location set, we use bayesian compressive sensing reconstruction algorithm to obtain the estimation of channel impulse response in the second stage, we get the better reconstruction signal to noise rate(RSNR) of channel impulse response, and the channel estimation result is used in signal demodulation, reduce the bit error rate. Further, we have designed the high rate of BPM-4PPM modulation scheme, and increase the data rate to150 Mbps. Considering the location and amplitude certainty priori information of modulation signal, we put forward the grouping location and approaching amplitude of the modified subspace pursuit algorithm to complete the work of signal demodulation, and further reduces the bit error rate of signal demodulation, improve the performance of the whole system. This paper expounds the IR-UWB systems with more excellent performance can be obtained under the framework of BCS than under the traditional framework of CS, extends the research category of IR-UWB system based on compressive sensing, and proposed a new thought of optional for the design of the low complexity and high performance IR-UWB systems in the future.