Research On Algorithms Of High Speed Wideband Signal Processing

Ultra-wideband(UWB) wireless communication technology, as an emerging technology, has many advantages compared with narrowband wireless communication technology. Therefor UWB technology has received the widespread attention in academia and industry in recent years. UWB receiver is the research emphasis of UWB communication, but the hardware implementation of UWB receiver still faces many problems.One of the main bottleneck is the UWB signal sampling. Because of the huge signal bandwidth, receiver needs to use high-speed analog-to-digital converter(ADC) to sample the signal. And because of the high complexity, high price and huge power consumption of high speed and high precision ADC, it is difficult to meet the needs of practical system.In order to solve this problem, we study it from two aspects, reducing sampling precision and reducing the sampling rate.Reducing the sampling precision can reduce the complexity of ADC, thereby to increase the sampling rate to Nyquist sampling rate. Monobit receiver has received a widespread attention, because of its simple structure. But high performance monobit receiver needs to use monobit ADC sample results to estimate its parameters such as amplitude of the signal, the threshold of ADC has great impact on parameter estimation.So we focus on the effect of a monobit receiver quantization threshold on the performance of the receiver. By analyzing the relation between Cramer-Rao lower bound and the quantitative threshold, we find that the optimum quantization threshold is not 0, that is commonly used as quantitative threshold, but the signal amplitude. Because the signal amplitude of unpredictability, we put forward three kinds of adaptive quantization threshold adjustment scheme, these schemes can achieve optimum quantitative threshold adaptively. Experiments show that compared with the fixed threshold, these three schemes of parameter estimation performance is better.The performance of the receiver show that compared with the fixed threshold estimate the parameters, parameters estimate by using adaptive threshold adjustment can promote receiver performance with 1-2dB, in high signal-to-noise ratio.Compressed sensing is a new way to reduce sampling rate, but due to the sparse reconstruction algorithm cannot meet both the requirements of reconstruction precision and speed, and large amounts of reconstructed Nyquist sampling data, compressed sensing methods do not be applied in the practical UWB systems. Referring to the compressive sensing sampling model, we proposed a low sampling rate UWB receiver. Using the finite character set signal characteristics, we recovery directly the signal information bits instead of Nyquist sampling sequence, thus can reduce the sampling rate of the signal and reduce the complexity of the calculation. The original mathematical problem of proposed receiver is a combinatorial optimization problem, we transform it into a concave minimization problem, and a surrogate function iteration descend algorithm is proposed to solve the problem. The analysis of the performance of proposed algorithm is presented.Simulation results show that the proposed receiver can reduce the sampling rate of the signal, and the proposed algorithm has a higher recovery precision and lower calculation complexity than infinite norm minimization algorithm and basis pursuit algorithm.