| Due to the high data rate, robustness to multipath fading, low power consumption, high-precision localization, low complexity and cost, impulse radio-ultra wideband (IR-UWB) is regarded as the most appropriate physical layer technology for short range wireless personal area network (WPAN).For wireless communication systems, in order to meet certain performance requirements, the receiver needs to collect multipath signal energy for improving signal-to-noise ratio (SNR). When inter symbol interference (ISI) causes serious deterioration of system performance, it is also needed to consider the suppression of ISI. Unlike the traditional wireless communication system with continuous carrier modulation, IR-UWB system uses extremely narrow pulses to carry information. As the radio wave propagation characteristics of non-continuous short impulse differ with that of narrowband signal, the ultra wideband (UWB) channel is different from the narrowband channel, which is embodied on the characteristics of dense multipath, frequency dependence and environment dependence. These factors bring more difficulties for IR-UWB receiver to realize multipath signal energy collection and suppression of ISI. We research on multipath signal energy collection and suppression of ISI combined with the characteristic of indoor UWB channel. It tries best to take advantages of the received impulse signal features to resolve effectively the two problems. The contents of this thesis are listed as following:I. The effect of UWB indoor channel characteristics on receiver performance. It is two important problems for system performance to collect multipath signal energy and suppress ISI, to obtain a certain performance, the two problems should be considered, both of them is related with channel. Different from the traditional narrowband channel, UWB channel possesses the characteristics of dense multipath, frequency dependence and environment dependence. At first, we analyze the effects of dense multipath, frequency dependence and environment dependence on multipath signal energy and suppression of ISI. Combined with current methods for the two problems, the drawbacks of current methods are pointed out. Then, the UWB indoor channel model is introduced, which provides the foundation for following investigations.II. The investigation on collection of multipath signal energy. For IR-UWB system, to achieve appropriate performance, it is needed to collect multipath signal energy, and thus improve the SNR. Because of the effects brought by the characteristics of dense multipath and frequency dependence, the energy of pulse signal is scattered in a large number of multipath signals, several dozens or even hundreds, and pulse waveform distortion for each path is different, these bring great difficulties to collect effectively multipath signal energy. Considering the impact brought by dense multipath and frequency dependence, in the existing investigations of collecting multipath signal energy, it is a feasible solution to estimate firstly the composite pulse multipath channel and then use it as the template to match the pulse signal to realize the collection of multipath signal energy. The effect of collecting multipath signal energy is influenced by the accuracy of estimation. The more accurate the composite pulse multipath channel is estimated, the better the effect is when the composite pulse multipath channel estimated is used as the template to match the pulse signal waveform, and thus the multipath signal energy can be collected more effectively. The average method is an effective method of estimating composite pulse multipath channel, which estimates the composite pulse multipath channel by averaging a certain number pilot pulse waveforms and can achieve certain accuracy of estimation. However, for average method, the interval between transmitted pilot pulses is limited by channel delay spread time, and hence it would take long time to estimate the composite pulse multipath channel. Aiming at the problem, a composite pulse multipath channel estimation method based on nearest neighbor (NN) rule is proposed in this thesis. The NN estimation method based on the premise that the received pulse waveform is still antipodal polarity when antipodal polarity pulse is transmitted, according to NN rule, the composite pulse multipath channel is estimated by the difference of pulse waveform polarity, thus the limit on interval between transmitted training impulses is not rigorous. Research shows that when the estimation time is reduced relative to the average method (reducing the interval of training pulse and adding the numbers of training pulse), the NN estimation method can still improve the accuracy of estimation. Hence, The NN estimation method can collect multipath signal energy more effectively, and improve the SNR, achieve better bit error rate (BER) performance.III. The investigation on suppression of ISI. For IR-UWB system, with the degree of ISI increasing, the impact of ISI on system performance is more and more serious, the suppression of ISI should also be considered. UWB channel is environment dependent, the channel is different in different indoor environment, then the degree of interference for the current symbol brought by adjacent symbols is different, and as a result, the degree of ISI is also different, this would influence the suppression of ISI. In the existing methods of suppressing ISI, equalization receiver is an optimal linear receiver, the equalization receiver adopts fractionally spaced linear filter structure, which is capable of performing the functions of matched filter and symbol spaced equalization, thus it can suppress effectively the ISI. For equalization receiver, the observation window length is an important parameter for suppressing ISI. In different channel, to suppress the ISI more effectively, the required length of optimal observation window is different. Especially for the actual system implementation, the channel information can't be obtained in advance. In order to suppress the ISI more effectively, this requires the equalization receiver to adjust adaptively the observation window length according the difference of channel. Aiming at the problem, an adjustable observation window length-fractionally spaced linear equalization (AOWL-FSLE) equalization receiver based on minimum mean squared error (MMSE) criterion is proposed. The receiver can adaptively adjust the observation window length according to difference of channel, achieve better steady mean squared error (MSE) performance, thus suppress ISI more effectively and improve BER performance. In addition, for IR-UWB system, considering in the indoor environment, some of the electronic devices generate the large value of the interference, the noise embodies obviously the character of non-Gaussian. The equalization receiver based on conventional MMSE criterion exhibits performance deterioration in presence of non-Gaussian noise. Therefore, an adjustable observation window length (AOWL) equalization receiver based on H∞criterion is proposed. The proposed receiver can adjust adaptively the observation window length according to difference of channel, thus ISI can be suppressed more effectively. Moreover, the proposed receiver is based on H∞criterion, which makes it robust to the effect brought by non-Gaussian noise, thus it is more appropriate to be realized for actual system. |
PDF Full Text Request