Research On Synchronization And Channel Estimation For UWB Impulse Radio Communications

Ultra-Wideband Impulse Radio (UWB-IR) Communication is a communication technology for high data rate transmission by transmitting extremely narrow pulses on the order of sub-nanoseconds. Since the spectral mask was released by the Federal Communications Commission (FCC) in February 2002, UWB-IR has attracted growing interest all around the world. Experts and scholars attach importance to UWB-IR application to the short range wireless communications. Compared with other short range wireless communication technology (such as Bluetooth, IEEE 802.11a/b/g), UWB-IR technology has several advantages including:ample multipath diversity, low-cost baseband transceivers, high data rates along with low transmission power, potentially large user capacity, and the ability to coexist with narrowband radio systems operating in frequency overlay similar to that of CDMA network. However, deploying UWB-IR communication still faces many challenges, in which low-complexity and accurate synchronization and channel estimation in multipath channel environment are crucial.In UWB-IR communication systems, ultra-short pulses bear information bits in a low-duty-cycle pattern and propagate in a dense multipath environment. Therefore, low-complexity and accurate synchronization and channel estimation are difficult task to complete. This paper makes a comparatively intensive study on synchronization and channel estimation for UWB-IR communication systems, proposing two channel estimation algorithms and five synchronization algorithms, summarized as follows:1 A blind channel estimation algorithm based on discrete wavelet transform. The proposed algorithm exploits an unconventional denoising fast Fourier transform (FFT) to perform deconvolution. The denoising FFT employs the discrete wavelet transform (DWT) to compute discrete Fourier transform (DFT), meanwhile suppressing the noise. Theoretical analysis and simulation results both show that the proposed approach greatly improves the estimation accuracy of CIR, compared with the traditional FFT-based algorithm. However, their computational complexities are on the same order.2 A channel estimation algorithm based on zero-tap detecting Kalman Filter. An improved kalman filter algorithm is proposed to estimate the channel impulse response (CIR) of time-varying channel. By fitting zero-tap detection (ZTD) algorithm into kalman filtering process, the proposed algorithm exploits the system-specific UWB receiver structure and priori statistical characteristics of UWB channel. As a result, the computational complexity is reduced dramatically and the performance is improved. Simulations and comparisons are implemented to illustrate the improvement of the proposed algorithm.3 Two data-aided synchronization algorithms dispensing with searching procedures based on orthogonal pulses. A data-aided (DA) synchronization algorithm dispensing with searching procedures is firstly developed. By adopting a stream of alternating orthogonal pulses, frame-level synchronization can be achieved without channel information. Without utilizing any searching procedure, the timing offset estimator is obtained in a closed form. Because exhaustive search can be avoided, both computational complexity and synchronization time are reduced. Furthermore, the proposed algorithm remains operational in "cold start-up" scenarios. Simulations and comparisons are provided to illustrate the advantages of our design over the alternatives. Through analyzing the structure of the firstly proposed DA algorithm, the accuracy of the timing estimator depends on the choice of the pattern of alternating orthogonal pulses. Based on this point, we redesign a new pattern of alternating orthogonal pulses, resulting in a new DA synchronization algorithm dispensing with searching procedures. This newly proposed one not only has all the advantages of the firstly proposed algorithm, but also further improves the accuracy of the timing estimator compared with the firstly proposed algorithm.4 A non data-aided synchronization algorithm dispensing with searching procedures based on orthogonal pulses. Compared with data-aided (DA) synchronization algorithm, non data-aided (NDA) synchronization algorithm, also called blind algorithm, reaches a comparatively low accuracy of synchronization. However, DA algorithm makes use of training sequence, therefore sacrificing the efficiency of transmission rate and the efficiency of power. Therefore, it is necessary to study the non data-aided (NDA) synchronization algorithm dispensing with searching procedures. By adopting a stream of alternating orthogonal pulses, this chapter proposes a NDA synchronization algorithm dispensing with searching procedure. The proposed algorithm exploits square and overlap-add operations to eliminate the impact of the unknown information bits and obtains the closed form of the timing estimator dispensing with searching procedures. Computer simulations are provided to illustrate the improvement of the accuracy of estimation, compared with the alternative algorithm. Besides computer simulations, performance analysis and analytical performance comparison are also provided.5 A data-aided two-step synchronization through energy detection. The proposed DA algorithm consists of two steps. Since a UWB-IR symbol is comprised of multiple frames, the first step tries to find the start of individual frames via square and overlap-add operation. Then the first frame in each symbol is identified by exploiting the cyclic autocorrelation of DS codes at the second step. Simulations are performed to validate the promising performance of the proposed scheme.6 A blind two-step synchronization through energy detection. The first step tries to find the start of individual frames by energy detection, and then the first frame in each symbol is identified by exploiting the randomicity of DS codes and information symbols at the second step. Achieving synchronization with pulse-level resolution, the proposed algorithm considerably reduces synchronization time and complexity due to the two-step synchronization structure, compared with the single-step algorithm. Simulations and comparisons confirm that the proposed approach outperforms the existing alternative in terms of normalized mean square error and bit error rate.