The Study On The Fast Multi-dimensional Acquisition Schemes For Direct-Sequence Spread-Spectrum Systems With Antenna Arrays

Recently, the incorporation of antenna arrays to direct-sequence spread-spectrum (DS/SS) systems has been widely considered and becomes more and more popular in modern communication systems, due to the significant performance improvements in spatial filtering and bandwidth efficiency. But the researching advance on acquisition, a bottleneck for DS/SS systems, is not satisfied and most schemes do not exploit the potential of antenna arrays well. In this dissertation, we study multi-dimensional acquisition of DS/SS systems based on antenna arrays and achieve some satisfied results through exploiting the potential of antenna arrays to increase the detection probability of code acquisition process.First, a three-dimensional acquisition scheme of DS/SS systems based on antenna arrays is proposed in this dissertation, which can work well even if the Doppler offset is up to 40kHz. And those conventional schemes can not work in this larger Doppler offset. For accelerating the searching process in frequency domain and estimating the Doppler offset accurately, FFT transform is used in this scheme. It is worth noticed that FFT can not work well in those systems with single antenna because the pre-intregation period must be shortly due to existing larger Doppler offset so that the appropriate signal-noise ratio(SNR) can not be obtained after despreading to guarantee the correctness of the Doppler estimation. But it is different for the systems with multi-antenna arrays because the gain obtained from antenna arrays can improve the SNR. Analysis in theory and numerical results show that the proposed scheme outperforms the conventional match-filtering scheme with the same complexity of hardware, especially in the lower SNRs.Second, a novel scheme of adaptive threshold setting is proposed for the rapidly-varying and fading channel. Based on the fact of the multipath component arriving in cluster, the scheme can eliminate the samples containing multipath signal components according to some rules and does not discard those larger samples simply. In this way, some larger samples merely containing noise component is reserved so that the estimation of noise power is more precisely and the threshold setting is more reasonably compared with the conventional schemes.Third, a code phase estimator is presented to improve the acquisition performance for DS/SS systems with larger Pseudo-Noise(PN) periods. In this scheme, the improving value and direction of the local PN generator is given based on the cross correlation between the incoming PN signal and the special auxiliary signal assuming that the channel is stationary and time-invariant in short time. In this way, the search can start from those code phase positions nearby sync cell with larger probability and reduce the mean acquisition time. For increasing the precision of phase estimator, some repetitious estimating outputs are also summed and averaged according to their sign and adopted the strategy of majority to decide which result is chosen at last. Analysis in theory and numerical results show that the proposed scheme can speedup the acquisition process several times compared with the conventional methods and the performance improves as the PN period increases.At last, a rapid timing acquisition scheme with multiple antennas is proposed for direct-spread ultra wideband(DS-UWB) systems, in which coarse estimation of initial frame is obtained in this dissertation based on the Cyclostationarity(CS) property inherent to UWB signals. We develop a data-aided approach to correlate the received waveform with the template taken from the received signals a symbol earlier to obtain the estimation based on the maximum likelihood(ML) criterion. Through utilizing spatial diversity from multiple antennas, the results of coarse estimation are very precise and the estimating errors are 20% at most, even less than one frame at higher SNRs.