Research And Implementation On Key Techniques Of Signal Combining In Arraying Based On Structured Characteristics Of Communication Signal

Deep space communication has a very special application background that its signals, which experience dramatic path loss, are very weak when reaching the ground. An effective way to cope with low signal-to-noise ratio(SNR) for deep space communication is using multi-antenna arraying techniques, which has become an important trend in deep space communication. The efficiency of combination depends on the estimation and compensation of the signal parameter-difference such as time-delay, phase. Thus, it is significant to investigate the estimation methods of signal parameter-difference under low SNR condition.To improve the performance of parameter-difference estimation under low SNR conditions, this paper mainly researches time-delay and phase differences estimation under low SNR conditions by taking full advantage of structured characteristics of digital communication signal including symbol-by-symbol information transmission and cyclostationary. This paper also designs and implements an antenna array combination system based on a FPGA-centered intermediate frequency(IF) signal processing platform. The main work and innovations obtained in this dissertation are as follows.1. Aiming at time-difference estimation on dual-antenna array reception, this paper proposes a time-difference estimation algorithm based on weighted timing information which is extracted from symbol-by-symbol information transmission in digital communication. Combining time-difference estimates from timing information and estimates from the Generalized Cross Correlation(GCC) method, the observation model for real time-difference is built. Based on the model and analysis on correlation of observation errors, the least squares estimation and weighted least squares estimation of time-difference are given separately, whose physical concepts are also analyzed. It is indicated that the best linear unbiased estimator(BLUE) of time-difference is equal to the weighted least squares estimation with the use of the covariance matrix of observation errors. The simulation results show that rational use of timing information in digital communication from symbol-by-symbol information transmission can improve the performance of time-difference estimation under low SNR conditions compared with the classical GCC algorithm.2. Aiming at phase-difference estimation on dual-antenna array reception, this paper gives a phase-difference estimation algorithm based on timing-auxiliary included in symbol-by-symbol information transmission in digital communication. The paper analyzes the impact of non-optimal sampling point on the phase-difference estimation from the perspective that the higher the SNR is, the more propitious it is to estimate parameters. The algorithm obtains the phase differences based on correlation of the optimal sampling point sequence from timing information and interpolating recovery. The simulation results show that the performance of phase differences estimation based on timing-auxiliary can be significantly improved under low SNR conditions compared with the traditional ways which use all sampling points for correlation.3. Aiming at parameter-difference estimation on multi-antenna array reception, this paper introduces a cyclic-correlation-based multi-antenna time-delay/phase alignment algorithm based on cyclostationary of communication signals. The algorithm achieves estimation and compensation of time-delay and phase differences by iteration, aligning the time-delay/phase of each single signal during the iteration adjustment process. In the iteration process, the algorithm also gives the method to construct combining reference using second-order cyclostationary statistics. When solving the problem of time-difference, the summed envelope of the cross-correlation between single signal and other signals served as the cyclic-correlation combining reference. when solving the problem of phase differences, the observation vector and observation matrix are built using cyclic cross-correlation function between single signal and the sum of other signals and cyclic autocorrelation function of the single signal. The simulation results show that the algorithm can effectively improve the time-delay/phase alignment performance as the algorithm makes full use of cyclostationary of communication signals when iteration.4. An 8-antenna combination system supporting a maximum bandwidth of 36 MHz is designed and implemented. The system is aiming at the actual requirements of deep space communication. Then several main works, such as design of the whole system based on a FPGA-centered digital IF signal processing platform,a hardware and software co-design program based on Qsys, an automatic gain control(AGC) module and Nios II scheduling module are implemented. At last, the paper shows the testing results of the system that the gain of the 8-antenna signal combination system can reach to an average of 8.4dB on the condition that the normalized SNR of single-antenna reception signal is-6dB, verifying the feasibility of the whole system.