低信噪比DS-SS/BPSK信号检测与参数估计技术研究

【作者】 沈振惠；

【导师】 唐斌；

【作者基本信息】 电子科技大学 ， 通信与信息系统， 2003， 硕士

【摘要】 本文研究了包括平方倍频检测、周期谱检测、时域相关检测和高阶统计量检测四种针对低信噪比下DS-SS/BPSK信号的检测与参数估计方法。平方倍频方法最初用于合作方间的DS-SS/BPSK信号检测，对它的结构稍作修改，可用于侦察检测中估计DS-SS/BPSK信号的频率，性能良好。将它与相关积累技术和功率谱集平均技术相结合，能提高该方法的时效性。周期谱检测方法将DS-SS/BPSK信号建模为周期平稳过程，通过对DS-SS/BPSK信号周期谱的切片进行谱峰搜索，可检测出信号频率。时域相关检测方法利用扩频信号的周期自相关特性，能在中频和基带对扩频码周期进行估计，仿真结果表明中频检测时性能较好。高阶统计分析技术是检测直接序列扩频信号的另一途径，本文提出了一类基于和四阶统计量切片的DS-SS/BPSK信号的检测与参数估计方法。其中，基于四阶累积量2-D切片的检测方法能够很好地抑制高斯噪声，并且兼顾了时域相关检测和平方倍频检测两种方法的特点，能够同时估计出载波频率和扩频码周期两个参数。以上方法除周期谱方法以外，都已在电子29所的实际系统上通过了验证，达到各项指标。更多还原

【Abstract】 In this paper, the detection and parameter estimation, are studied for DS-SS/BPSK signal with low signal-to-noise ratio, including square detection, cyclic spectrum detection, correlation detection and higher-order statistics-based detection. The square detection, which was used in detection for DS-SS/BPSK signal between cooperations, is modified to estimate the frequency of non-cooperative weak DS-SS/BPSK signals with an acceptable performance. When combined with correlation accumulate technique and power spectrum collection-average technique, the temporal efficiency is enhanced. In the cyclic spectrum detection, the DS-SS/BPSK signal is modeled as cyclostationary signal. By searching the peaks on the slice of DS-SS/BPSK signals cyclic spectrum, the carrier frequency can be estimated. The correlation detection using the periodicity of the direct sequence autocorrelation function can estimate the symbol period at intermediate frequency or at baseband. Simulation results indicate that the former is superior to the latter when the demodulated frequency is unequal to the carrier frequency. The detection and parameter estimation based on high-order statistics are another approach for DS-SS/BPSK signals. In this paper, a sort of detectors based on fourth-order statistics slices are presented to detect the presence of signal and estimate the carrier frequency and symbol period of DS-SS/BPSK signal. Because the detector based on the fourth-order cumulant 2-D slice, which has better ability to suppress Gaussian noise, possesses both characters of square detector and correlation detector, it can estimate not only the carrier frequency, but also the symbol period.The above-mentioned detection and parameter estimation methods, but cyclic spectrum detection, have all been run in the real system and the results reach the desired indexes.更多还原