ID 原文 译文
6584 为提高估计方法的性能,提出一种基于模糊C类均值(fuzzy C-means,FCM)聚类的信源数估计方法。 To improve the performance of the estimation method is proposed based on fuzzy C mean (fuzzy C - means, FCM clustering of source number estimation method.
6585 在此方法中,信号的协方差阵的特征值及其差值,对协方差阵酉变换得到的盖氏圆半径及其差值作为FCM聚类过程中的样本指标。 In this approach, the signal of the eigenvalues of the covariance matrix and its difference, the covariance matrix of unitary transformation to get guy's circle radius and difference as the samples in the process of FCM clustering index.
6586 最终判断样本归属时,引入容错概率的概念。 The final judgment sample belonging, introduce the concept of probability of fault tolerance.
6587 在接受一定程度容错的前提下,提高了算法的性能。 In a certain degree of fault tolerance under the premise of improving the performance of the algorithm.
6588 在不同噪声类型下进行仿真的结果表明,所提出的信源数估计方法性能较好。 Simulation results show that under the different types of noise, the proposed source number estimation method has good performance.
6589 针对较低信噪比条件下二进制偏移载波信号的捕获问题,提出了一种基于批处理和Teager-Kaiser(TK)算子的联合捕获算法。 In view of the low SNR of binary offset carrier signal capture problem, put forward a kind of based on batch and Teager - Kaiser joint capture algorithm (TK) operator.
6590 该方法首先对接收信号和组合扩频码进行批处理和平均处理,并将接收信号转换到频域进行多次循环移位; This method first and combination of received signal spread spectrum code for batch and average processing, and will receive signal into frequency domain for multiple cyclic shift;
6591 然后,将得到的最佳频偏补偿序列与组合扩频码进行圆周相关运算得到最大相关值; Then, will be the best sequence and the combination of spread spectrum code deviation compensation on circular related operation to get maximum value;
6592 最后,引入相关峰能量最大值与次大值的比值作为门限判决变量,比较了所提TK算子法、差分相干累积算法和批处理算法的检测概率。 Finally, the introduction of correlation peak ratio of the maximum energy with great value as a threshold decision variable, compared the proposed TK operator method, differential coherent accumulation algorithm and batch detection probability of the algorithm.
6593 仿真结果表明,在同一条件下,TK算子法检测概率比差分相干算法提高了约2dB,比批处理算法提高了约6dB,并减少了捕获时间。 The simulation results show that under the same conditions, TK operator method than differential coherent detection probability algorithm by about 2 db, is about 6 db higher than batch algorithm, and reduce the time of capture.