ID	原文	译文
8234	该方法通过设计噪声统计特性梯度检测函数、敏感噪声统计特性的实际变化情况,利用窗口自适应函数实时计算窗口宽度,使得窗口在预设区间内自适应滑动,以适应实际噪声的变化。	This method through the design gradient noise statistical properties testing function, sensitive noise statistical characteristics of the actual changes, using the adaptive window function real-time calculation window width, make the window in the preset range adaptive sliding, to adapt to the change of the actual noise.
8235	仿真实验表明,基于滑动窗口的新息自适应组合导航算法可以有效跟踪噪声统计特性的实时变化,可同时兼顾自适应跟踪精度和跟踪灵敏度。	Simulation experiments show that the new rate adaptive integrated navigation based on sliding window algorithm can effectively track the noise statistical properties of real time change, can be both adaptive tracking precision and sensitivity.
8236	针对先验信息残缺的非合作电子对抗背景下的低截获概率雷达信号识别问题,提出一种基于改进的半监督朴素贝叶斯的识别算法。	In view of the a priori information incomplete non-cooperative electronic countermeasures under the background of low probability of intercept radar signal recognition problem, put forward a kind of based on improved a semi-supervised simple bayesian recognition algorithm.
8237	该算法首先提取出4种低截获概率(low probability of intercept,LPI)雷达信号的双谱对角切片作为识别特征;	The algorithm firstly extracted four kinds of low probability of intercept (low aim-listed probability of intercept, LPI) radar signal spectrum and diagonal slice as recognition characteristics;
8238	针对传统的半监督朴素贝叶斯(semi-supervised Na?ve Bayes,SNB)在更新训练样本集过程中会产生迭代错误的不足,利用改进的SNB(Revised SNB,RSNB)算法构建分类器,完成对测试样本的识别。	In view of the traditional half supervision and naive bayes (semi - supervised Na?Ve Bayes theorem, the SNB) produces in the process of updating the training sample set of iterative error, using the improved SNB (Revised SNB, RSNB) algorithm to construct classifier, complete the recognition of test sample.
8239	该方法通过在无标记样本集生成的置信度列表中选取置信度较高的样本添加到有标记样本集中,再利用预测后的分类结果对分类器参数(即特征期望向量珡mi和方差向量σi)进行改进,有效解决了传统算法分类精度低且分类性能不稳定等缺点。	By this method in unmarked sample set to generate confidence confidence higher samples selected from the list to add to the marked sample concentration, the classification of the reuse after predicting results of classifier parameters (i.e., characteristic vector 珡 mi expectation and variance vector sigma (I) was improved, effectively solve the traditional low classification accuracy and shortcomings and so on classification performance is not stable.
8240	理论分析和仿真结果表明,在LPI雷达信号识别问题,相比于SNB算法和传统的主成分分析加支持向量机法(principal component analysis-support vector machine,PCA-SVM),该算法具有更高的分类识别率和更好的分类性能。	Theoretical analysis and simulation results show that the LPI radar signal recognition problem, compared to the SNB algorithm and the traditional principal component analysis and support vector machine method (principal component analysis, support vector machine, PCA - SVM), the classification of the algorithm has higher recognition rate and better classification performance.
8241	针对当前type-Ⅱ准循环低密度奇偶校验(quasi-cyclic low-density parity-check,QC-LDPC)码的校验矩阵中存在权重为2的循环矩阵(weight-2circulant matrices,W2CM)导致Tanner图更容易产生短环,从而影响迭代译码收敛性的问题,基于完备循环差集(cyclic difference sets,CDS)提出了一种围长为8的type-ⅡQC-LDPC码的新颖构造方法。	In view of the current type - Ⅱ quasi-cyclic low density parity check (quasi - cyclic low density parity check, QC LDPC) existing in the check matrix code weight for 2 cyclic matrix (weight - 2 circulant matrices, W2CM) lead to Tanner graph is more likely to produce short loop, which affects the convergence of iterative decoding problem, based on the complete cycle of difference set (cyclic difference sets, CDS) put forward a type of girth is 8 - Ⅱ QC - novel construction method of LDPC codes.
8242	该方法构造的校验矩阵由权重为0的零矩阵、权重为1的循环置换矩阵和W2CM组成,保留了type-ⅡQC-LDPC码的具有更高最小距离上界的优点,改善了码的纠错性能;	The method to construct the check matrix by the weight of 0 zero matrix and weight of 1 cycle of permutation matrix and W2CM, retained the type -- Ⅱ QC LDPC code has the advantages of higher upper and lower bounds of the minimum distance to improve the error correction code performance;
8243	且Tanner图中无4、6环的出现,译码时具有较快的收敛速度。	No 4, 6, ring and Tanner graph, has faster convergence speed when decoding.