The code tracking loop is a key component for user positioning. The pseudorange information of Bei Dou B1 signals has been fused and changed for vector tracking, so a correlation output model for complex scenarios is ...The code tracking loop is a key component for user positioning. The pseudorange information of Bei Dou B1 signals has been fused and changed for vector tracking, so a correlation output model for complex scenarios is designed to prevent the propagation of error and valuate the signal performance. The relevant software and hardware factors that affect the output are analyzed.A single channel time-division multiplexing(TDM) method for multicorrelation data extraction is proposed. Statistical characteristics of the correlation output data for both vector and scalar structures are evaluated. Simulation results show that correlation outputs for both structures follow normal or Chi-squared distributions in normal conditions, and the Gamma distribution in harsh conditions. It is shown that a tracking model based on the multi-channel fusion hardly changes the probability distribution of the correlation output in the normal case, but it reduces the ranging error of the code loop, and hence the tracking ability of the code loop for weak signals is improved. Furthermore, vector tracking changes the pseudorange characteristics of channels anytime, and affects the mutual correlation outputs of the code loops in the abnormal case. This study provides a basis for the subsequent design of autonomous integrity algorithms for vector tracking.展开更多
针对特征分解方法在实现非等功率同步直接序列码分多址(DS-CDMA)信号伪码序列盲估计时存在的处理数据向量不能太长以及不能工作于非平稳环境中的问题,引入了一种由主分量分析实现自适应特征提取的在线无监督学习(LEAP)神经网络(NN)。首...针对特征分解方法在实现非等功率同步直接序列码分多址(DS-CDMA)信号伪码序列盲估计时存在的处理数据向量不能太长以及不能工作于非平稳环境中的问题,引入了一种由主分量分析实现自适应特征提取的在线无监督学习(LEAP)神经网络(NN)。首先将已分段的一周期DS-CDMA信号作为NN的输入信号,用NN各权值向量的符号函数代表DS-CDMA信号各用户的伪码序列,然后通过不断输入信号来反复训练权值向量直至收敛,最终DS-CDMA信号各用户的伪码序列就可以通过各权值向量的符号函数重建出来。此外,采用变步长以提高收敛速度。理论分析与仿真实验表明,LEAP NN至少可以实现-20 d B信噪比下10个用户的非等功率同步DS-CDMA伪码序列盲估计,并且比传统的Sanger NN具有更快的收敛速度。展开更多
The energy preserving average vector field （AVF） method is applied to the coupled Schr6dinger-KdV equations. Two energy preserving schemes are constructed by using Fourier pseudospectral method in space direction di...The energy preserving average vector field （AVF） method is applied to the coupled Schr6dinger-KdV equations. Two energy preserving schemes are constructed by using Fourier pseudospectral method in space direction discretization. In order to accelerate our simulation, the split-step technique is used. The numerical experiments show that the non-splitting scheme and splitting scheme are both effective, and have excellent long time numerical behavior. The comparisons show that the splitting scheme is faster than the non-splitting scheme, but it is not as good as the non-splitting scheme in preserving the invariants.展开更多
In this paper, a pair of Mond-Weir type higher-order symmetric dual programs over arbitrary cones is formulated. The appropriate duality theorems, such as weak duality theorem, strong duality theorem and converse dual...In this paper, a pair of Mond-Weir type higher-order symmetric dual programs over arbitrary cones is formulated. The appropriate duality theorems, such as weak duality theorem, strong duality theorem and converse duality theorem, are established under higher-order （strongly） cone pseudoinvexity assumptions.展开更多
This paper deals with a class of inertial gradient projection methods for solving a vari-ational inequality problem involving pseudomonotone and non-Lipschitz mappings in Hilbert spaces.The proposed algorithm incorpor...This paper deals with a class of inertial gradient projection methods for solving a vari-ational inequality problem involving pseudomonotone and non-Lipschitz mappings in Hilbert spaces.The proposed algorithm incorporates inertial techniques and the projection and contraction method.The weak convergence is proved without the condition of the Lipschitz continuity of the mappings.Meanwhile,the linear convergence of the algorithm is established under strong pseudomonotonicity and Lipschitz continuity assumptions.The main results obtained in this paper extend and improve some related works in the literature.展开更多
基金supported by the National Natural Science Fundation of China(41474027)
文摘The code tracking loop is a key component for user positioning. The pseudorange information of Bei Dou B1 signals has been fused and changed for vector tracking, so a correlation output model for complex scenarios is designed to prevent the propagation of error and valuate the signal performance. The relevant software and hardware factors that affect the output are analyzed.A single channel time-division multiplexing(TDM) method for multicorrelation data extraction is proposed. Statistical characteristics of the correlation output data for both vector and scalar structures are evaluated. Simulation results show that correlation outputs for both structures follow normal or Chi-squared distributions in normal conditions, and the Gamma distribution in harsh conditions. It is shown that a tracking model based on the multi-channel fusion hardly changes the probability distribution of the correlation output in the normal case, but it reduces the ranging error of the code loop, and hence the tracking ability of the code loop for weak signals is improved. Furthermore, vector tracking changes the pseudorange characteristics of channels anytime, and affects the mutual correlation outputs of the code loops in the abnormal case. This study provides a basis for the subsequent design of autonomous integrity algorithms for vector tracking.
文摘针对特征分解方法在实现非等功率同步直接序列码分多址(DS-CDMA)信号伪码序列盲估计时存在的处理数据向量不能太长以及不能工作于非平稳环境中的问题,引入了一种由主分量分析实现自适应特征提取的在线无监督学习(LEAP)神经网络(NN)。首先将已分段的一周期DS-CDMA信号作为NN的输入信号,用NN各权值向量的符号函数代表DS-CDMA信号各用户的伪码序列,然后通过不断输入信号来反复训练权值向量直至收敛,最终DS-CDMA信号各用户的伪码序列就可以通过各权值向量的符号函数重建出来。此外,采用变步长以提高收敛速度。理论分析与仿真实验表明,LEAP NN至少可以实现-20 d B信噪比下10个用户的非等功率同步DS-CDMA伪码序列盲估计,并且比传统的Sanger NN具有更快的收敛速度。
基金supported by the National Natural Science Foundation of China(Grant No.91130013)the Open Foundation of State Key Laboratory of HighPerformance Computing of China
文摘The energy preserving average vector field （AVF） method is applied to the coupled Schr6dinger-KdV equations. Two energy preserving schemes are constructed by using Fourier pseudospectral method in space direction discretization. In order to accelerate our simulation, the split-step technique is used. The numerical experiments show that the non-splitting scheme and splitting scheme are both effective, and have excellent long time numerical behavior. The comparisons show that the splitting scheme is faster than the non-splitting scheme, but it is not as good as the non-splitting scheme in preserving the invariants.
基金Supported by the National Natural Science Foundation of China(No.11431004,11271391 and 11201511)the Natural Science Foundation of Chongqing(CSTC2014pt-sy00001,CSTC2015jcyj A00005)the Education Committee Project Research Foundation of Chongqing(KJ1500309,KJ1400519)
文摘In this paper, a pair of Mond-Weir type higher-order symmetric dual programs over arbitrary cones is formulated. The appropriate duality theorems, such as weak duality theorem, strong duality theorem and converse duality theorem, are established under higher-order （strongly） cone pseudoinvexity assumptions.
文摘This paper deals with a class of inertial gradient projection methods for solving a vari-ational inequality problem involving pseudomonotone and non-Lipschitz mappings in Hilbert spaces.The proposed algorithm incorporates inertial techniques and the projection and contraction method.The weak convergence is proved without the condition of the Lipschitz continuity of the mappings.Meanwhile,the linear convergence of the algorithm is established under strong pseudomonotonicity and Lipschitz continuity assumptions.The main results obtained in this paper extend and improve some related works in the literature.
