Aiming at the problem of rank deficiency of orbit determination algorithm equation caused by weak navigation signal, poor geometric structure of signal and small number of visible satellites of a GEO satellite, a weighted least square orbit determination algorithm with additional constraints are proposed based on the relative geostationary characteristics of GEO satellite, which can improve the success rate of orbit determination of GEO satellite navigation signal when the number of navigation satellites are less than 4 and the measurement value is poor. Shorten the initialization time of the receiver. Based on the characteristics of strong radial constraint, weak lateral constraint, strong lateral constraint and strong longitudinal constraint of the ground-based orbit measurement data, the feasibility of the above algorithm is verified by the space-based heterogeneous data fusion orbit determination test. The results show that the accuracy of the two kinds of orbit determination test results are the same, and the root mean square error of the total position in the overlapping trackarc is superior to 5 m. The root mean square error of the total velocity is superior to 2 mm/s. It is concluded that the effectiveness of GNSS system supporting orbit determination of high orbit spacecraft is verified, which provides technical support for the application of GNSS orbit determination mission in the future.

The openness of Android operating system not only brings convenience to users, but also leads to the attack threat from a large number of malicious applications (apps). Thus malware detection has become the research focus in the field of mobile security. In order to solve the problem of more coarse-grained feature selection and larger feature loss of graph structure existing in the current detection methods, we put forward a method named DGCNDroid for Android malware detection, which is based on the deep graph convolutional network. Our method starts by generating a function call graph for the decompiled Android application. Then the function call subgraph containing the sensitive application programming interface (API) is extracted. Finally, the function call subgraphs with structural features are trained as the input of the deep graph convolutional network. Thus the detection and classification of malicious apps can be realized. Through experimentation on a dataset containing 11,120 Android apps, the method proposed in this paper can achieve detection accuracy of 98.2%, which is higher than other existing detection methods.

During re-entry objects with low-eccentricity orbits traverse a large portion of the dense atmospheric region almost every orbital revolution. Their perigee decays slowly, but the apogee decays rapidly. Because ballistic coefficients change with altitude, re-entry predictions of objects in low-eccentricity orbits are more difficult than objects in nearly circular orbits. Problems in orbit determination, such as large residuals and non-convergence, arise for this class of objects, especially in the case of sparse observations. In addition, it might be difficult to select suitable initial ballistic coefficient for re-entry prediction. We present a new re-entry prediction method based on mean ballistic coefficients for objects with low-eccentricity orbits. The mean ballistic coefficient reflects the average effect of atmospheric drag during one orbital revolution, and the coefficient is estimated using a semi-numerical method with a step size of one period. The method is tested using Iridium-52 which uses sparse observations as the data source, and ten other objects with low-eccentricity orbits which use TLEs as the data source. We also discuss the performance of the mean ballistic coefficient when used in the evolution of drag characteristics and orbit propagation. The results show that the mean ballistic coefficient is ideal for re-entry prediction and orbit propagation of objects with low-eccentricity orbits.