Table of Contents Author Guidelines Submit a Manuscript
International Journal of Aerospace Engineering
Volume 2018, Article ID 7547128, 13 pages
Research Article

The Influence of Orbital Element Error on Satellite Coverage Calculation

School of Computer Science, Hubei Key Laboratory of Intelligent Geo-Information Processing, China University of Geosciences, Lumo Road, Wuhan 430074, China

Correspondence should be addressed to Xiaoyu Chen; moc.621@liamyxc

Received 29 November 2017; Revised 5 April 2018; Accepted 14 May 2018; Published 5 July 2018

Academic Editor: Enrico C. Lorenzini

Copyright © 2018 Guangming Dai et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.


This paper studies the influence of orbital element error on coverage calculation of a satellite. In order to present the influence, an analysis method based on the position uncertainty of the satellite shown by an error ellipsoid is proposed. In this error ellipsoid, positions surrounding the center of the error ellipsoid mean different positioning possibilities which present three-dimensional normal distribution. The possible subastral points of the satellite are obtained by sampling enough points on the surface of the error ellipsoid and projecting them on Earth. Then, analysis cases are implemented based on these projected subastral points. Finally, a comparison report of coverage calculation between considering and not considering the error of orbital elements is given in the case results.

1. Introduction

The calculation of satellite and constellation coverage of ground targets is a basic problem of Earth observation systems. Various kinds of satellites and constellations are available and selectable according to different requirements of mission design. As the basic purposes of constellation design, continuous coverage and maximum regional coverage are often considered. Hence, many proposed methods were devoted to how to optimize these two goals. Cote [1] offered tips on how to maximize satellite coverage while cruising in the high-latitude West Coast waters of British Columbia. Yung and Chang [2] proposed a method for maximizing satellite coverage at predetermined local times for a set of predetermined geographic locations. Draim [3] described a new four-satellite elliptic orbit constellation giving continuous line-of-sight coverage of every point on Earth’s surface.

During the implementation of a mission, the visible time windows of important targets are also usually considered. For example, Lian et al. [4] proposed two static and dynamic models of observation toward Earth by agile satellite coverage to decompose area targets into small pieces and compute visible time window subarea targets. Wang et al. [5] analyzed four parameters to determine the coverage area of satellite. It also introduced and compared two current test methods of satellite area, which presents the composition of the required test equipment and test processes, and proved the effectiveness of the tests through engineering practices. Besides the numerical analysis methods, heuristic search algorithms such as particle swarm optimization [6] were also applied to the satellite coverage calculation. Huang and Dai [7] proposed an optimal design of constellation of multiple regional coverage based on NSGA-II with the aid of multiobjective optimization.

The error influence of orbital elements on the satellite positioning was mainly considered in the field GIS and radar location [810]. Analysis methods proposed in these papers usually employed a model that can transform the possible position of a satellite to the points in an error ellipsoid. Nelson [11] introduced and discussed some basic properties of the error ellipsoid. For example, the basic parameters of the error ellipsoid were depicted by the axial direction and the axial ratio among each axis. In the study of satellite positioning error, the error ellipsoid was used in the positioning accuracy analysis of global navigation satellite system [12]. Some papers [1315] introduced that the axis ratio of error ellipsoid for GPS should be controlled within a certain range. Wang et al. [16] selected eight sites to study the different shapes of the error ellipsoid in the global range. Alfano [17] showed the effects of positional uncertainty on the Gaussian probability computation for orbit conjunction. A model of position uncertainty for the TIS-B System was proposed [18]. Shi et al. [19] introduced a distributed location model and algorithm with position uncertainty. Trifonov [20] introduced position uncertainty measures for a particle on the sphere.

However, for coverage calculation, the influence of the orbital element error has not been considered, which is the research purpose of this paper. Because only the error influence of initial orbital elements on coverage calculation is studied in this paper, hence, the influence of other variables, such as perturbation, are not considered here. Because of the nonlinear function between satellite position and orbital elements, assuming the error of orbital elements presents normal distribution, therefore, the uncertainty of the satellite position will also present normal distribution in the three-dimensional space. The relationship between the error of satellite position and orbital element error can be obtained according to the knowledge of error propagation and the function relationship between the satellite position and orbital elements, which is the basis of coverage uncertainty analysis.

The rest of this paper is organized as follows. In Section 2, the proposed method is introduced. In Section 3, preparations for cases are given. In Section 4, five analysis cases are implemented. Section 5 is devoted to the analysis of experimental results and summary of this paper. Section 6 gives some conclusions.

2. Coverage Uncertainty Related to Orbital Elements

The uncertainty analysis method proposed in this paper includes four steps, which are summarized as follows: Step 1:obtain the position vector covariance matrix based on the given orbital element error matrix and transformation matrix.Step 2:calculate the eigenvalues and eigenvectors of the position vector covariance matrix. The eigenvalue is the axial length of the error ellipsoid, and the eigenvector is the axial direction of the error ellipsoid. Then, perform the rotation and translation operations on the error ellipsoid.Step 3:extract the edge points of the error ellipsoid projection on the surface of Earth.Step 4:obtain the coverage analysis results to ground targets based on the extracted edge points.

In the above four steps, the implementations of steps 1 and 2 are to obtain the necessary parameters about error ellipsoid. Steps 3 and 4 are utilized to project all the points of error ellipsoid surface on Earth for the further coverage calculation. Process simulation presented in Figure 1 shows that each extracted point of the projection edge is regarded as the worst position of satellite positioning, and all the analysis work are implemented based on these points. The method employed to extract the edge points is a convex hull algorithm shown in Table 1. The analysis method metioned in step 4 is introduced in Table 2 with more details. Sections 2.1 and 2.2 describe the specific process of obtaining an error ellipsoid.

Figure 1: Simulation of uncertainty analysis process.
Table 1: Convex hull algorithm.
Table 2: Method of analyzing error influence.
2.1. Obtain the Covariance Matrix of the Satellite Position

Considering the nonlinear relationship between the satellite position and orbital elements, the covariance matrix and propagation matrix between orbital elements and position uncertainty can only be obtained by solving partial derivative, which can be shown by where is an square matrix. In this paper, represents the covariance matrix of satellite position in directions , , and . is the error matrix of orbital elements.

When is a nonlinear function of , such as the partial derivative of by partial derivative of needs to be obtained by the following method:

Then, (5) can be obtained based on (3) and represented by

The nonlinear function between the satellite position vector and orbital elements is shown in (6), (7), and (8), where , , , , , and are orbital elements and and are functions of , , and . Vector is decomposed into three vectors in the directions , , and , and in each direction, the partial derivatives of orbital elements can be obtained. For example, the partial derivative of by that of , , , , , and can be presented in (9). Similarly, gaining the partial derivatives of and by that of , , , , , and , respectively, can finally obtain the error propagation matrix.

2.2. Rotation and Translation of the Error Ellipsoid

Assume that the major axis direction of error ellipsoid is

So the rotation angle of major axis around axis should be

Then, the rotation angle of major axis around axis is

Subsequently, all points need to be rotated around axis . Based on the previous rotations, the third rotation angle is equal to the angle between axis and middle axis of error ellipsoid.

Finally, if the standard position of satellite is , the point needs to be translated to a new position.

2.3. Feature of the Error Ellipsoid

It is assumed that the measurement errors of six orbital elements present normal distribution. So the position error of satellite also presents three-dimensional normal distribution [5, 16]. The distribution density of possible satellite position in three-dimensional space is where is the position increment.

The denominator of (13) is a constant, while the numerator is changeable. So all points with the same distribution density can be presented by

Equation (14) is a presentation of ellipsoid cluster, and is a magnification factor. If is symbolized by (14) can also be shown by

Because it is not convenient to show the ellipsoid in coordinate system , the principal axis coordinate system is employed to present it. For a real symmetric matrix , an orthogonal matrix can be utilized to make and

According to (18), (15) can be shown by

Equation (19) can be expanded to

Assume that each axis of the error ellipsoid is symbolized by , , and ; hence,

As mentioned before, the position of the satellite is an uncertainty position distribution within error ellipsoid interior. This distribution probability can be presented by giving a variable substitution, such as

The probability in the ellipsoid is converted to the probability in the ball as follows:

In addition, considering that (24) can be expanded to where the probability of the point distribution in the ellipsoid varies with the change of as follows: (i)When , the probability of the point distribution in the error ellipsoid is 19.9%.(ii)When , the probability of the point distribution in the error ellipsoid is 73.9%.(iii)When , the probability of the point distribution in the error ellipsoid is 95%.

3. Preparation for Coverage Case

3.1. Calculation of the Error Ellipsoid

According to the measurement accuracy in actual engineering project, the error matrix of orbital elements is set as follows: where is a conversion factor between angle and radian.

The orbital elements of a satellite are shown in Table 3. In addition, as a requirement of parameter transformation, the conversion relationship between eccentric anomaly and true anomaly is needed:

Table 3: Data of orbital elements.

The covariance matrix of the satellite position is obtained as follows:

The feature vector of is

The data of axis length and rotation angle are shown in Tables 4 and 5, respectively. The standard position coordinate of the satellite position is (3600.6, 5193.1, 3648.4). The final simulation of the error ellipsoid with different views is shown in Figure 2.

Table 4: Data of axis length.
Table 5: Data of rotation angle.
Figure 2: Space distribution of error ellipsoid.
3.2. Edge Extraction of Projection on Earth

The projection of error ellipsoid on Earth is a curved surface, so in order to obtain the edge points, a mapping on plane is more convenient. 421 edge points are obtained in the case by using convex hull algorithm, with different views of simulation results shown in Figure 3.

Figure 3: Result of edge extraction.

4. Numerical Case

Five cases with different configures are designed and implemented in this paper. Firstly, the uncertainty influence of single satellite positing on multiple point target coverage is given in case 1. Then, in order to deeply understand the shape change of “error ellipsoid” when the satellite moves to different positions, the changes of axial length and axial ratio are given in case 2. Lastly, the uncertainty influence of a constellation positioning on multiple point targets, constellation positioning on a single area target, and constellation positioning on two area targets are given in cases 3 to 5, respectively.

The reference system used in this paper is ECI. When the satellite is in the initial position perigee, the latitude and longitude of the satellite are 55.2°E and 30°N, respectively. The orbit and ground point targets are shown in Figure 4 with 3D and 2D views.

Figure 4: Satellite orbit and point targets.
4.1. Influence of Single Satellite Position Uncertainty on Multiple Point Target Coverage

The purpose of case 1 is to explain the necessity of considering orbital element error in coverage calculation. Six point targets are set in this case, whose coordinates are presented in Table 6, obtaining the different minimum coverage requirements of these point targets in terms of geocentric angles considering and not considering positioning uncertainty, respectively. Case results listed in Table 7 show that the geocentric angle needs to be larger to cover all the aimed targets when considering error of orbital elements. The results shown in Figure 5 obviously present the relative percentage of needed geocentric angle when considering uncertainty.

Table 6: Coordinate of the six point targets.
Table 7: Results of case 1.
Figure 5: Simulation comparison of case results.
4.2. Shape Change of the “Error Ellipsoid”

The purpose of case 2 is to research on the shape change of the error ellipsoid. Analyze the axial length variation while true anomaly changes from 0° to 360°. Focus on the shape change by studying the variations of axial length and axial ratio, which are shown in Figures 6 and 7, respectively. Interesting conclusions can be drawn from the case results. The most significant one is that the 3D shape of error ellipsoid is a time-dependent function, and the error ellipsoid is rather slender when true anomaly changes around 110°. While anomaly reaches around 180°, the shape is stouter.

Figure 6: Changes of axial length.
Figure 7: Changes of axial ratio.
4.3. Uncertainty Influence of a Constellation Positioning on Multiple Point Targets

The implementation of case 3 is to study the coverage uncertainty of constellation containing 3 satellites whose orbital elements are shown in Table 8. Two point targets A (3159.6162, 5472.44908, and 3648.4) and B (1097.44509, 6222.5133, and 3648.4) shown in Figure 8 are set in this case. The latitude and longitude of these two points are (60°E, 30°N) and (80°E,30°N), respectively. Case results in Table 9 show that when considering positioning uncertainty, the minimum required covering radius is larger than that of not considering uncertainty. If it is needed to cover both points A and B at the same time, the minimum radius must be 1484.017928 km.

Table 8: Orbital element of each satellite.
Figure 8: Positions of points A and B.
Table 9: Minimum coverage radius of satellites.
4.4. Uncertainty Influence of Constellation Positioning on a Single Area Target

We have mentioned that 421 edge points are obtained in the case by using convex hull algorithm. In order to reduce the calculation expense, 54 edge points are chosen evenly to implement the coverage case containing 96,576 ground target points. Only when each of the 96,576 sampling points can be covered by the satellite, whose subastral point is any one of the 54 edge points, the area is deemed covered completely. Assuming that the target is a circular area, and the radius of this area is 100 km, compute the minimum sensor angle meeting the coverage requirements of this area when the angle of the satellite sensor changes from 7°. Case results are shown in Tables 10, 11, 12, and 13. In order to present this change significantly, relevant simulation results of these data are shown in Figures 912, respectively. As well, the percentage change of area coverage is also shown in Figure 13.

Table 10: Coverage results of a circular area when the sensor angle is 7°.
Table 11: Coverage results of a circular area when the sensor angle is 7.4°.
Table 12: Coverage results of a circular area when the sensor angle is 7.7°.
Table 13: Coverage results of a circular area when the sensor angle is 8°.
Figure 9: Coverage results of a circular area when the sensor angle is 7°.
Figure 10: Coverage results of a circular area when the sensor angle is 7.4°.
Figure 11: Coverage results of a circular area when the sensor angle is 7.7°.
Figure 12: Coverage results of a circular area when the sensor angle is 8°.
Figure 13: Percentage change of area coverage.
4.5. Uncertainty Influence of Constellation Positioning on Two Area Targets

Assuming that A and B are circular areas with circle . The minimum coverage radius should be calculated with respect to the requirement of the complete coverage from satellite 2. From case results shown in Table 14, we know that the minimum coverage radius of satellite 2 for coverage requirement is 1684.017928 km.

Table 14: Minimum coverage radius for each target.

5. Results and Discussion

Above cases have shown the influence of orbital element errors on coverage calculation. It is obvious that the minimum angle requirement is larger when considering error of orbital elements. From case 1, it can be seen that the influence of orbital element error on coverage is more significant when the required geocentric angle is smaller. As the required geocentric angle increases, the impact of orbital element error on coverage will be smaller and smaller. From case 2, an interesting phenomenon can be seen that when true anomaly is about 180°, and it is a turning point for the changes of axial length and axial ratio. Cases 3–5 also show the analysis results of different coverage problems based on the Monte Carlo simulation for area targets. These cases show that under the case settings in this paper, the extra requirements of sensor angle grows by about 14.3%. The analysis implemented in this paper is based on the fixed positions of satellite or constellation. Hence, in future research, this work can be extended to analyze the real-time coverage error related to positioning uncertainty.

6. Conclusions

In this paper, we present a method of accurately calculating the constellation coverage when considering position uncertainty. The position uncertainty of satellite is shown by an “error ellipsoid,” whose spatial shape is also analyzed in the paper. For the illustration of proposed method, coverage calculations for a single satellite and constellation are both considered to carry out the coverage cases. The cases show that when considering the position uncertainty related to influence of orbital error, in order to surely cover the target points, the covering radius needs to be larger than not considering position uncertainty. The work presented in this paper also has important reference value for the reliability analysis of constellation design, which can effectively deal with the influence of position uncertainty.

Conflicts of Interest

The authors declare that there is no conflict of interests regarding the publication of this paper.


This work is supported by National Natural Science Foundation of China under Grant no. 61472375, the 13th Five-Year Pre-Research Project of Civil Aerospace in China, the Joint Funds of Equipment Pre-Research and Ministry of Education of the People’s Republic of China under Grant no. 6141A02022320, Fundamental Research Funds for the Central Universities under Grant nos. CUG2017G01 and CUG160207, and the Reform and Research of Graduate Education and Teaching under Grant no. YJS2018308.


  1. J. Cote, Maximizing Satellite Coverage, Pacific Yachting, 2012.
  2. K. W. Yung and D. C. Chang, U.S. Patent No. 7,198,230, U.S. Patent and Trademark Office, Washington, DC, USA, 2007.
  3. J. E. Draim, “A common-period four-satellite continuous global coverage constellation,” Journal of Guidance, Control, and Dynamics, vol. 10, no. 5, pp. 492–499, 1987. View at Publisher · View at Google Scholar · View at Scopus
  4. Z. Lian, Y. Tan, Y. Xu, and J. Li, “Static and dynamic models of observation toward Earth by agile satellite coverage,” in Proceedings of International Workshop on Planning and Scheduling for Space, pp. 1–6, Darmstadt, Germany: ESOC, 2011.
  5. H. B. Wang, F. Xue, and C. J. Liu, “Test method of satellite coverage area and its engineering application,” Radio Communications Technology, vol. 3, p. 19, 2010. View at Google Scholar
  6. J. A. Ojeda Romero, Dual Satellite Coverage using Particle Swarm Optimization Doctoral dissertation, Virginia Tech, 2014.
  7. X. Huang and G. Dai, “An optimal design of constellation of multiple regional coverage based on NSGA-II,” International Journal of Advancements in Computing Technology, vol. 4, no. 21, pp. 295–303, 2012. View at Publisher · View at Google Scholar · View at Scopus
  8. E. S. Fekpe, T. K. Windholz, and K. Beard, “Probabilistic approach for modeling and presenting error in spatial data,” Journal of Surveying Engineering, vol. 135, no. 3, pp. 101–112, 2009. View at Publisher · View at Google Scholar · View at Scopus
  9. E. D. Kaplan, “Understanding GPS: principles and application,” Journal of Atmospheric and Solar-Terrestrial Physics, vol. 59, no. 5, pp. 598-599, 2005. View at Google Scholar
  10. P. Misra and P. Enge, “Global Positioning System: Signals, Measurements and Performance,” Tech. Rep., Ganga-Jamuna Press, 2011. View at Google Scholar
  11. W. Nelson, Use of Circular Error Probability in Target Detection, MITRE Corporation, Bedford MA, USA, 1988.
  12. J. C. Roh and D. W. Waters, “Position and velocity uncertainty metrics in GNSS receivers,” Tech. Rep., US Patent US8525727, 2013. View at Google Scholar
  13. E. M. Mikhail and F. E. Ackermann, Observations and Least Squares, Harper and Row, New York, NY, USA, 1976.
  14. M. Nomura and T. Tanaka, “GPS positioning method under condition of only three acquired satellites,” SICE Journal of Control, Measurement, and System Integration, vol. 2, no. 4, pp. 206–212, 2009. View at Publisher · View at Google Scholar
  15. F. D. Nunes, F. M. G. Sousa, and J. M. N. Leitao, “Performance analysis of altimeter-aided GNSS receiver for indoor scenarios,” in The 7th Conference on Telecommunications, Santa Maria da Feria, Portugal, May 2009.
  16. M. Wang, L. Ma, L. Zhang, and H. Ji, “Analysis of horizontal positioning error distribution of the Chinese area positioning system,” Astronomical Research & Technology, vol. 9, no. 3, pp. 522–526, 2012. View at Google Scholar
  17. S. Alfano, “Relating position uncertainty to maximum conjunction probability,” Journal of the Astronautical Sciences, vol. 53, no. 2, pp. 193–205, 2004. View at Google Scholar
  18. Z. Qingzhu, Z. Jun, Z. Yanbo, and L. Wei, “Navigation accuracy category-position models and estimate position uncertainty calculations for tis-b system,” Chinese Journal of Aeronautics, vol. 22, no. 4, pp. 419–425, 2009. View at Publisher · View at Google Scholar · View at Scopus
  19. Y. L. Shi, Q. M. Cui, S. Q. Cao, and X. F. Tao, “Distributed location model and algorithm with position uncertainty,” The Journal of China Universities of Posts and Telecommunications, vol. 21, no. 2, pp. 48–56, 2014. View at Publisher · View at Google Scholar · View at Scopus
  20. D. A. Trifonov, “Position uncertainty measures on the sphere,” in Proceedings of the Fifth International Conference on Geometry, Integrability and Quantization, pp. 211–224, Institute of Biophysics and Biomedical Engineering, Bulgarian Academy of Sciences, 2004.