自從全球定位系統使用至今，愈來愈多個人化產品使得現代人使用導航定位系統的要求也趨於增加，其中以定位精度與定位速度最為要求。若使用傳統GPS接收器以提供定位，通常下載所需導航訊息(Broadcast Ephemeris and Satellite Almanac)須花費近12.5分鐘，因此目前許多研究目標即為縮短使用者第一次定位時間，以達到快速定位。
本研究中主要使用International GNSS Service (IGS)所提供之GPS精密星曆與Bernese軟體產生之精確衛星觀測量，並使用擴展型卡爾曼濾波器(Extended Kalman Filter, EKF)，將具有衛星位置或包含衛星速度之觀測量轉換為符合典型GPS接收器可使用的時效性精確型廣播星曆之15個軌道參數格式，特別是觀測量偏微分模型是一相當複雜的數學推導，因此透過加入狀態誤差、觀測量誤差等，可驗證其推導之正確性。
使用時效性精確型廣播星曆進而預估下段時間之衛星位置，與原始精密星曆比較，可得位置平均誤差從原本廣播星曆所提供之3公尺降至約0.4公尺內，亦可善加利用包含衛星速度觀測量以降低資料數目並強化衛星定位的準確。不同濾波器之運算方式不同，對衛星解算上精度之影響及其所需觀測量數目多寡等，分析EKF與Batch Least Squares Filter這兩種濾波器之實用性及差異。

Global Positioning System (GPS) has become more popular today. With the increase of personal mobile products, more demands from navigation and positioning systems are issued. Among them, the higher accuracy in position and velocity are one of the important key factors. A traditional GPS receiver usually takes 12.5 minutes to download the navigation message to provide location-based services, so that shortening the time-to-first-fix (TTFF) to achieve fast-positioning is a common goal of many research tasks. Since the precise ephemeris can be downloaded from International GNSS Service (IGS), it is possible to use the ephemeris to strengthen TTFF performance.
The primary approach of this study is to use Extended Kalman Filter (EKF) that converts the above IGS precise orbit (as measurement) into the precise broadcast ephemeris (as estimated state). Since the measurement partial differential is a complex mathematics model, therefore, adding the state error, the measurement error, and etc. with EKF approach can improves the accuracy of the state and the derived model.
In this study, the satellite’s positions at the sequential epochs have been predicted by the EKF-estimated precise broadcast ephemeris and compared with the true precise orbit. The average error of the satellite position can be reduced from 3 meters to 0.4 meters. When decreasing the number of measurements, combining with the satellite’s velocity is useful to enhance the accuracy of satellite position. In this study, the accuracy comparison between EKF and Batch Least Squares Filter (BLSF) has been analyzed. Under the same condition, the accuracy obtained with using EKF and BLSF are 0.4 meters and 0.62 meters, respectively. Clearly, the result shows that the orbit accuracy using EKF approach is better than that using BLSF.

[1]Elliott D. Kaplan, Christopher J. Hegarty(2005), Understanding GPS Principles and Application 2nd ed., ARTECH HOUSE, BOSTON, pp.1-3.
[2]Frank van Diggelen(2009), A-GPS: assisted GPS, GNSS, and SBAS, ARTECH HOUSE, BOSTON.
[3]Abidin H.Z.(1993), Computational and geometrical aspects of on-the-fly ambiguity resolution. Ph.D. dissertation, Dept. of Surveying Engineering Tech. Rep. No. 164, University of New Brunswick, Frederiction, New Brunswick, Canada.
[4]Wells D., Beck N., Delikaraoglou, D., Kleusberg A., Krakiwsky, E.J., Lachapelle, G., Langley, R.B., Nakiboglu, M., Schwarz, K.P., Tranqullia, J.M. and Vanicek, P.(1986), Guide to GPS Positioning, Canadian GPS Associates.
[5]Green G.B., Massatt PP.D. and Rhodus N.W.(1989), The GPS21 primary satellite constellation. Navigation, vol. 36. No.1, Spring, pp. 9-24.
[6]Bagley L.C. and Lamons J.W.(1992), NAVSTAR joint program office and a status report on the GPS program. Procs. 6th Int. Geodetic Symp. on Satellite Positioning, Columbus, Ohio, 17-20 March, 1992, pp. 21-30.
[7]Navigation Center(2010), IS-GPS-200E,
From http://www.navcen.uscg.gov/, pp. 42-43, 91-94
[8]International GNSS Service(2012), IGS Product Availability, From http://igscb.jpl.nasa.gov/components/prods_cb.html
[9]Rolf Dach, Urs Hugentobler, Pierre Fridez, Michael Meindl(2007), Bernese GPS Software Version 5.0, AIUB.
[10]Luis Serrano, Donghyun Kim, Richard B. Langley, Kenji Itani, Manmi Ueno(2004), A GPS Velocity Sensor: How Accurate Can It Be? A First Look, Department of Geodesy and Geomatics Engineering, University of New Brunwisk, Canada.
[11]Wen Zhang, Mounir Ghogho, Luis Enrique Aguado. Extension of GPS Broadcast Ephemeris to Determine Satellite Velocity and Acceleration, University of Leeds, Leeds, UK.
[12]Kalman R.E.(1960), A New Approach to Linear Filtering and Prediction Problems, Journal of Basic Engineering, vol. 82, pp. 35-45.
[13]Minkler G., Minkler J.(1993), Theory and Application of Kalman Filtering, Palm Bay, FL: Magellan Book Company.
[14]Gelb, A.(1974), Applied Optimal Estimation, Cambridge: M.I.T. Press
[15]Brown R.G. and Hwang P.Y.C.(1997), Introduction to Random Signals and Applied Kalman Filtering 3rd ed., JOHN WILEY & SONS.
[16]Bucy R.S. and Senne K.D.(1971), Digital Synthesis of Non-Linear Filters, vol. 7, Automatica, pp. 287.
[17]Godha S.(2006), Performance Evaluation of Low Cost MEMS-Based IMU Integrated with GPS for Land Vehicle Navigation Application, MSc Thesis. Department of Geomatics Engineering, University of Calgary, Calgary, Canada.
[18]Ching-Shun Ho(2007), Optimal Filtering Theory and Its Applications handout. Unpublished manuscript, IAA NCKU.
[19]Arthur Gelb(1974), Applied Optimal Estimation, Cambridge, Mass.: M.I.T. Press.
[20]Andrew H. Jazwinski(1970), Stochastic Process and Filtering Theory, New York: Academic Press.
[21]Guochang Xu(2007), GPS Theory, Algorithm and Applications 2nd ed., Springer, pp. 248-249.
[22]曾清涼、儲慶美(1999)，GPS衛星測量原理與應用第二版，國立成功大學衛星資訊研究中心，頁2-31─2-32，4-5，第五章。
[23]陳家宏(2009)，GPS廣播星曆估算─利用最小平方法，碩士論文，國立成功大學航空太空工程學系。
[24]張清義(2011)，應用GPS速度加強廣播星曆，碩士論文，國立成功大學航空太空工程學系。
[25]賴文鉦(2003)，GPS雙頻虛擬衛星發射器之設計、製作與測試，碩士論文，國立成功大學電機工程學系。
[26]魏子翔(2011)，時效性與週效性GPS廣播星曆預估，國立成功大學航空太空工程學系，頁25。
[27]何慶雄、翁錦堂、景國恩，GPS衛星精密軌道計算，國立成功大學衛星資訊中心，國立成功大學航空太空工程學系，頁2-3。