Degraded GPS Signal Measurements With A Stand-Alone High Sensitivity Receiver

Transcription

1 Degraded GPS Signal Measurements With A Stand-Alone High Sensitivity Receiver G. MacGougan, G. Lachapelle, R. Klukas, K. Siu, Department of Geomatics Engineering L. Garin, J. Shewfelt, G. Cox, SiRF Technology Inc. BIOGRAPHIES Glenn MacGougan is a MSc. student in the Department of Geomatics Engineering, University of Calgary. In 2000 he completed a BSc. in Geomatics Engineering at the same institution. He has previous experience in GPS related R&D at NovAtel Inc. and Trimble Navigation. He expects to complete his MSc in September Dr. Gérard Lachapelle is Professor and Head of the Department of Geomatics Engineering, University of Calgary. He has been involved with GPS developments and applications since More information on site Dr. Richard Klukas is an Assistant Professor in the Department of Geomatics Engineering at the University of Calgary. He holds BSc and MSc degrees in Electrical Engineering and a PhD in Geomatics Engineering, all from the University of Calgary. His research interests include all aspects of wireless location. Lap Kee Siu is a BSc. student of Electrical Engineering at the University of Calgary. He currently works as an internship student in the Department of Geomatics Engineering. He expects to complete his BSc. in May Lionel J. Garin, Lead GPS Architect, SiRF Technology Inc., has over 20 years of experience in GPS and communications fields. Previous to that, he worked at Ashtech, SAGEM and Dassault Electronique. He is the inventor of the "Enhanced Strobe Correlator" code and carrier multipath mitigation technology. He holds an MSEE equivalent degree in digital communications sciences and systems control theory from Ecole Nationale Superieure des Telecommunications, France and BS in physics from Paris VI University. Geoffrey F. Cox holds a B.A. degree in Geology and Chemistry at the University of Maine and a MEng. in Geomatics Engineering from the University of Calgary. From 1996 to 2000, he worked in various areas of GPS. Mr. Cox joined SiRF Technology, Inc. in the fall of 2000 as a Senior Applications Engineer. John L. Shewfelt received a B.Sc. in Electrical Engineering from the University of California Santa Barbara in He has since been involved in the design, development, test and integration of advanced avionics and guidance systems. In 2000, he joined SiRF Technology Inc. as Applications Engineering Manager to facilitate the integration of GPS technology into embedded products and platforms. ABSTRACT The use of GPS for personal location using cellular telephones or other devices requires signal measurements under both outdoor and indoor situations. The outdoor environment may range from clear to shaded/blocked signal measurements. The indoor environment may range for single floor wooden constructions to high-rise buildings and underground facilities. In this paper, a high sensitivity receiver that operates in unaided stand-alone mode is tested under a range of shaded signal environments, ranging from residential outdoor areas to urban canyons to residential houses. The measurement analysis is performed in both the observation and position domains. The results show that the receiver tested can yield measurements with C/N 0 Presented at ION National Technical Meeting, San Diego, January

2 degradations in excess of 20dB-Hz, as compared to lineof-sight measurements. Position results are a function of the geometry of the remaining satellites, which in turn is a function of the environment. INTRODUCTION GPS signal deterioration occurs through signal masking caused by natural (e.g. foliage) and man-made (e.g. buildings) obstructions, ionospheric scintillation, Doppler shift, multipath, jamming, evil waveforms, and receiver and antenna effects. The impact of anyone of the above can result in partial to total loss of signal tracking and/or tracking errors, depending on the severity of the effect and the receiver tracking characteristics. These effects are evident in a receiver's measure of the carrier to noise density ratio C/N 0. Tracking errors, especially if undetected by the receiver firmware, can result in large position errors. Partial loss of tracking results in geometry degradation, which in turn affects position accuracy. The L1 C/A Code repeats every millisecond. This can be used advantageously by the GPS receiver in that the signal can be integrated for extended periods in order to obtain a higher signal to noise ratio. Chansarkar & Garin (2000) describes the use of GPS signals at very low power levels using long dwell times. In terms of unaided GPS, this integration can be performed coherently for up to 20ms. The maximum coherent integration time is due to the navigation bit boundaries. Furthermore, non-coherent integration, which is basically integration of the squared signal, can be performed for long periods of time relative to the coherent integration interval. Using the full coherent interval and long non-coherent integration times, weak signal tracking in degraded environments is possible. Prior investigations into the use of low power GPS signals using long dwell times have been performed by Peterson et al. (1997), Moeglein & Krasner (1998), Garin et al (1999), and van Diggelen & Abraham (2001). Testing at the University of Calgary focussed on the use of long integration times for the stand-alone case that is no network aiding. There is a strong need to characterize unaided receiver performance under GPS signal deterioration to extend the use of GPS to a range of new applications. RECEIVER DESCRIPTION The receiver type under test is a high sensitivity (HS) unaided SiRF receiver unit. For comparison purpose, a standard model SiRF receiver and a NovAtel OEM4 receiver are added. The latter is not expected to perform well under signal degradation as its firmware is optimized for high accuracy performance but is used to show the range of performance now possible under signal degradation. Two identical HS SiRF units were used. The receivers are shown in Figure 1. Table 1 outlines the major similarities and differences between the SiRF standard and high sensitivity (HS) receivers. The receivers use the same basic hardware architecture; however, the integrations times used are significantly different, as the high sensitivity receiver uses long periods of non-coherent integration. In addition, the Kalman filters used in the receivers are different. The filter differences become evident in the position domain analysis. All receivers used herein output raw measurements. These were recorded for subsequent analysis. Figure 1: Receivers Used During Testing Table 1: Comparison of Standard and High Sensitivity SiRF Receivers ANALYSIS CONVENTIONS The following colour convention was used when comparing the results from the different receivers. The OEM4 receiver results are plotted in red, the standard SiRF results in green and the HS SiRF receiver results in blue and purple, respectively. In the figures shown in this paper, the test warm-up period data is often shown in grey or in a faded colour scheme. This data is not included in the test statistics. Presented at ION National Technical Meeting, San Diego, January

3 Some figures show the testing period data only but from multiple test runs. This data is not contiguous in time but is presented in a continuous fashion to better display the similarities and differences between test runs. Black vertical lines delineate the different test runs. In figures displaying the calculated fix density values for the receivers, an overlaid light blue dot is used to indicate 2D fix versus 3D fix. GPS SIMULATION TESTING - TRACKING THRESHOLD TEST In recent years, advances in simulation technology have contributed to the development of state-of-the-art hardware GPS simulators. Spirent Communications Inc. makes the STR-4760 GPS simulator. The simulator is described thoroughly in GSS, The simulation unit, shown in Figure 2, used by the University of Calgary, consists of a control computer (shown on the left) and two 16 channel L1 or 8 channel L1/L2 hardware simulator boxes (shown on the right). GPS signals are often attenuated by propagation through different mediums (atmosphere, foliage, buildings, etc). This results in an effective decrease in the carrier strength component of C/N 0. In addition, broadband sources of RF interference also decrease C/N 0 by increasing the ambient noise density. The simulator allows real-time control of the signal level (± 20dB with respect to 160dBW) for each satellite corresponding to one channel of signal output. This power level is referred to as the relative channel power for the simulator as used in this paper. The signal level can be set equally and varied by the same amount for all channels. The C/N 0 threshold for tracking weak signals can be determined by lowering the signal level slowly until the receiver is no longer able to track the satellite. Figure 2: Spirent (GSS) STR-4760 Simulator A test was designed, based on MacGougan et al. (2001), using the simulator to assess the tracking threshold of the four test receivers. A constellation based on the real almanac for GPS week 1148 was used to produce simulated GPS signals. A 20 minutes warm-up period with undegraded signal tracking ensured that all receivers had enough time to lock on all eight simulated satellites. All the satellites used the same signal power, which was lowered from +10dB to -20dB relative channel power in 0.5 db steps. A five-minute period with the lowest power level possible (-20dB) finished off the test. The set-up for this test is shown in Figure 3. Figure 3: Tracking Threshold Test Set-up The tracking results, in terms of the average C/N 0, for all satellites tracked, and the associated simulator relative channel power are shown in Figure 4. The OEM4 receiver tracks to about 9dB relative channel power. The standard SiRF receiver does not fare much better as it tracks down to -10dB relative channel power. The tracking ability of the HS SiRF receivers is at or better than the -20dB relative channel power. The number of satellites tracked and the associated simulator relative channel power is shown in Figure 5. The NovAtel and standard SiRF receivers lose almost all satellites simultaneously while the high sensitivity receivers lose only 3 satellites gradually. The 3D error in position for each receiver and the associated simulator relative channel power is shown in Figure 6. Strong correlation between low signal levels and higher position error is evident. However, the HS receivers are still able to provide position output; the five satellites that they track provide a good geometry. Presented at ION National Technical Meeting, San Diego, January

4 Figure 4: Signal Power Levels During Simulator Tracking Threshold Test Figure 6: 3D Position Error During Simulator Tracking Threshold Test FIELD TESTS The aim of the field tests was to take the high sensitivity receivers into increasingly difficult environments in terms of GPS signal tracking. Thus, testing was first performed under signal masking conditions and fast fading due to roadside trees and foliage. Subsequent testing took place under more severe signal masking and multipath conditions in downtown Vancouver, B.C, Canada and Calgary, Alberta, Canada. Finally, testing inside two types of residential garages was performed to assess conditions with no line of sight on any of the satellites. Figure 5: Number of Satellites Tracked During Simulator Tracking Threshold Test FIELD TEST SET-UP AND DESCRIPTION The test set-up used for the base station is shown in Figure 7. A high performance antenna, namely the NovAtel 600 model, was mounted at a surveyed location with a clear view of the sky. The test set-up used for the rover receivers is depicted in Figure 8. The rover receivers were always initialized under open sky conditions for a 20-minute period. Test statistics only refer to the data after this warm-up period. The receivers are tested in parallel using a common antenna (NovAtel 600 model) or an inline low noise amplifier or LNA in the case of the simulation test. An LNA (also part of the antenna) acts to set the signal Presented at ION National Technical Meeting, San Diego, January

5 conditions (i.e. noise floor) to very similar values in each receiver (Van Dierendonck, 1995). In order to ensure valid comparison, the signal conditions experienced by each receiver must be very similar. In real conditions, this is achieved by splitting the signal from a common active antenna. In simulation mode, this is achieved with the use of an LNA prior to the signal splitter. vehicle, when sufficient satellite coverage was available to update the INS with an accuracy level to be described later. The antenna was mounted on the roof of the vehicle approximately one metre above the INS. In the case of the garage tests, the reference positions inside the garage were measured by extending the GPS position of a nearby point using a standard survey method. Figure 7: Field Test - Base Station Set-up Figure 8: Field Test - Rover Set-up The modes of field-testing include: Vehicular testing in a residential area of Calgary with foliage beside and overhanging much of the road Vehicular testing in downtown Calgary and Vancouver Static test inside a wooden frame garage and a concrete wall garage. The vehicular tests have speeds of up to 50 km/h. A carrier phase differential GPS/INS system was used to provide the reference data for the positions of the test FIELD TEST MEASURES The analysis of results of the field-testing focuses on the observation and position domains. Position domain analysis is performed by comparison of the test receiver s positions with positions derived from carrier phase differential GPS/INS using NovAtel's Black Diamond system when available. Comparison is only undertaken when the position errors estimated by the GPS/INS system are better than 5 m in latitude and longitude and 10 in height, at the 1 sigma level. This ensures that the ensuing derived statistics are meaningful. Analysis focuses on the receiver output position for the SiRF model receivers. The NovAtel OEM4 receiver is used to collect raw data for post mission analysis. This raw data is used in a least squares position solution using height constraints (when the geometric dilution of precision, GDOP, exceeds 5.0). This was done in order to compare the 2D fix capability of the SiRF receivers with that derived from a geodetic quality receiver. The SiRF receivers employ a Kalman filtering and the software to process the OEM4 data does not. The effects of the Kalman filter will be noticeable during comparison of kinematic testing periods with fewer instantaneous outliers evident. The focus of the analysis in the position domain is on horizontal and vertical errors. Vertical performance is expected to be poor due to degraded geometry and the typically poor performance of GPS in terms of height. Errors are computed by subtracting the reference values from the true values. Signal quality, fading, availability, and dilution of precision (DOP) are also be analyzed. Signal fading is calculated by the difference of the C/N 0 measurements between like model receivers at the rover and base stations in order to eliminate receiver C/N 0 estimation biases. HDOP is assessed based on the receiver output HDOP for the SiRF receivers and the computed value based on satellites tracked for the OEM4 receiver. The measurement output of a GPS receiver typically includes pseudorange, Doppler, and carrier phase observations. The L1 pseudorange is the primary measurement used in single point positioning and thus is the focus of the measurement domain analysis. By constraining the positions of the receiver to known reference positions, an assessment of the errors in the pseudoranges can be made from the residuals of the constrained least-squares solution. C 3 NAVG 2TM, a Presented at ION National Technical Meeting, San Diego, January

6 software package developed at the University of Calgary, is capable of this type of analysis. Fix density is also a useful measure of a receiver s capabilities in an environment with signal masking and multipath. Fix density values can be determined directly from the SiRF receiver output and from the processing of the OEM4 data. However, a more pessimistic test measure was used in the comparison of the receivers. The 2D fix density was calculated based on the number of satellites used in epoch-to-epoch solutions during the test period. If three satellites were used in a solution, a 2D fix was assumed. Similarly, a 3D fix was assumed when four or more satellites were present. Fix density is thus the percentage of epochs with valid fixes based on a 1 Hz data rate for the entire test period. This measure is pessimistic as it is possible to obtain a 2D fix with two satellites used in solution by using filtering, clock coasting and height fixing (e.g., Lachapelle et al.,1997).in summary the following measures are used: Observation Domain: Residuals: as estimated from a least squares solution constrained to known rover positions Fading: C/N 0 (base) C/N 0 (rover) using like type receivers HDOP: Horizontal Dilution of Precision Number of satellites used in solution Position Domain: 2D Fix Density: percentage of epochs with 3 satellites used in solution 3D Fix Density: percentage of epochs with 4 or more satellites used in solution 2D Error and height error. RESIDENTIAL TESTING UNDER TREES AND FOLIAGE A common problem for many GPS receivers is the fast fading and tracking problems induced by trees and roadside foliage. A vehicular data set was thus collected in a residential area of Calgary with foliage beside and overhanging much of the road. This environment is shown in Figure 9. One to three level houses with trees lining the roadside characterize this older area of Calgary known as Mount Royal. Three test runs were performed on September 27, 2001 beginning at approximately 16:10:00, 16:50:00, and 17:35:00 UTC time. Unfortunately, the standard model SiRF receiver was not available during these test runs and is not included in the analysis. There were also problems with the GPS/INS processing for test 1 so a differential code solution using OEM4 receivers was used during the limited epochs when the estimated accuracy was better than 5 m in latitude and longitude and 10 m in height. Figure 9: Vehicular Test in Residential Area The number of satellites used in the position solutions for the receivers for all three test-runs is shown in Figure 10. The associated fix densities for each receiver are shown in Figure 11. The results indicate that satellite availability is improved for the high sensitivity receivers, as they are capable of 3D fix during most of the test runs. The OEM4 reverts to 2D fix more frequently. Figure 10: Residential Test - Number of Satellites Used In Solutions Presented at ION National Technical Meeting, San Diego, January

7 The 2D errors for all three test runs and associated statistics are shown in Figure 12. The associated HDOP values are shown in Figure 13. The height errors during the test runs are shown in Figure 14. More geometry degradation is present due to fewer satellite observations available for the OEM4 receiver. In addition, more instantaneous outliers are present in the results from the OEM4 receiver due to the use of unfiltered epoch-byepoch processing. The use of a Kalman filter for periods with very poor geometry and a lack of quality observations would reduce these outliers significantly, as it has for the case of the SiRF receivers. Figure 13: Residential Test - HDOPs Figure 11: Residential Test - Fix Density Values Figure 14: Residential Test - Height Errors The run-to-run trajectories (based on the receivers output) for the two HS SiRF receivers are very similar and that for one receiver is shown in Figure 15. The NovAtel OEM4 receiver run-to-run trajectory is shown in Figure 16. The instantaneous outliers in the OEM4 derived positions are also evident in this figure. Figure 12: Residential Test - 2D Position Errors Presented at ION National Technical Meeting, San Diego, January

8 Figure 17: Residential Test Correlation Between Fading and Pseudorange Residuals for PRN 11 Figure 15: Residential Test - HS66 Run-To-Run Horizontal Positions Figure 16: Residential Test - OEM4 Run-To-Run Horizontal Positions In terms of signal fading and residuals, a low elevation satellite, rising from 15 to 24, was investigated during run 2. Figure 17 shows the absolute value of the residuals for PRN 11 along with the corresponding calculated fading values for one of the HS receivers. Not only are fast fading effects observable but also a strong correlation exists between the fading and the residual values. URBAN CANYON TESTING The challenges for satellite-based navigation posed by the downtown environment of many cities are numerous. Large buildings tend to mask much of the sky and induce large multipath and fading errors. The general problem with using GPS in urban canyons translates into a lack of satellite availability. However, high sensitivity GPS receivers may be able to track the degraded signals of satellites that are typically masked. This may be by tracking a multipath signal or the combination of the weakened direct signal and multipath signals. The addition of these degraded observations could allow position determination when otherwise impossible by standard means. In order to address the capabilities of the HS SiRF receivers in urban canyons, tests were performed in downtown Calgary and Vancouver. Both cities have concentrations of tall buildings. DOWNTOWN VANCOUVER Testing took place on October 13 th and 14 th, 2001, with five test runs and two test runs, respectively. The test trajectory used for all runs is shown in Figure 18 along with some pictures of the testing environment. Each test lasted approximately 30 minutes, following a twentyminute warm-up period under clear sky conditions. Presented at ION National Technical Meeting, San Diego, January

9 Figure 18: Vancouver Test Trajectory The number of satellites used in the position solutions for the receivers for all seven test-runs are shown in Figure 19. The associated fix densities for each receiver are shown in Figure 20. The results indicate that the availability is again improved for the HS receivers, as they are capable of 3D fix for 91.9% and 93.6% of the test runs. The standard SiRF receiver and the OEM4 unit both revert to 2D fix more often with 11.2% and 15.2% 2D fixes respectively versus 1.2% and 1.0% 2D fixes for the high sensitivity receivers. The 2D errors for all seven test runs and associated statistics are shown in Figure 21. Major outliers are present for all four receivers. In order to better understand the performance of the receivers, outliers larger than three standard deviations (3 ) were removed and the statistics were recomputed. These results are shown in Figure 22. The use of degraded measurements combined with HS filtering effects leads to large errors and poor performance in this environment. The associated HDOP values are shown in Figure 23. The improvement in HDOP for the HS receivers is again evident; although, this is not clearly reflected in the 2D position error statistics. In terms of vertical position errors, the performance of the receivers is shown in Figure 24 along with corresponding statistics. These results emphasize the need for proper filtering of the observations in the position solution. Observations are useful only when weighted appropriately. Figure 19: Vancouver Test - Number of Satellites Tracked Figure 20: Vancouver Test - Fix Density Values Presented at ION National Technical Meeting, San Diego, January

10 Figure 21: Vancouver Test - 2D Position Errors Figure 23: Vancouver Tests - HDOPs Figure 22: Vancouver Test - 2D Position Errors With Outliers Removed Figure 24: Vancouver Test - Height Errors DOWNTOWN CALGARY Three successive tests were conducted in downtown Calgary on September 27 th. No reference position system was used to provide a truth trajectory. However the trajectory used is along straight East-West and North- South streets and good accuracy performance can easily be derived from plotting the test positions on an existing map. Also, fix density information and the number of satellites used in the solutions still provides useful information in term of solution availability. These results are shown in Figure 25 and Figure 26, respectively, Presented at ION National Technical Meeting, San Diego, January

11 followed by Figure 27 depicting the HDOP during the test runs. The results indicate that availability is again improved for the HS SiRF receivers. The standard model SiRF receiver and the OEM4 unit both revert to 2D fix more often. Figure 28, Figure 29, Figure 30, and Figure 31 show the run-to-run trajectories for the HS receivers, the standard model receiver, and the OEM4 receiver, respectively. The estimated HS receiver trajectory degradation is likely due to internal filtering problems. Figure 27: Calgary Test - HDOPs Figure 25: Calgary Test - Fix Density Values Figure 28: Calgary Test HS66 SiRF Unit Run-To-Run Horizontal Positions Figure 26: Calgary Test - Number of Satellites Tracked Figure 29: Calgary Test - HS14 SiRF Unit Run-To- Run Horizontal Positions Presented at ION National Technical Meeting, San Diego, January

12 Figure 32: Wood Frame and Concrete Wall Garage Test Environments Figure 30: Calgary Test - ST30 SiRF Unit Run-To- Run Horizontal Positions Figure 31: Calgary Test - OEM4 Unit Run-To-Run Horizontal Positions INDOOR TESTING By taking a GPS receiver inside, the direct line-of-sight component of the signals relied upon for effective navigation is compromised. Attenuation of the direct signal propagating through various types of materials is expected as well as increased error due to multipath. Testing was performed in two different types of residential garages as shown in Figure 32. A wood frame garage and then a garage with concrete walls located under a living room were tested with and without the garage door closed. For the purposes of this paper, the worst-case results corresponding to the concrete structured garage with the door closed are presented. The test in the concrete garage with the door closed began with a 20-minute warm-up period followed by a testduration of 60 minutes. The antenna was moved from a surveyed point outside the garage to a surveyed point inside the garage and the door was closed. The warm-up period for the following figures will be shown using faded colours. The number of satellites tracked by the receivers is shown in Figure 33. The associated fix density values are then shown in Figure 34. The results clearly indicate the failure of the standard GPS receivers to track signals inside while the HS receivers maintain usable observations of five to eight satellites during the test. In addition, the fix density values indicate 3D availability for 99 percent of the test period. The 2D errors for the receivers tested are shown in Figure 35 along with associated statistics. The HDOP values for HS14 are shown in Figure 36 along with the number of satellites used in the solutions. There is no significant degradation of HDOP for the HS receivers. For a more intuitive representation of the horizontal errors a scatter plot of the horizontal position is shown in Figure 37. Presented at ION National Technical Meeting, San Diego, January

13 Figure 35: Concrete Garage Test - 2D Errors Figure 33: Concrete Garage Test - Number of Satellites Tracked Figure 36: Concrete Garage Test - HDOPs Figure 34: Concrete Garage Test - Fix Density Values Figure 37: Concrete Garage Test - HS14 Horizontal Positions Presented at ION National Technical Meeting, San Diego, January

14 CONCLUSIONS The use of high sensitivity GPS receivers in unaided stand-alone mode results in higher availability of observations in residential areas, urban canyons, and some indoor environments. 3D Position fixes were obtained more frequently with the HS receivers than the standard receivers tested under foliage and in urban canyons. In the indoor environment tested, the standard receivers could not operate while the high sensitivity unit could still provide positions with accuracy better than 50m. The tracking threshold of the high sensitivity receivers was tested using a GPS hardware simulator and found to be at least 9 to 10dB lower than the standard mode GPS receivers tested. The ability to provide pseudorange measurements and positions, when otherwise impossible using standard tracking, has clear advantages for users in terms of availability. In general, better DOP results from more usable observations. However, position degradation due to increased noise and multipath on the measured pseudoranges results from the use of degraded observations. More work remains to be done to improve position reliability. In a vehicular case, the use of additional sensors, such a low cost rate gyro, will more than likely improve availability and reliability (e.g. Stephen & Lachapelle 2001). Indoor, the use of MEMS accelerometers and miniature direction finding sensors is likely to be necessary to improve these characteristics. The tests conducted herein dealt only with satellite reacquisition in degraded environments. Acquisition in such an environment is more difficult and requires specific tests. Stephen, J., and G. Lachapelle (2001). Development and Testing of a GPS-Augmented Multi-Sensor Vehicle Navigation System. The Journal of Navigation, Royal Institute of Navigation, 54, MacGougan, G., G. Lachapelle, M. E. Cannon, J. Gee, and M. Vinnins (2001) GPS Signal Degradation Analysis Using a Simulator. Proceedings of the Institute of Navigation ION Annual Meeting-2001 (June 10-13, 2001, Albuquerque, New Mexico). Moeglein, M. and N. Krasner (1998) An Introduction to SnapTrack Server-Aided GPS Technology. Proceedings of the Institute of Navigation ION GPS-98 (September 15-18, 1998, Nashville, Tennessee), Peterson, B., D. Bruckner, and S. Heye (1997) Measureing GPS Signals Indoors. Proceedings of the Institute of Navigation ION GPS-97 (September 16-19, 1997, Kansas City, Missouri), van Diggelen, F. and C. Abraham (2001) Indoor GPS, The No-Chip Challange. GPS World, 12(9), Van Dierendonck, A. J. (1995), GPS Receivers, Global Positioning Systems: Theory and Applications, Vol I, ed. B.W. Parkinson and J.J. Spilker Jr. (1996), American Institute of Aeronautics and Astronautics, Washington DC, pp Presented by G. MacGougan at ION National Technical Meeting, San Diego, January REFERENCES Chansarkar, M. and L. J. Garin (2000) Acquisition of GPS Signals at Very Low Signal to Noise Ratio. Proceedings of the Institute of Navigation ION National Technical Meeting-2000 (January 26-28, 2000, Anaheim, California), Garin, L. J., M. Chansarkar, S. Miocinovic, C. Norman, and D. Hilgenberg (1999) Wireless Assisted GPS-SiRF Architecture and Field Test Results. Proceedings of the Institute of Navigation ION GPS-99 (September 14-17, 1999, Nashville, Tennessee), GSS, (2000), STR Series Multichannel Satellite Navigation Simulator Reference Manual, Global Simulation Systems, doc. # DGP00032AAC, Issue 10-00, April Lachapelle, G., S. Ryan, M. Petovello, and J. Stephen (1997) Augmentation of GPS/GLONASS For Vehicular Navigation Under Signal Masking. Proceedings of the Institute of Navigation ION GPS-97, Presented at ION National Technical Meeting, San Diego, January