| GISdevelopment.net ---> Asian GPS --->Proceedings --->2002 --->GPS in Atmospheric Sciences |
Scintillation effect on WAAS Reference Station (WRS) signal
Arjun Singh
Airports Authority of India
Telephone No. +91-11-4649778 (R ),+91-11-4655718 (o)
Email: arjun_gps@yahoo.com
Shamsher Singh
Airports Authority of India
Telephone No. +91-11-7553748 ( R), +91-11-4641018 (o)
Email : shamsher_rana@rediff.com
Dr D.C Reddy
Vice Chancellor (retired)Osmania university, Hyderabad India
Telephone No: + 91-40-7552778 (R ),+ 91-40-7098066 (o )
Email: p_laxminarayana@yahoo.com
Dr P. Laxminaryana
Navigation Electronic Research and Training Unit, Osmania University
Telephone No.+91-40-3544765 (R )
Email: dc_reddy2@yahoo.com
Arjun Singh
Airports Authority of India
Telephone No. +91-11-4649778 (R ),+91-11-4655718 (o)
Email: arjun_gps@yahoo.com
Shamsher Singh
Airports Authority of India
Telephone No. +91-11-7553748 ( R), +91-11-4641018 (o)
Email : shamsher_rana@rediff.com
Dr D.C Reddy
Vice Chancellor (retired)Osmania university, Hyderabad India
Telephone No: + 91-40-7552778 (R ),+ 91-40-7098066 (o )
Email: p_laxminarayana@yahoo.com
Dr P. Laxminaryana
Navigation Electronic Research and Training Unit, Osmania University
Telephone No.+91-40-3544765 (R )
Email: dc_reddy2@yahoo.com
Abstract
For reliable operation of WRS which is part of the SBAS (satellite Based Augmentation System) , GPS satellite signal observation should be available continuously at each WRS. Loss of signal availability, leading to degradation of the SBAS performance can occurs during periods of Ionospheric scintillation. Scintillation effects arise from small-scale irregularities in electron density, which is most common observed in the high latitude auroral region and the low latitude equatorial anomaly region. The auroral region expands equator ward during intense geomagnetic storms and scintillation effects may be observed in southern united state and Europe. The equatorial scintillation is expected to peak during the years of solar maximum, while high latitude storms activity is expected to peak in the year 2001-2003. SBAS are currently being implemented in United States Wide Area Augmentation System (WAAS), European Geo Satellite Navigation Overlay System ( EGNOS) , Japan, Multifunction Transportation Satellite Augmentation System ( MSAS) and India GPS / GLONASS and Geo-Satellite Augmented Navigation (GAGAN) regions which may be affected , to some extent, by the presence of scintillation. The impact of the scintillation activity is quantified by loss of phase and code observation, number of satellite observations lost simultaneously and duration and spatial extent of degraded receiver tracking performance.
The GAGAN will become fully operational in the year 2010 for providing Ionospheric correction to the GPS signal users flying over Indian Airspace. The GPS data is taken in fair weather condition at Delhi but GPS signal has experienced scintillation in both days for certain periods. It was also observed in the position error vs number of sample plot that the errors (Longitude, Latitude and Altitude) fluctuating during 2030 to 0030 time (IST) The Indian continent has diverse climate and it is also near to magnetic equator and also the Ionosphere variation over the Indian Airspace is not uniform. Therefore the scintillation intensity will vary at every two degrees latitude. For GAGAN project, it is suggested that more study in different WRS location is required before starting the modeling of the Ionosphere over Indian Airspace.
Introduction
The ionosphere is a diverse medium, in which radio frequency signals are refracted by an amount dependent upon signal frequency and Ionospheric electron density. In region of small-scale irregularities in electron density, rapid random phase variation can be produced by phase irregularities in the emerging wave front. Diffraction of the signal also leads to variations in signal amplitude-refers to as amplitude scintillation (or amplitude fading, for degradation in signal strengths). The WRS signal is also affected by the atmospheric condition i.e. medium of propagation, particularly the ionosphere. Trans-Ionospheric condition for GPS signal are perturbed due to group delay of the signal and fluctuations in the GPS signal characteristics (amplitude and phase) caused by irregularities in the electron density distribution of the ionosphere which causes the range error measurement.
To classify the Ionospheric conditions, the entire globe has been divided into three distinct regions i.e. equatorial, mid-latitude and high latitude zones. The equatorial region extends from the magnetic equator to about ±30-degree. The high latitude region covers locations above 65-degree around the magnetic poles. In between the above two boundaries, lies the mid-latitude zone. The Indian subcontinent extends from the magnetic equator, touching Trivandrum near the tip of the peninsula, to the mid-latitude zone in the north.
The equatorial region has two very prominent features: (1) the equatorial anomaly also called the Appleton Anomaly which is the latitudinal variation of high ambient F-region electron density at the magnetic equator and two crests around ±30 degree. During afternoon and early evening hours and (2) very severe irregularity structures in the electron density distribution. These irregularity structures also known as ‘bubbles’ are actually depleted ionization regions in the form of bananas or peeled orange sections over the magnetic equator and have vertical and spatial extents, extending from a few to hundreds of kilometers. The irregularities cause severe fluctuations known as scintillations in the signal strength. This phenomenon becomes a major issue for navigation applications because fluctuations provide additional stresses to the GPS / GLONASS receiver tracking loops and can induce cycle slips or even complete loss of lock.
Background
Post– sunset in the equatorial ionosphere routinely cause rapid phase and amplitude fluctuations of radio waves propagating through the disturbed regions. The intensity of scintillation is positively correlated with the solar cycle and associated signal fades will often exceeds 20dB at L-band frequencies during solar maximum. The effects of such an environment on the performance on GPS/ GLONASS navigation systems is poorly understood [6]. Ionospheric scintillations in the equatorial region have been studies extensively over the last two decades on VHF,UHF L-band etc. In Fig. 1 three zones have been identified i.e. most intense, intense and moderately intense scintillations in India. The Indian subcontinent covers the geographic latitude range of 10 to 45 degree N and the magnetic equator touches the Southern tip of the Peninsula. The well-known equatorial anomaly occurs with the northern crest along the line joining Kolkata in the east and Ahmedabad in the west. In other works, this region covers the northern half of the equatorial belt in the Asian Zone and extends into the transitional latitudes from the equatorial to middle latitudes. Ionospheric observations in Indian subcontinent have been extensive and the contributions made from India in terms of the global aspects of equatorial scintillation are significant. However, most of the observations have been oriented towards understanding the physical plasma processes [2]. No serious attempt has, however, been made to derive models of Ionospheric scintillation, which can be used for aviation application in SBAS. It will be very critical while implementing the GAGAN for Indian Airspace.
Fig-1 Identify the different zone where the scintillations most intense, intense and moderate
Characteristics of scintillation
It has been well-established fact that the signal strength and frequency of scintillation is highly correlated with 11 years cycle of sunspot number. The percentage occurrence of scintillation is also a function of the local time of a day, season, and magnetic activity. Equatorial scintillation has been found to increase significantly with an increase in the solar activity level,. The control is more pronounced at all-equatorial locations around the anomaly crest than at the equator [1]. It has also been observed that there is the dramatic increase in the scintillation activity during the equinoxes (February to April and August to November) and December solstice months with the increase of sunspot number. It has also been observed that scintillation is very prominent during post sunset hour of these periods. This related to the equatorial Ionospheric plasma processes. It is also interesting to observe that scintillation occurrence in the June solstice is spares and almost independent of solar activity.
It has been established that the equatorial irregularities can frequently produce signal fading of L- band transmissions from Geo Earth Orbiting Satellite (GEOS) in excess of 20 to 25dB at locations like Ascension Island (7.9 degree S, 14.4 degree W geographic 31 degree S dip), Kolkata 22.58 degree N, 88.38 degree E geographic, 32 degree N dip) and Hong Kong (22.12 degree N, 114.12 degree E geographic, 30 degree N dip) situated near the crest of the equatorial anomaly [2].
Equatorial scintillation was observed too much stronger (i.e. with greater impact on receiver tracking performance) than the high latitude scintillation. Percentage of corrupt observations often exceeds 40% for station located near the equatorial anomaly peak, in the local time sector 2000-2300. Receiver tracking performance in the equatorial region exhibits clear seasonal variations with peaks in the winter month and a dependence on the solar cycle. Based on observed static and a solar maximum in mid 2000 it is anticipated that the high levels of degraded tracking performance will continue in the equatorial region until early 2002[3].
In terms of WAAS, the scintillation will have significant effects on the WRS network i.e. if VSAT or terrestrial network is used. The impact of the scintillation will be interpreted with respect to Grid Ionosphere Vertical Error (GIVE) bounds, as derived for each grid Ionospheric estimates that will be transmitted to the users. The GIVE values reflect accuracy and reliability derived from networks of Ionospheric observations in the local area surrounding the grid points. Only the current observations will be used in the derivations of the GIVE values, such that an error bound reflects current statistics. If no observations are available within the local area, the GIVE will be marked unavailable for a given grid points.
Scintillation data base
The amplitude and phase scintillation measurements can be done with an Ionospheric Scintillation Monitor (ISM), a GPS receiver specifically designed to track the scintillation behavior of the atmosphere. The ISM receiver is a modified high performance single frequency GPS C/A code receiver equipped with wide bandwidth tracking loop. The wide bandwidth tracking loops improve the receiver ability to maintain signal lock during active periods. The receiver modifications include the use of a low noise, highly stable, oven controlled crystal oscillator as the receiver’s frequency reference. The receiver is also modified for high data rate processing of signal power and carrier phase measurements for calculation of scintillation –parameters in near real time.
The ISM calculates several parameters useful for scintillation analysis. The S4 parameter, measuring amplitude scintillation, is normalized standard deviation of the incoming signal intensity given by:
S4 =√ [2 > - 2 ] / 2
Where “<> “ denotes the average value of detrended signal intensity over a 60 HZ period. The signal intensities are measured at 50 Hz and detrended with 6th order butterfly low pass filter. The measure of phase scintillation is based on the standard deviation of carrier phase signals that have been detrended with 6th order butterfly high pass filter. The standard deviation computed over 1,3,10,30 and 60 second interval. The ISM reports will be for all scintillation calculation at 60 seconds intervals near real time [5].
Intensity of scintillation is measured by different indices. The most commonly used index for scientific studies is S4, which is the second moment of the signal fluctuation around the mean signal level. The communication Engineers however have found SI (dB) as a simpler measure for describing scintillation. SI (dB) is obtained by scaling the difference in dB levels between the third peaks from the maximum to the third peak level from minimum of the signal. Empirical relation between the two measurements is available [7]. In this presentation the intensity of scintillation will be represented by SI (dB). Three case of scintillation corresponding to low value of S4 (<0.3), moderates value (0.3 to 0.6) and high values (>0.6) has been considered [4].
A comparison of scintillation effects on GPS / GLONASS signal with GEOS signal was observed at one station, which shows the considerable difference in scintillation occurrence pattern. Incase of GEOS, scintillation occurs when the clouds of Ionospheric irregularities drift across sub Ionospheric point of the satellites. At Kolkata, scintillation on 1.5 GHz channels of INMARSAT normally occurs in between local sunset and midnight. L-band scintillations normally moderate to intense during high solar activity years. VHF scintillation is usually saturated and very fast in the initial stage. In the post midnight hours the rate of fading is appreciably reduced although the signal exhibits saturated scintillation before gradually restoring to its steady level [2].
In case of GPS/ GLONASS constellation, 10 to 15 satellites together distributed over the sky different sections of the ionosphere around the station under varying look angles. In the post sunset hours, a number of GPS and GLONASS satellites signal often experience scintillation and a result scintillations appear to be more extensive for such system. Even in the post mid night hours, when no scintillation activity is observed GEOS microwave links, the L band GPS signals experience severe fluctuations on signal strength (C/ N0). The deep fading of GPS signal strength (C/N0 ) is interrupted in terms of propagation angle of the particular satellite ray path with respect to geomagnetic field direction. From low latitude station in the northern hemisphere, the orbiting satellite signals will be viewed “end on” through a field tube aligned equatorial plasma bubble introducing more intense C/ N0 fluctuation than in the case of the Geo Satellite which normally passes through a small section of the bubble.
The intense fluctuation of C/ N0 of the GPS signals when there were negligible or mild scintillation on the L band signal of a GEOS signal was initially puzzling. This has prompted to check the calibration of the GPS C/ N0 fluctuation with respect to that of the GEOS signal receiving system. The signal from a GPS satellites exhibiting moderate to intense but not saturated scintillations, traveling near the INMARSAT sub Ionospheric point was chosen for calibration. It has been observed there is a good correspondence between scintillation indexes of the two satellites when two sub Ionospheric points are close to each other [2].
Occurrence of scintillation
The scintillation occurs only when its strength exceeds the SI value of 3dB during 2000 to 2400hrs (IST) has been considered.. Scintillation increases both in frequency and amplitude (dB) with solar activity. Year 1999 to2000 was the peak period of the solar cycle and strong scintillation was experienced each day during this period. It is interesting to note that during February to April and August to November the scintillation effect was quite frequent during solar cycle. With increase in solar activity GPS + GLONASS signals are likely to be affected by scintillation in the equatorial region. Location near the equatorial anomaly crest would show the degradation of the navigation in a peculiar way. In the Northern Hemisphere scintillation may simultaneously render more than on GPS/GLONASS satellites in the southern sky unusable for navigation purpose. In addition to southern links any GPS/GLONASS satellites near overhead will also show intense scintillations due to higher ambient level of ionization around the crest of the equatorial anomaly. Fig- 2 shows the variation of scintillation index vs Time of IMMARSAT & GPS + GLONASS satellites for four hours at Kolkata. As a results the accuracy of position fixing may degrade to unacceptable limits
Fig-2 Variation of scintillation index of IMMARSAT & GPS+ GLONASS
Experimental data
The Asia pacific Air Navigation planning implementation Regional Group (APANPIRG) of International Civil Aviation Organization (ICAO) in its meeting held in October 2000, had decided to initiate a Regional GPS measurement campaign and Civil Aviation Administration of Singapore was assigned the task of coordination activity in the region. Accordingly, Civil Aviation Administration of Singapore had invited the Civil Aviation Administration of states in the region to participate in the GPS measurement campaign to determine “Normal and Peak Excursion on the GPS performance”. The measurement campaign carried out on 25-09-01 and 09-10-01 with each measurement lasting 24 hours period i.e. 00:00:00: to 23:59:59. The ASTECH (GPS+ GLONASS) 12-channel receiver was used for raw data down load and GPS receiver antenna was positioned at Latitude 28° 38. 22149’ N, Longitude, 77° 10.29609’ E, Altitude 192.37467 Meters. The ASTECH (GPS+ GLONASS) receiver is proved to be very robust at tracking the carrier signal amplitude and phase, but it experienced scintillation induced navigation outages on both days in nighttime.
Fig –3(a): Error vs time (∆Latitude, ∆Longitude and ∆Altitude)
The data sample was taken from 15:10:00 to 17:08:58(UTC) and 10 GPS satellites were visible during the sample period. The position error was calculated as coordinate of the reference point i.e. location of the GPS receiver antenna, is known. The position value was down loaded in the form of Latitude, Longitude and Altitude at the rate of one Hz . One sample of data is approximately for two hours. The position error is calculated with respect to the reference point. The DOP value is less than four during sample time. Therefore the GPS receiver has calculated the position with best geometry of the satellites, which are visible during sample time. The errors are ∆Latitude, ∆Longitude and ∆Altitude & calculated in terms of meters.
Fig –3(b): All visible GPS satellites SNR Vs time
Fig -3(c ): SNR Vs Time for SV-4
The position error vs time (UTC) are plotted for the all-visible satellite during sample period with good dilution of precision (DOP). It is observed in Fig 3-a that the position error plot contains the spikes in the measurement during entire sample period. The spikes over plots are due to scintillation. In further analysis it was observed that GPS receiver was tracking 10 satellites but 8 satellites were locked for position data measurement with good geometry. The two satellites could not lock due to scintillation or low elevation angle. In detail analysis, GPS receiver had also recoded the SNR of the all-visible GPS satellites. Fig 3-b shows the plot between SNR of all visible satellites vs Time (UTC) and it is observed some of the satellites SNR are fluctuating badly.
In detail analysis, it was found that GPS Satellites Vehicle (SV) 28,27,26,2,7, 8,11,4 and locked and their signal were affected by Scintillation during the sample period. The SV 28,27,26,7 signals was less affected, SV 11 was moderate and SV 8 , 4 were severely affected by scintillation. In Fig 3-c the severely effected satellites signal is plotted SNR vs Time (UTC). It is observed that GPS SV–4 signal was badly affected and some of the time satellite is unlocked which has definitely degrade the measurement performance.
The second sample of the data from 17:10:00 to 19:08:58 (UTC) was taken immediately after first sample of data logging was completed. During the sample period, the total number of satellites tracked and locked is 7. The similar exercise was repeated. Fig4-a is plot between position error vs Time (UTC) and it was observed that mid of the sample the spikes are prominently visible but not in continuous in nature.
Fig –4(a): Error vs time(∆Latitude, ∆Longitude and ∆Altitude)
Fig -4(b ): All visible GPS satellites SNR Vs time
Fig 4-b is plotted between all visible satellites SNR vs Time (UTC) and observed that mid of the plot is fluctuating during same period of sample. To further strengthen the observation individual GPS satellites SNR vs time (UTC) is plotted as Fig4-c to know about which satellite is badly affected by scintillation. The SV 28,24,20,9,7,4,2 were tacked and locked by receiver. SV 24 was severely affected. SV 20, 7 and 9 were less affected by scintillation.
Fig -4(c ): SNR vs time
The GPS data logging was carried out on 08-10-2001 for 24 Hours and raw data was recorded from 14:20:01 to 16:19:59 (UTC). The eight GPS satellites were in view but 7 satellites SV28, 11,27,26,8,7 and 4 were locked and their data was used for position calculation. The similar exercise was carried out which has been done in the previous section. Fig 5-a is plotted between position errors and Time (UTC). It observed that mid of the data, spikes are visible otherwise entire observations are free from spikes. The spikes are due to scintillation effects on GPS signal. To further strengthen the observation Fig 5-b is plotted between all GPS satellites in view vs Time (UTC). Similar spikes are observed in the mid of the data sample. In further analysis, to check which GPS satellites signal is badly affected by scintillation . In Fig 5-c the SNR vs Time (UTC), the SNR of GPS SV –2 is cutting off in the mid of the observation i.e. due to scintillation the GPS SV-2 is unlocking. Therefore the spikes are in the mid of the observation due to GPS SV-2.
Fig -5(a ): Error vs time(∆Latitude, ∆Longitude and ∆Altitude)
Fig -5(b) All visible GPS satellite
Fig –5(c): SNR vs time
The second sample of data for two hours was taken just after the first sample data recording was completed. The data was recorded from 16:21to 18:44:58(UTC). Fig 6-a is plotted between position errors vs time (UTC) and end of the data sample small spikes are visible otherwise entire data is free from the scintillation. The sample data for two hours has experienced least scintillation effect on 08-10-2001. The other parameters e.g. DOP, Errors and SNR were well within the limits. Fig 6-b is plotted between all GPS satellites in view vs Time (UTC) , shows least variation in SNR . Fig 6-c is plotted between SNR vs Time for most affected satellites and SV-20 shows that GPS receiver is unable to lock. It may be due to low elevation angle of the SV-20 in view or may be just appearing in the sky.
Fig -6(a ): Error vs time(∆Latitude, ∆Longitude and ∆Altitude)
Fig -6(b) All visible GPS satellites Vs Time
Fig –6(c): SNR vs time
Discussion
In the GPS data measurement campaign the data analysis shows that few of the GPS satellites were severely affected by scintillation but again affected GPS satellites has to be identified There is possibility that some or all of the affected satellites may provide good geometry and DOPs which will impact on position measurement accuracy as well as in the error calculation at WRS. The instantaneous position solution obtained during strong scintillation is compared with those of just before the onset of scintillation and after the disappearance of the scintillations. The variation of latitude and longitude of the present location determine by GG24 receiver shows slow variation with time even when links are undisturbed. When scintillation is observed on GPS signals the values of latitude and longitude shows large and fast deviation superimposed on the slow variations. In position measurement time if all lock GPS satellites signal are affected by scintillation may introduce large deviation in position measurement but if few are affected because of their location in sky may not show much deviation in the measurement.
Conclusion
The data collected here underscore the importance of GPS application during solar cycle. The varied behavior of GPS signal suggest that end to end test to be conducted for critical GPS application to insure that any potential degrading effects, such as those caused by scintillation are identified and accurately accessed. Understanding these effects remains an important goal and presents the formidable challenge for development of accurate model for region.
It is observed that in the month of September 2001 the scintillation effect was more prominent than October 2001.The data recorded at Delhi which falls in the region of the moderate scintillation which has been shown in the Fig-1. If this the case then other WRS which falls in the intense and most intense zone need more observation for modeling the Ionosphere for Indian Airspace. The seasonal change may also cause the GPS signal fading needs more study.
The GAGAN, will provide Ionospheric corrections to GPS signals for aircraft in the Indian Airspace for both en-route navigation and precision landing The effects of the scintillation is important for GAGAN, since WRS located all over the India will be under influence of equatorial scintillation. Finally with expansion of the GPS navigation on the global basis, it is too important to access the potential effects of scintillation navigation performance and availability in regions where scintillation is most active.
Acknowledgement
The author would like to thanks Dr. P. Banerjee , Deputy Director, Time and Frequency division of National Physical Laboratory for his kind support in the GPS data logging activity and valuable suggestion in the analysis of the data. Thanks to Shri S.H. Khan Executive Director (CNS-P) Airports Authority of India for extending full support and encouraged my team during data logging period.
References
For reliable operation of WRS which is part of the SBAS (satellite Based Augmentation System) , GPS satellite signal observation should be available continuously at each WRS. Loss of signal availability, leading to degradation of the SBAS performance can occurs during periods of Ionospheric scintillation. Scintillation effects arise from small-scale irregularities in electron density, which is most common observed in the high latitude auroral region and the low latitude equatorial anomaly region. The auroral region expands equator ward during intense geomagnetic storms and scintillation effects may be observed in southern united state and Europe. The equatorial scintillation is expected to peak during the years of solar maximum, while high latitude storms activity is expected to peak in the year 2001-2003. SBAS are currently being implemented in United States Wide Area Augmentation System (WAAS), European Geo Satellite Navigation Overlay System ( EGNOS) , Japan, Multifunction Transportation Satellite Augmentation System ( MSAS) and India GPS / GLONASS and Geo-Satellite Augmented Navigation (GAGAN) regions which may be affected , to some extent, by the presence of scintillation. The impact of the scintillation activity is quantified by loss of phase and code observation, number of satellite observations lost simultaneously and duration and spatial extent of degraded receiver tracking performance.
The GAGAN will become fully operational in the year 2010 for providing Ionospheric correction to the GPS signal users flying over Indian Airspace. The GPS data is taken in fair weather condition at Delhi but GPS signal has experienced scintillation in both days for certain periods. It was also observed in the position error vs number of sample plot that the errors (Longitude, Latitude and Altitude) fluctuating during 2030 to 0030 time (IST) The Indian continent has diverse climate and it is also near to magnetic equator and also the Ionosphere variation over the Indian Airspace is not uniform. Therefore the scintillation intensity will vary at every two degrees latitude. For GAGAN project, it is suggested that more study in different WRS location is required before starting the modeling of the Ionosphere over Indian Airspace.
Introduction
The ionosphere is a diverse medium, in which radio frequency signals are refracted by an amount dependent upon signal frequency and Ionospheric electron density. In region of small-scale irregularities in electron density, rapid random phase variation can be produced by phase irregularities in the emerging wave front. Diffraction of the signal also leads to variations in signal amplitude-refers to as amplitude scintillation (or amplitude fading, for degradation in signal strengths). The WRS signal is also affected by the atmospheric condition i.e. medium of propagation, particularly the ionosphere. Trans-Ionospheric condition for GPS signal are perturbed due to group delay of the signal and fluctuations in the GPS signal characteristics (amplitude and phase) caused by irregularities in the electron density distribution of the ionosphere which causes the range error measurement.
To classify the Ionospheric conditions, the entire globe has been divided into three distinct regions i.e. equatorial, mid-latitude and high latitude zones. The equatorial region extends from the magnetic equator to about ±30-degree. The high latitude region covers locations above 65-degree around the magnetic poles. In between the above two boundaries, lies the mid-latitude zone. The Indian subcontinent extends from the magnetic equator, touching Trivandrum near the tip of the peninsula, to the mid-latitude zone in the north.
The equatorial region has two very prominent features: (1) the equatorial anomaly also called the Appleton Anomaly which is the latitudinal variation of high ambient F-region electron density at the magnetic equator and two crests around ±30 degree. During afternoon and early evening hours and (2) very severe irregularity structures in the electron density distribution. These irregularity structures also known as ‘bubbles’ are actually depleted ionization regions in the form of bananas or peeled orange sections over the magnetic equator and have vertical and spatial extents, extending from a few to hundreds of kilometers. The irregularities cause severe fluctuations known as scintillations in the signal strength. This phenomenon becomes a major issue for navigation applications because fluctuations provide additional stresses to the GPS / GLONASS receiver tracking loops and can induce cycle slips or even complete loss of lock.
Background
Post– sunset in the equatorial ionosphere routinely cause rapid phase and amplitude fluctuations of radio waves propagating through the disturbed regions. The intensity of scintillation is positively correlated with the solar cycle and associated signal fades will often exceeds 20dB at L-band frequencies during solar maximum. The effects of such an environment on the performance on GPS/ GLONASS navigation systems is poorly understood [6]. Ionospheric scintillations in the equatorial region have been studies extensively over the last two decades on VHF,UHF L-band etc. In Fig. 1 three zones have been identified i.e. most intense, intense and moderately intense scintillations in India. The Indian subcontinent covers the geographic latitude range of 10 to 45 degree N and the magnetic equator touches the Southern tip of the Peninsula. The well-known equatorial anomaly occurs with the northern crest along the line joining Kolkata in the east and Ahmedabad in the west. In other works, this region covers the northern half of the equatorial belt in the Asian Zone and extends into the transitional latitudes from the equatorial to middle latitudes. Ionospheric observations in Indian subcontinent have been extensive and the contributions made from India in terms of the global aspects of equatorial scintillation are significant. However, most of the observations have been oriented towards understanding the physical plasma processes [2]. No serious attempt has, however, been made to derive models of Ionospheric scintillation, which can be used for aviation application in SBAS. It will be very critical while implementing the GAGAN for Indian Airspace.
Fig-1 Identify the different zone where the scintillations most intense, intense and moderate
Characteristics of scintillation
It has been well-established fact that the signal strength and frequency of scintillation is highly correlated with 11 years cycle of sunspot number. The percentage occurrence of scintillation is also a function of the local time of a day, season, and magnetic activity. Equatorial scintillation has been found to increase significantly with an increase in the solar activity level,. The control is more pronounced at all-equatorial locations around the anomaly crest than at the equator [1]. It has also been observed that there is the dramatic increase in the scintillation activity during the equinoxes (February to April and August to November) and December solstice months with the increase of sunspot number. It has also been observed that scintillation is very prominent during post sunset hour of these periods. This related to the equatorial Ionospheric plasma processes. It is also interesting to observe that scintillation occurrence in the June solstice is spares and almost independent of solar activity.
It has been established that the equatorial irregularities can frequently produce signal fading of L- band transmissions from Geo Earth Orbiting Satellite (GEOS) in excess of 20 to 25dB at locations like Ascension Island (7.9 degree S, 14.4 degree W geographic 31 degree S dip), Kolkata 22.58 degree N, 88.38 degree E geographic, 32 degree N dip) and Hong Kong (22.12 degree N, 114.12 degree E geographic, 30 degree N dip) situated near the crest of the equatorial anomaly [2].
Equatorial scintillation was observed too much stronger (i.e. with greater impact on receiver tracking performance) than the high latitude scintillation. Percentage of corrupt observations often exceeds 40% for station located near the equatorial anomaly peak, in the local time sector 2000-2300. Receiver tracking performance in the equatorial region exhibits clear seasonal variations with peaks in the winter month and a dependence on the solar cycle. Based on observed static and a solar maximum in mid 2000 it is anticipated that the high levels of degraded tracking performance will continue in the equatorial region until early 2002[3].
In terms of WAAS, the scintillation will have significant effects on the WRS network i.e. if VSAT or terrestrial network is used. The impact of the scintillation will be interpreted with respect to Grid Ionosphere Vertical Error (GIVE) bounds, as derived for each grid Ionospheric estimates that will be transmitted to the users. The GIVE values reflect accuracy and reliability derived from networks of Ionospheric observations in the local area surrounding the grid points. Only the current observations will be used in the derivations of the GIVE values, such that an error bound reflects current statistics. If no observations are available within the local area, the GIVE will be marked unavailable for a given grid points.
Scintillation data base
The amplitude and phase scintillation measurements can be done with an Ionospheric Scintillation Monitor (ISM), a GPS receiver specifically designed to track the scintillation behavior of the atmosphere. The ISM receiver is a modified high performance single frequency GPS C/A code receiver equipped with wide bandwidth tracking loop. The wide bandwidth tracking loops improve the receiver ability to maintain signal lock during active periods. The receiver modifications include the use of a low noise, highly stable, oven controlled crystal oscillator as the receiver’s frequency reference. The receiver is also modified for high data rate processing of signal power and carrier phase measurements for calculation of scintillation –parameters in near real time.
The ISM calculates several parameters useful for scintillation analysis. The S4 parameter, measuring amplitude scintillation, is normalized standard deviation of the incoming signal intensity given by:
Intensity of scintillation is measured by different indices. The most commonly used index for scientific studies is S4, which is the second moment of the signal fluctuation around the mean signal level. The communication Engineers however have found SI (dB) as a simpler measure for describing scintillation. SI (dB) is obtained by scaling the difference in dB levels between the third peaks from the maximum to the third peak level from minimum of the signal. Empirical relation between the two measurements is available [7]. In this presentation the intensity of scintillation will be represented by SI (dB). Three case of scintillation corresponding to low value of S4 (<0.3), moderates value (0.3 to 0.6) and high values (>0.6) has been considered [4].
A comparison of scintillation effects on GPS / GLONASS signal with GEOS signal was observed at one station, which shows the considerable difference in scintillation occurrence pattern. Incase of GEOS, scintillation occurs when the clouds of Ionospheric irregularities drift across sub Ionospheric point of the satellites. At Kolkata, scintillation on 1.5 GHz channels of INMARSAT normally occurs in between local sunset and midnight. L-band scintillations normally moderate to intense during high solar activity years. VHF scintillation is usually saturated and very fast in the initial stage. In the post midnight hours the rate of fading is appreciably reduced although the signal exhibits saturated scintillation before gradually restoring to its steady level [2].
In case of GPS/ GLONASS constellation, 10 to 15 satellites together distributed over the sky different sections of the ionosphere around the station under varying look angles. In the post sunset hours, a number of GPS and GLONASS satellites signal often experience scintillation and a result scintillations appear to be more extensive for such system. Even in the post mid night hours, when no scintillation activity is observed GEOS microwave links, the L band GPS signals experience severe fluctuations on signal strength (C/ N0). The deep fading of GPS signal strength (C/N0 ) is interrupted in terms of propagation angle of the particular satellite ray path with respect to geomagnetic field direction. From low latitude station in the northern hemisphere, the orbiting satellite signals will be viewed “end on” through a field tube aligned equatorial plasma bubble introducing more intense C/ N0 fluctuation than in the case of the Geo Satellite which normally passes through a small section of the bubble.
The intense fluctuation of C/ N0 of the GPS signals when there were negligible or mild scintillation on the L band signal of a GEOS signal was initially puzzling. This has prompted to check the calibration of the GPS C/ N0 fluctuation with respect to that of the GEOS signal receiving system. The signal from a GPS satellites exhibiting moderate to intense but not saturated scintillations, traveling near the INMARSAT sub Ionospheric point was chosen for calibration. It has been observed there is a good correspondence between scintillation indexes of the two satellites when two sub Ionospheric points are close to each other [2].
Occurrence of scintillation
The scintillation occurs only when its strength exceeds the SI value of 3dB during 2000 to 2400hrs (IST) has been considered.. Scintillation increases both in frequency and amplitude (dB) with solar activity. Year 1999 to2000 was the peak period of the solar cycle and strong scintillation was experienced each day during this period. It is interesting to note that during February to April and August to November the scintillation effect was quite frequent during solar cycle. With increase in solar activity GPS + GLONASS signals are likely to be affected by scintillation in the equatorial region. Location near the equatorial anomaly crest would show the degradation of the navigation in a peculiar way. In the Northern Hemisphere scintillation may simultaneously render more than on GPS/GLONASS satellites in the southern sky unusable for navigation purpose. In addition to southern links any GPS/GLONASS satellites near overhead will also show intense scintillations due to higher ambient level of ionization around the crest of the equatorial anomaly. Fig- 2 shows the variation of scintillation index vs Time of IMMARSAT & GPS + GLONASS satellites for four hours at Kolkata. As a results the accuracy of position fixing may degrade to unacceptable limits
Fig-2 Variation of scintillation index of IMMARSAT & GPS+ GLONASS
Experimental data
The Asia pacific Air Navigation planning implementation Regional Group (APANPIRG) of International Civil Aviation Organization (ICAO) in its meeting held in October 2000, had decided to initiate a Regional GPS measurement campaign and Civil Aviation Administration of Singapore was assigned the task of coordination activity in the region. Accordingly, Civil Aviation Administration of Singapore had invited the Civil Aviation Administration of states in the region to participate in the GPS measurement campaign to determine “Normal and Peak Excursion on the GPS performance”. The measurement campaign carried out on 25-09-01 and 09-10-01 with each measurement lasting 24 hours period i.e. 00:00:00: to 23:59:59. The ASTECH (GPS+ GLONASS) 12-channel receiver was used for raw data down load and GPS receiver antenna was positioned at Latitude 28° 38. 22149’ N, Longitude, 77° 10.29609’ E, Altitude 192.37467 Meters. The ASTECH (GPS+ GLONASS) receiver is proved to be very robust at tracking the carrier signal amplitude and phase, but it experienced scintillation induced navigation outages on both days in nighttime.
Fig –3(a): Error vs time (∆Latitude, ∆Longitude and ∆Altitude)
The data sample was taken from 15:10:00 to 17:08:58(UTC) and 10 GPS satellites were visible during the sample period. The position error was calculated as coordinate of the reference point i.e. location of the GPS receiver antenna, is known. The position value was down loaded in the form of Latitude, Longitude and Altitude at the rate of one Hz . One sample of data is approximately for two hours. The position error is calculated with respect to the reference point. The DOP value is less than four during sample time. Therefore the GPS receiver has calculated the position with best geometry of the satellites, which are visible during sample time. The errors are ∆Latitude, ∆Longitude and ∆Altitude & calculated in terms of meters.
Fig –3(b): All visible GPS satellites SNR Vs time
Fig -3(c ): SNR Vs Time for SV-4
The position error vs time (UTC) are plotted for the all-visible satellite during sample period with good dilution of precision (DOP). It is observed in Fig 3-a that the position error plot contains the spikes in the measurement during entire sample period. The spikes over plots are due to scintillation. In further analysis it was observed that GPS receiver was tracking 10 satellites but 8 satellites were locked for position data measurement with good geometry. The two satellites could not lock due to scintillation or low elevation angle. In detail analysis, GPS receiver had also recoded the SNR of the all-visible GPS satellites. Fig 3-b shows the plot between SNR of all visible satellites vs Time (UTC) and it is observed some of the satellites SNR are fluctuating badly.
In detail analysis, it was found that GPS Satellites Vehicle (SV) 28,27,26,2,7, 8,11,4 and locked and their signal were affected by Scintillation during the sample period. The SV 28,27,26,7 signals was less affected, SV 11 was moderate and SV 8 , 4 were severely affected by scintillation. In Fig 3-c the severely effected satellites signal is plotted SNR vs Time (UTC). It is observed that GPS SV–4 signal was badly affected and some of the time satellite is unlocked which has definitely degrade the measurement performance.
The second sample of the data from 17:10:00 to 19:08:58 (UTC) was taken immediately after first sample of data logging was completed. During the sample period, the total number of satellites tracked and locked is 7. The similar exercise was repeated. Fig4-a is plot between position error vs Time (UTC) and it was observed that mid of the sample the spikes are prominently visible but not in continuous in nature.
Fig –4(a): Error vs time(∆Latitude, ∆Longitude and ∆Altitude)
Fig -4(b ): All visible GPS satellites SNR Vs time
Fig 4-b is plotted between all visible satellites SNR vs Time (UTC) and observed that mid of the plot is fluctuating during same period of sample. To further strengthen the observation individual GPS satellites SNR vs time (UTC) is plotted as Fig4-c to know about which satellite is badly affected by scintillation. The SV 28,24,20,9,7,4,2 were tacked and locked by receiver. SV 24 was severely affected. SV 20, 7 and 9 were less affected by scintillation.
Fig -4(c ): SNR vs time
The GPS data logging was carried out on 08-10-2001 for 24 Hours and raw data was recorded from 14:20:01 to 16:19:59 (UTC). The eight GPS satellites were in view but 7 satellites SV28, 11,27,26,8,7 and 4 were locked and their data was used for position calculation. The similar exercise was carried out which has been done in the previous section. Fig 5-a is plotted between position errors and Time (UTC). It observed that mid of the data, spikes are visible otherwise entire observations are free from spikes. The spikes are due to scintillation effects on GPS signal. To further strengthen the observation Fig 5-b is plotted between all GPS satellites in view vs Time (UTC). Similar spikes are observed in the mid of the data sample. In further analysis, to check which GPS satellites signal is badly affected by scintillation . In Fig 5-c the SNR vs Time (UTC), the SNR of GPS SV –2 is cutting off in the mid of the observation i.e. due to scintillation the GPS SV-2 is unlocking. Therefore the spikes are in the mid of the observation due to GPS SV-2.
Fig -5(a ): Error vs time(∆Latitude, ∆Longitude and ∆Altitude)
Fig -5(b) All visible GPS satellite
Fig –5(c): SNR vs time
The second sample of data for two hours was taken just after the first sample data recording was completed. The data was recorded from 16:21to 18:44:58(UTC). Fig 6-a is plotted between position errors vs time (UTC) and end of the data sample small spikes are visible otherwise entire data is free from the scintillation. The sample data for two hours has experienced least scintillation effect on 08-10-2001. The other parameters e.g. DOP, Errors and SNR were well within the limits. Fig 6-b is plotted between all GPS satellites in view vs Time (UTC) , shows least variation in SNR . Fig 6-c is plotted between SNR vs Time for most affected satellites and SV-20 shows that GPS receiver is unable to lock. It may be due to low elevation angle of the SV-20 in view or may be just appearing in the sky.
Fig -6(a ): Error vs time(∆Latitude, ∆Longitude and ∆Altitude)
Fig -6(b) All visible GPS satellites Vs Time
Fig –6(c): SNR vs time
Discussion
In the GPS data measurement campaign the data analysis shows that few of the GPS satellites were severely affected by scintillation but again affected GPS satellites has to be identified There is possibility that some or all of the affected satellites may provide good geometry and DOPs which will impact on position measurement accuracy as well as in the error calculation at WRS. The instantaneous position solution obtained during strong scintillation is compared with those of just before the onset of scintillation and after the disappearance of the scintillations. The variation of latitude and longitude of the present location determine by GG24 receiver shows slow variation with time even when links are undisturbed. When scintillation is observed on GPS signals the values of latitude and longitude shows large and fast deviation superimposed on the slow variations. In position measurement time if all lock GPS satellites signal are affected by scintillation may introduce large deviation in position measurement but if few are affected because of their location in sky may not show much deviation in the measurement.
Conclusion
The data collected here underscore the importance of GPS application during solar cycle. The varied behavior of GPS signal suggest that end to end test to be conducted for critical GPS application to insure that any potential degrading effects, such as those caused by scintillation are identified and accurately accessed. Understanding these effects remains an important goal and presents the formidable challenge for development of accurate model for region.
It is observed that in the month of September 2001 the scintillation effect was more prominent than October 2001.The data recorded at Delhi which falls in the region of the moderate scintillation which has been shown in the Fig-1. If this the case then other WRS which falls in the intense and most intense zone need more observation for modeling the Ionosphere for Indian Airspace. The seasonal change may also cause the GPS signal fading needs more study.
The GAGAN, will provide Ionospheric corrections to GPS signals for aircraft in the Indian Airspace for both en-route navigation and precision landing The effects of the scintillation is important for GAGAN, since WRS located all over the India will be under influence of equatorial scintillation. Finally with expansion of the GPS navigation on the global basis, it is too important to access the potential effects of scintillation navigation performance and availability in regions where scintillation is most active.
Acknowledgement
The author would like to thanks Dr. P. Banerjee , Deputy Director, Time and Frequency division of National Physical Laboratory for his kind support in the GPS data logging activity and valuable suggestion in the analysis of the data. Thanks to Shri S.H. Khan Executive Director (CNS-P) Airports Authority of India for extending full support and encouraged my team during data logging period.
References
- Das Gupta , A. A. Maitra & S. Basu “ Occurrence of night time VHF scintillation near equatorial anomaly crest in the Indian sector” Radio Science ,161455,1981.
- Studies on potentiality of GLONASS for positioning and Timing vis-à-vis application of GPS by Ministry of Information & Technology, & Defence Research and Development Organization, Ministry of defence, Government of India
- S. Skone and K. Knudsen “ Impacts of Ionospheric Scintillation on SBAS Performance “ ION GPS2000, 19-22 September 2000 Salt Lake city , UT
- Yannick Beniguel, I.E.E.A. “ Characterization of equatorial scintillations for trans Ionospheric links” ION GPS2000, 19-22 September 2000 Salt Lake city , UT
- Patricia H. Doherty, Susan H. Delay and Cesar E. Valladares, John A,. Kulbuchar” Ionospheric Scintillation effects in the equatorial and Auroral regions” ION GPS2000, 19-22 September 2000 Salt Lake city, UT
- K.M. Groves , S. Basu , J.M. quinn, T.R. Pedersen, K Falinski, A Brown and R Silva and P Ning “ A comparison of GPS Performance in a Scintillation Environment at Ascension Island” ION GPS2000, 19-22 September 2000 Salt Lake city , UT
- Whitney H.E. , J Aarons& C. Malik” A proposed index for measuring Ionospheric Scintillation” planet space science,171069,1969