Characteristics of mobile satellite L-band signal in mid-latitude region: GPS approach

Transcription

1 Indian Journal of Radio & Space Physics Vol 40, April 2011, pp Characteristics of mobile satellite L-band signal in mid-latitude region: GPS approach B Voon Pai*, W A W Z Abidin, A K Othman, H Zen & T Masri Department of Electronics Engineering, Faculty of Engineering, Universiti Malaysia Sarawak, Sarawak, Malaysia voon_pai@yahoo.com Received 21 October 2010; revised received and accepted 11 March 2011 Mobile satellite L-band signal that propagates from the transmitting satellite to the receiver experiences impairment mainly due to shadowing and multipath effects. In the present study, open space data has been obtained for Fukuoka, Japan (mid-latitude) using portable global positioning system (GPS) satellite receiver. The received signal performance, fading characteristic and cumulative distribution function of certain GPS satellites over the sky of Fukuoka have been observed and analysed. A general mathematical model representing the signal strength for the open space environment has been deduced based on the collected data. The outcome of this study can be used in future research to determine the effect of different mobile satellite environment on the arriving mobile satellite signal in improving the quality of service perceived by mobile satellite users. Keywords: Mobile satellite L-band signal, Mobile signal performance, Satellite propagation data, Signal-to-noise ratio (SNR) PACS No.: Ua 1 Introduction Mobile satellite (MS) system has been designed to provide communication services to area where terrestrial communication is not available. Wide area coverage, higher bandwidth and the ability to serve users while they are on the move are the main characteristics of the MS system and reflect the reliability of the system to the mobile users 1. Therefore, its reliability is higher and more flexible in term of navigation system on the moving vehicles. It also serves as an alternative to the current communication system which has a high potential for further development in order to support the ever increasing demand of personal and multimedia communications 2. MS communication system operates in L-band. The L-band, with frequencies 1-2 GHz, is particularly effective in providing rapid and flexible communication through mobile and portable terminal or transportable earth station 3-6. The voice and data communication services especially can be provided through the link between the satellite and the ground terminal. Mid-latitude is the region in the latitude range of North and South, respectively 7. For the present study, data obtained in mid-latitude region of Fukuoka, Japan (33 35 N of equator and E of Greenwich) (ref. 8) has been analysed. The attenuation effects such as shadowing and multipath degrade the received signal quality of the MS system. Multipath fading is due to the arrival of the reflected line-of-sight (LOS) signals of different amplitude and phase which can cause signal fluctuations 9. These reflections are due to the surrounding environment such as buildings, trees and electric poles. In addition, shadowing effect is caused by presence of obstacles that impede the visibility of the LOS signal from the satellite such as building and trees 9. However, the current study only concentrated on satellite received signals in open space environment. A simple and low-cost data acquisition GPS system had been used to carry out measurement for this study. Common methods used to carry out experimental work utilize the existing geostationary satellite network or airborne platform such as airplane and helicopter However, both methods are complicated and costly. Moreover, these methods limit the number of studies that can be carried out for the MS link. GPS system is a constellation of Medium-Earth- Orbit (MEO) satellites arranged on six planes where

2 106 INDIAN J RADIO & SPACE PHYS, APRIL 2011 at least four satellites can be placed equidistantly 12. It transmits signal at L-band ( MHz) and is mainly used for navigation purposes. Current GPS receiver provides important satellite propagation data which can be extracted from the raw GPS signal. This data can give the insight for the MS signal link quality. Apart from being portable, GPS receiver is cheap and readily available. Utilizing the GPS receiver greatly reduces the time of experimentation. Another advantage is that the portable GPS receiver is easy to manage as compared to the existing geostationary satellite network or airborne platform. This GPS receiver has patch antenna of radiation pattern and allows perfect signal reception from boresight with the response attenuated as elevation angle decrease 13. Satellite signals are received via the right-hand circularly polarized (RHCP) antenna 14. Typical coverage is 160 with gain variations from about 2.5 dbi at zenith to near unity at an elevation angle of 15. Below 15 of elevation, the gain is usually negative. Therefore, propagation data of up to twelve different satellites or more can be measured simultaneously using the GPS receiver at any instant of time for different number of propagation environments and with different propagation parameters such as elevation and azimuth. The significant propagation parameters especially signal-to-noise ratio (SNR), elevation and azimuth have been used to study and predict the propagation signal performance of the received MS signal in open space environment. By using the portable GPS receiver, measurements have been carried out under fine weather with clear open sky. Raw GPS propagation data has been automatically recorded for later analyses. The MS channel characteristics have been evaluated in terms of the SNR, signal fading time series with respect to the elevation and azimuth angles. Further, the fading characteristic of the signals can be pointed out through the graphs of average SNR of certain GPS satellites against the elevation angles. The cumulative distribution function (CDF) of SNR difference at any given probability has also been presented. Moreover, comparison among different GPS satellites propagation have been observed based on each characteristic. In addition, the general mathematical model of the SNR with respect to the elevation angle has been obtained. 2 Experimental details The experiment has been set up in open space environment. The GPS receiver, connected to a laptop, has been placed at the top of a building with the display facing the sky with no obstacles within 30 m radius of the receiver. Hence, the LOS signals can be obtained from GPS satellites. Furthermore, this minimized the effects of multipath and also shadowing effects due to the presence of obstacles such as buildings and trees. The measurements were taken under sunny and clear sky conditions in order to eliminate the attenuation due to rain, storms, heavy cloud and strong wind. A portable and cost-effective data acquisition system has been used for the measurements. It is a method of communication between computer and GPS receiver to receive message from the satellites. This commercially available, portable handheld GPS receiver can track several satellites at the same time and provide propagation data such as SNR, elevation angles and azimuth angles. The GPS receiver is connected to the computer via serial port. The GPS signal currently tracked by the receiver and the sentences collected were saved for analysis. Data from each satellite has been identified from their respective pseudorandom number (PRN); PRN2 indicates signal from satellite with PRN of 2. The data has been collected at Hakozaki Campus in Kyushu University, Japan for this study. In order to validate the obtained data, same measurements needed to be repeated. For all measurements, the display of GPS receiver had been placed flat facing the sky. This arrangement enabled SNR of many satellites visible above the receiver in the sky. The receiver also has the ability to reject signals arriving from behind and this removed the effect of multipath signal due to the ground. From this set up, data for a number of satellites for different elevation and azimuth had been measured simultaneously. A sample of raw data obtained from GPS receiver by a computer is shown in Fig. 1. This data is in the National Marine Electronic Association (NMEA) data format 15, a specification that defines the interface between various pieces of marine electronic equipment. GPS receiver communication is defined within this specification. The idea of NMEA is to send a line of data called a sentence that is totally self contained and independent from other sentences. For this study, the sentence of interest is the GPGSV

3 PAI et al.: MOBILE SATELLITE L-BAND SIGNAL CHARACTERISTICS IN MID-LATITUDE REGION 107 (Satellite in View) sentence which provides important propagations data (SNR, elevation and azimuth). Since propagation data of different satellites arriving from different directions can be recorded at any time, it is, therefore, possible to compare the received signal quality between those satellites from a similar set of measured raw data. In order to decode the required parameters in NMEA sentences, NMEA extractor had been developed. This program is developed using Visual C++ programming language and the NMEA sentences are parsed according to their corresponding satellites from raw data. After extracting the corresponding satellite propagation parameters for those visible satellites, analysis of the measured satellite signals is necessary. The analyzed data can be displayed in different forms as shown in Fig. 2. The extracted data can be presented into graphical form with respect to SNR, elevation and azimuth angles during the visible periods for each satellite. The graphs have been plotted in the same cartesian plane in order to observe the relationship of the SNR, elevation and azimuth angles. Hence, the SNR pattern can be observed and analyzed with respect to the corresponding elevation and azimuth angles. In order to plot the graph of average SNR against elevation angles, taking the average value of SNR of the total number of data collected and generated it with the elevation angles. The graphs have been plotted to determine the fading characteristic. These graphs can be obtained by getting the average value of SNR for each elevation angles for each of those selected satellites. Then, the normalized SNR values of the average SNR have been calculated. Only data collected above or equal to 15 of elevation have been considered for analysis. The analysis is done by plotting all the outputs in the same graph. In addition, CDF is another way to analyze the data obtained. This has been done by calculating the cumulative probability of SNR difference between each SNR value with the SNR reference (SNR ref ) value of 44 dbhz. Lastly, general mathematical model has be derived from the relationship of the overall average SNR with the elevation angles in open space environment. This model has been obtained by using the curve fitting method. The process flow of this study has been summarized in Fig Results and Discussions There are a number of satellites visible on the sky of Fukuoka, Japan as listed in Table 1. The satellites which appeared for more than 6 hours are PRN 1, 2, 3, 4, 5, 7, 9, 10, 11, 14, 16, 17, 18, 19, 20, 22, 24, 26, 28, 29 and 30. These 21 satellites are considered as operational satellites. In contrast, satellites which appear for less than 6 hours are PRN 6, 8, 13, 15, 21, 23, and 27. These satellites also have lower elevation angles which are less than 55. These are considered as redundant backup satellites. For the operational satellites, satellites of PRN 2, 3, 4, 10, 11, 14, 16, 17, 19, 24, 26, and 29 have variable azimuth angles of 50 o Fig. 1 GPS receiver raw data Fig. 2 Data analysis in different forms Fig. 3 Process flow for the study

4 108 INDIAN J RADIO & SPACE PHYS, APRIL 2011 (actual 200 ) to 89 (actual 356 ) and then decrease directly to 0 and increase to 40 (actual 160 ) as well as the variable elevation angles increase from 0 to around 88 and then decrease to 0. However, satellites of PRN 1, 5, 7, 9, 18, 20, 22, 28, and 30 has variable azimuth angles without decrease to 0. These satellites also have variable elevation angle increase from 0 to around 88 and then decrease to 0. The satellites of PRN 1, 2, 3, 7, 9, 11, 14, 17, 18, 19 and 28 have been chosen for analysis. These satellites appear for 6 hours or more, have maximum Table 1 Local time, period and maximum elevation angle of visible satellites Satellite Local time, hhmmss Period Maximum Elevation ( ) h50m38s h40m46s h44m26s h14m48s h17m20s h8m32s h54m34s h30m26s h16m16s h19m56s h6m22s h8m46s h38m24s h34m28s h4m h36m26s h49m h28m18s h18m40s h37m10s h8m24s h6m28s h27m24s h12m20s h7m48s h58m20s h32m20s h22m18s h27m28s h16m26s h28m44s h46m50s h24m34s h36m22s h34m42s h22m20s 75 elevation angle of 70 or greater, and have elevation angle that starts and ends at 0 (complete set data). The graphs showing an open space measured SNR with respect to their corresponding elevation and azimuth angles have been plotted. The measurements have been carried out on a rooftop of building. The azimuth angle had been divided by a factor of 4 for the entire graphs so that the axis of elevation and azimuth are on same axis. It is observed that the SNRs for those different satellites show similarity in their respective SNR pattern over the elevation and azimuth angles as shown in Figs 4 and 5. The SNRs of the signals arriving at elevation angles greater than or equal to 15 (θ 15 ) have been found to be greater than or equal to 44 dbhz (SNR ref ) most of the time which indicates no significant fading effect for those satellites 9. For elevation angles less than 15 (θ <15 ), SNR starts to drop below 44 dbhz SNR ref due to the design of the antenna as the gain is low at lower elevation angles 16. For this type of portable receiver, the loss of signal occurs when the SNR value is less than 30 dbhz. It has been indicated as 0 dbhz by the receiver 17. The azimuth angles patterns are different from one another as observed in Figs 4 and 5. Therefore, it is clear that SNR values are not affected by the change of the azimuth angles. Moreover, both have the same shape of elevation angles throughout the time period. Hence, it can be concluded that SNR is determined by the elevation angle. In Fig. 4, the time axis has the range s. The data of PRN1 is collected for 6 hrs 50 min 38 s. The signal started at 02:12:28 hrs and lasted until 09:03:06 hrs. The data of PRN2 in Fig. 5 is collected for 7 hrs 40 min 46 s which started at 05:51:04 hrs and lasted until 13:35:30 hrs. These different satellites raw data can be obtained from the same measurements. Another advantage of using this type of GPS receiver has been that it enabled one to obtain SNR values for different elevation angles. The elevation range achievable is between 0 and 88 as shown in Figs 6 and 7. The measurements using the existing geostationary satellites network have to be carried out from different locations in the world in order to obtain data for different elevation angles. On the other hand, measurement using an airborne platform (helicopter or airplane) would enable measurement over different elevation range but are very costly. Therefore, measurement done using the GPS receiver definitely

5 PAI et al.: MOBILE SATELLITE L-BAND SIGNAL CHARACTERISTICS IN MID-LATITUDE REGION 109 reduce the cost of research and also make it possible to study the MS link over a wide elevation range without moving from one location to another 9. In addition, the received signal with a small sequence of fluctuations along the time period are due to multipath effect as for open space, the signal will experience less than 5 db drops from peak-to-peak 11. The SNR values for each satellite signal at every elevation angle has been added and the average values obtained. Then, overall average SNR for those selected satellites for each elevation angles has been calculated and plotted as shown in Fig. 8. The SNRs are always in the range of dbhz whereas it is always greater than or equal to 44 dbhz for θ 15. Since there is no significant fading effect for θ 15, hence the results of this open space data has been used as reference in order to determine the effect of different MS environment on the arriving MS signal. At θ<15, SNR values drop drastically to 0 dbhz as expected. Figure 9 shows the fading characteristics for the open space measurements of eleven selected satellites that orbit the sky of Fukuoka measured using GPS receiver. Only the SNR values for θ 15 are Fig. 4 Graph of SNR, elevation and azimuth versus time of PRN1 Fig. 6 Graph of SNR, elevation and azimuth versus time of PRN3 Fig. 5 Graph of SNR, elevation and azimuth versus time of PRN2 Fig. 7 Graph of SNR, elevation and azimuth versus time of PRN7

6 110 INDIAN J RADIO & SPACE PHYS, APRIL 2011 considered. Since the satellite signals experience 5 dbhz fluctuation from peak to peak because of the multipath effect and radio noise, hence the signals are considered to experience fading if and only if the SNR values of satellites are below the range. However, all signals have shown good results since all the SNR values are always within the range. Therefore, these results imply that this characteristic for open space measurement is similar and significant attenuation effects are not present in Fukuoka, Japan. Furthermore, CDF characteristics for the open space measurements of eleven satellites that orbit the Fig. 8 Overall average SNR against elevation angle sky of Fukuoka are shown in Fig. 10. The CDF curves show good agreement between each other implying that this characteristic for open space measurement is similar and without presence of significant attenuation effects. For example, at 10% probability, SNR differences of 6 8 dbhz have been obtained for these visible satellites in Fukuoka. Figure 11 shows different types of regression plots with respect to the original data. Figure 11(a) shows the linear regression with the R 2 value equal to which means that only 30.15% of the original data can be calculated from the regression equation. However, Fig. 11(b) shows the polynomial second order regression curve where it has higher R 2 value as compared to linear regression. Furthermore, the polynomial sixth order regression also has been shown in Fig. 11(c) with R 2 value equal to which has the highest value as compared to others. However, it is not a smooth curve as needed and its corresponding equation is too long. Figures 11(d, e and f) display the exponential, logarithmic and power regression, respectively. From these three types of regression, logarithmic regression has higher R 2 value that is In addition, it also has smoother curve and near to the original data as observed. Exponential regression only has R 2 value of and it is a linear line. Power regression also has a smooth curve close to the original data, but its R 2 value of is lower as compared to logarithmic regression. As a result, the ideas of logarithmic and polynomial regressions are used. The best fit Fig. 9 Graphs to check fading characteristic Fig. 10 The CDF of satellites that are visible across the sky of Fukuoka

7 PAI et al.: MOBILE SATELLITE L-BAND SIGNAL CHARACTERISTICS IN MID-LATITUDE REGION 111 regression shows the relationship of the SNR with respect to the elevation angle of all the satellites in open space and is plotted by using MATLAB (ref. 18) as shown in Fig. 12. This regression has R 2 value of which means that 98% of the variance in the SNR values can be accounted for by this regression equation. Therefore, this regression gives the general mathematical model that represent the overall signal performance in open space as: 4 3 SNR = ( lnθ) ( lnθ) ( lnθ) ( lnθ) (1) where, θ, is elevation angle. This is a fourth order logarithmic polynomial equation with respect to Fig. 11 Regression curves Fig. 12 Best fit regression of data obtained

8 112 INDIAN J RADIO & SPACE PHYS, APRIL 2011 elevation angle. By using this mathematical model, the SNR values at any elevation angle can be expected without measuring the received signal from the GPS receiver. 4 Conclusions The present work has been carried out for open space environment in Fukuoka (mid-latitude) by using the commercially available GPS receiver. The ability of GPS receiver to provide SNRs of many GPS satellites at any instant of time allows comparison to be made with each other. The SNRs measured from different number of satellites have been found to be consistent over a wide range of elevation angle and azimuth angle with SNR exceeding SNR ref for θ 15 o. The use of the portable GPS receiver allows measurement to be carried out for a wide elevation range. Hence, elevation of up to 90 o has been observed. Furthermore, all signals have shown similar results and no significant attenuation effects was present since all the SNR values have always been within the 5 db range of fluctuation from peak to peak. In addition, the CDF characteristic for different GPS satellites that orbit Fukuoka showed good agreement between each other. Therefore, general mathematical model representing the signal strength for the open space environment is obtained based on the data collected. The outcome of this research work can be used in future research work as reference to determine the effect of different MS environment on the arriving MS signal. Acknowledgements The authors would like to thank Universiti Malaysia Sarawak for the financial support under the Dana Khas Penyelidikan References 1 John F & Ramjee P, IP/ATM Mobile satellite network (Artech House, Boston, London), 2002, Papathanassiou A, Salkintzis A K & Mathiopoulos P T, A comparison of study of the uplink performance of W-CDMA and OFDM for mobile multimedia communications via LEO satellites, Personal Communications, IEEE, 8 (3) (2001) pp Louis J I, Satellite communications systems engineering: Atmospheric effects, satellite link design and system performance, 1st ed (John Willey, UK), 2008, Gerard M & Michel B, Satellite communications systems: Systems, techniques and technology, 4th ed (John Willey, England), 2002, Takashi L, Satellite communication: System and its design technology (Ohmsha, Japan), 2000, 4. 6 Ray E S & Y Fun Hu, Mobile satellite communication networks (John Wiley, England), 2001, Arthur N S & Alan H S, Elements of physical geography (John Wiley, New York), 1984, planetbiomes.org/climate.htm. 8 Location of Japan, file/chap1/gaikyo-e.htm, Abidin W A W Z, Fujisaki K & Tateiba M, Empirical study on mobile satellite link using portable GPS receiver: Proc. JC-SAT 2006 (South Korea), 2006, pp Bråten L E, Amaya C & Rogers D V, Statistical characterization of land mobile satellite propagation environments using a photogrammetric technique: Proc URSI General Assembly 2002 (Maastricht, Netherlands), 2002, Julius G & Wolfhard J V, Handbook of propagation effects for vehicular and personal mobile satellite systems: Overview of experimental and modeling results, 1998, http: // 12 Michael K, The Global Positioning System and GIS: An introduction, 2nd ed. (Taylor & Francis, London), Abidin W A W Z, Fujisaki K & Tateiba M, GPS receiver and mobile satellite signal performance for equatorial region, Malaysia: Proceedings of ISAP2007, (Niigata, Japan), 2007, pp Manandhar D & Shibasaki R, Possibility analysis of polarization diversity scheme for multipath mitigation in GPS receivers, 2004 International Symposium on GNSS/GPS, Sydney, Australia, 6 8 Dec Dale D, NMEA data, 20 August Kaplan E D & Hegarty C J, Understanding GPS: Principles and applications, 2nd ed (Artech House, London), Abidin W AW Z, Fujisaki K & Tateiba M, Novel approach to determine the effects of MS environment using the portable GPS receiver with built-in antenna, Am J Appl Sci (USA), 5 (8) (2008) pp Aqua Phoenix, Curve fitting and regression: Matlab Lecture 9.3, http: // page3.html,