Report on DME interference on GPS/L5 (third version, July 99)

Report on DME interference on GPS/L5 (third version, July 99) Draft I. Introduction This paper is the third report to Direction Generale de l Aviation Civile (DGAC) of a study on the potential risk of interference of DME (962-1213 MHz) on GPS/L5 third civil signal planned at 1176.45 MHz +/- 24 MHz. The aim of the first version of this report was to show that using the known characteristics of both systems, the airborne co-site scenario could lead to interference from DME CW spurious radiation. This first report was presented as a working paper at GNSSP 3 [8] and IUT WG 8D [9]. In the second report the analysis was extended to the potential risk of interference on the ground-air scenario due to the nominal pulsed DME signal transmitted by one ore more ground transponder and it was stated that in addition to the CW airborne co-site problem identified in the first report, further studies should indicate if nominal DME pulsed signal interference could be handled by GPS/L5 to confirm or infirm the compatibility between these systems. This second report was presented as a working paper at AWOG 6 [15]. In this third report we include and discuss the results of recent simulations in US showing that the use of the proposed L5 signal is not compatible with DME pulsed signals on high en route flight level. Only close to the ground, when a small number of DME ground transmitters are in line of sight of the airborne L5 user, are both systems possibly compatible. In this paper, we also introduce some modifications to the CW airborne co-site interference estimation taking into account a refined assumption on the L5/DME airborne antennas isolation as well as up to date allowed airborne DME spurious CW power data. GPS/DME Report, Third Version Draft, B.Roturier 03/09/01 1

II. Interference reference scenarios Although the definitive signal structure should be definitively approved in August 99, RTCA papers [1], [16] have given some detailed indication on the possible characteristics of the new GPS/L5 signal structure. Among the main features one may note: center frequency: 1176.45 MHz signal bandwidth: 24 MHz received power in a 0dB antenna: - 154 dbw (+6 db as compared to GPS/L1 C/A) Two obvious scenarios where harmful interference on an airborne GPS/L5 user could occur from DME transmitters are: The radiation in GPS/L5 band of several DME transponder (ground transmitter). The 1 Mhz spaced DME channel 77X (1164 MHz) to 101X (1188 Mhz) radiating into the 24 Mhz around 1176 Mhz are all candidates to pulsed mode interference from the nominal DME signal. Moreover, since the 24 Mhz passband filter centered at 1176.45 Mhz can not have an infinitely steep cutoff rate, a certain number of DME channel above 101X and below 77X should also interfere with GPS/L5. The radiation in GPS/L5 band of a DME interrogator (airborne transmitter). In this case, due to channel assignation in DME band (no airborne transmitter channel are allocated in the 24 Mhz around 1176 Mhz), only DME spurious CW radiation are considered. These scenarios are successively examined in the following II.1 and II.2 paragraphs. GPS/DME Report, Third Version Draft, B.Roturier 03/09/01 2

II.1. Ground (DME) to Airborne (GPS/L5 user) Interference II.1.1. Introduction In this scenario, we compare the characteristics of the DME pulsed interference to the currently acceptable thresholds defined in GNSSP for GPS/L1 C/A, since from [1] we may anticipate that the resistance of GPS/L5 to pulsed mode interference should be roughly the same than GPS/L1 C/A. In [4], paragraph B.3.7.3.3. Pulsed Interference', table 1 (Table B.3.7-5.) is specified showing the level of resistance of a GNSS receiver to pulsed interference: GPS and SBAS Frequency range 1575.42 Mhz +/- 10 MHz Interference threshold 0 dbw (Pulse peak power) Pulse width <=125 µs, <= 1 ms* Pulse duty cycle <= 10 % * Applies to GPS receivers without SBAS Table 1. Interference thresholds for pulsed interference The 10% pulse duty cycle limit result from a maximum allowed degradation of the GPS L1 C/A receiver post correlation signal to noise density in presence of pulsed interference. In the following we aim to assess representative values of the pulse peak power, pulse width and pulse duty cycle of DME so that to compare them to the numbers given in table 1. This analysis give an indication on the number of DME which may interfere with GPS/L5 without significant degradation of L5 service. Finally we present the results of a US simulation taking into account the real distribution of today DME/TACAN over continental US. This simulation allow to assess L5 service degradation. II.1.2. DME Pulse Peak Power Considering in first approximation that free space propagation occur with a relative distance d, the interference DME peak power at an aircraft GPS receiving antenna port is: P = ERP G 1 ground DME GPS/ L5 λ 4πd 2 (1) Where ERP ground DME is the effective radiated peak power supplied at the DME transmitting antenna port by the DME transponder and G GPS/L5 is the gain of the airborne GPS antenna at L5. In [2], the power budget for air-ground DME links show that a typical figure for ERP ground DME is: ERP ground DME 40 dbw (2) We wish in the following to compute the lowest possible value of the P 1 interference peak power, that is when the DME transponder is at the radio horizon and the aircraft is flying at the high reference flight level FL 400 (40 000 ft) as in figure 1. GPS/DME Report, Third Version Draft, B.Roturier 03/09/01 3

Figure 1. Ground to air interference scenario In this case and in standard propagation conditions, the distance d is approximately given by the formula: And the free space loss term in Eq. (1) is then: d( Nm) 123. 40000( ft) 246 Nm (3) λ 4πd 2 ( db) < 147 db (4) Considering a 0 db gain for G GPS/L5, we obtain for the minimum peak value of P1: P 1min 107 dbw (5) The nominal power of GPS/L5 in a 0 db gain receiving antenna should be -154 dbw, hence the minimum peak interference power is still 47 db higher than GPS/L5. This calculation show that all the DME in view of an aircraft are potential sources of pulsed mode interference. In the case where the aircraft is at only 10 m of the DME transponder, the free space loss is -54 db and the maximum peak interference may be assessed at: P 1max 14 dbw (6) In this case, the maximum interference peak power is 140 db above GPS/L5. From table 1, it may be seen that this value is under the 0 dbw peak power threshold, hence a GNSS receiver should be able to handle the incident power from a DME ground transponder at any position in space. II.1.3. DME Pulse Width According to [6], The DME half amplitude pulse width characteristics are 3.5µs +/- 0.5µs. However, in this study we are interested in the total time duration of the pulse which is approximately the double, that is 7µs. One also should not forget that since the GPS receiver will be saturated by DME pulses, a desaturation time which may reach 1µs [10] has to be added to compute the effective pulse width, which is the amount of time during which the GPS signal is not usable. Hence in the following we consider that this effective pulse width is: DME Effective Pulse Width = 7µs + 1µs =8µs (7) GPS/DME Report, Third Version Draft, B.Roturier 03/09/01 4

II.1.4. DME Duty Cycle The last important parameter of the interfering DME signal is its pulse duty cycle. It is not possible to obtain a single number for this pulse duty cycle since it is a non stationary variable mainly depending upon: the number of aircraft interrogating at a given time, the number of pulses sent by the airborne interrogator which may vary with the mode or the generation of hardware. New generation interrogators need less pulses to acquire and track the DME information, typically 10 pairs of pulses by second instead of 30 pairs of pulses by second for older ones. the number of DME in the L5 band at a given time. For the moment we consider the case of a single transmitting DME transponder. In Annex 10 [6], the DME transponder transmission rate is bounded by 800 pulse pairs per second as a minimum value and 4800 pulse pairs per second as a maximum. However, today the DME are set to a maximum transmission rate of 2700 pulse pairs per second and statistics show that in a heavily loaded area such as Paris area, typical values are 1200 to 1500 pulse pairs per second reaching 2000 pulse pairs per second in intense traffic conditions. Hence it seems very unlikely that the 4800 pulse pairs per second transmission rate be ever reached, this value resulting from an earlier analysis of an intensively loaded scenario with older generation equipment needing 2 to 3 times more pulse pairs per second. Consequently, to reach reasonable worst case figures for the pulse duty cycle we consider in the following that a single DME transponder could have a continuous transmission rate of 2700 pulse pairs per second 1. With these hypothesis, we may assess that the DME pulse duty cycle for a single tranponder will be bounded by: Single DME pulse duty cycle < 2700*2*8/1 000 000 = 4.3 % (8) It has to be noted that in other studies a lower 2,5% pulse duty appears. II.1.4. Comparison with Table 1 Comparing (6), (7) and (8) with the second, third and fourth lines of table 1, it may be seen that a single (pulse cycle < 4.3%) or two (pulse cycle < 8.6 %) and even three (pulse cycle < 12.9 %) 2 simultaneously transmitting DME in GPS/L5 should have negligible effect on the GPS signal. If higher duty cycles are acceptable with new signal processing techniques such as pulse blanking or auto-blanking [10], (in [11] a 20% duty cycle is said to be possible), a larger number of simultaneously transmitting ground tranponder in view of an elevated airborne GPS receiver could be accepted. 1 It has to be noted that today TACAN (military DME operating in the same band) transmit at this same fixed rate of 2700 pulse pairs per second. 2 In practice it should be very unlikely that both three DME transmitters transmit at the maximum rate of 2700 pairs of pulses per second so that the practical duty cycle would have a high degree of probability to be lower than 10%. GPS/DME Report, Third Version Draft, B.Roturier 03/09/01 5

II.1.5. Results of US studies and conclusion At RTCA SC-159 WG1-3 rd civilian frequency meeting held May 13, 1999 at Washington DC, several papers [17], [18], [19] presented the results of a detailed analysis of the degradation of the GPS L5 receiver post correlation signal to noise density in presence of DME/TACAN interference over continental US using the today distribution of DME ground transmitters. The main GPS/L5 receiver assumptions where that: either a 5.5 db/mhz or a 10 db/mhz attenuation of input signals beyond +/- 10 MHz of 1176.45 MHz (up to 70 db max. attenuation) existed due to the RF input filter pulse blanking technique was employed investigated receiver altitudes where 40000 ft and 5000 ft. The main result of these simulations where that whichever the 5.5 db/mhz or 10 db/mhz mask was employed, due to the large number of ground DME/TACAN stations contributing to the pulse duty cycle, the post correlation signal to noise density degradation was acceptable only if a number of DME/TACAN ranging between 10 and 20 where removed inside L5 band. This solution seem very difficult to adopt in Europe. If all DME/TACAN channels are to be conserved in Europe, then the only solution to obtain a full compatibility between DME/TACAN pulsed signal and GPS in all phases of flight would be that US adopt a smaller bandwidth signal so that the number of interfering DME be reduced 3. 3 It has to be noted that if a smaller bandwidth is used, for a given insertion loss, a passband filter with a steeper cutoff rate may be used, hence also reducing the number of outband DME interferers. GPS/DME Report, Third Version Draft, B.Roturier 03/09/01 6

II.2. Airborne (DME) to Airborne (GPS/L5 user) Interference In the second scenario, considering the isolation I between the aircraft DME transmitting antenna and the GPS receiving antenna, the interference DME peak power at an aircraft GPS receiving antenna port is: Paircraft DME P2 = (9) I Where P aircraft DME is the peak power supplied at the DME airborne antenna port by the DME interrogator. II.2.1. Isolation estimation A straight estimation of isolation between GPS and DME antennas is difficult since this parameter depend on the size of the aircraft, the wavelength and the relative position (same or opposite sides of the airplane). In VHF band, due to the use of several (up to three) communication antennas for a long time, some validated and accurately measured data exist. On large Airbus or Boeing aircraft, typical same side isolation is 35 db or better and 45 db opposite side isolation [12]. On the contrary on smaller aircraft, measurements [13] have shown that same side isolation is restricted to about 20 db and opposite side isolation to 34 db. It has been suggested [14] that free space loss factor be used to assess isolation between DME and L5 antennas. Although, the use of the free space loss factor may give an indication of the large scale variation of the isolation with frequency (which should theoretically be a variation of 20 db/by decade), the complex phenomena resulting from the interaction of electromagnetic waves with aircraft framework (such as resonance, diffraction, depolarization,..) may lead to (usually unpredictably) lower isolation at some frequencies 4. If we start from VHF isolation data, and add a conservative 10 db to take into account some variation due to non free space phenomena around the large scale trend of isolation increasing with frequency, we obtain a set of isolation I ranging from 30 db to 55 db. Another source of data is isolation at L band in the case of double DME antennas airborne installation. It is stated in [7] par. 7.2.1.4 that typical isolation between both (same side) DME antennas are between 20 and 40 db. If we add 10 db of isolation to take into account the fact that DME and GPS antennas should usually be located on opposite sides we again obtain a set of isolation I ranging from 30 db to 50 db. Hence in the following we consider a set of isolation I between GPS and DME antennas which seems coherent with both above analysis ranging from 30 (smaller aircraft) to 60 db (larger aircraft) 5. 4 For example in [13], measurements at 118, 127 and 136 MHz have shown that upper-upper isolation was respectively 23, 20 and 19 db, that is a decrease of isolation with increasing frequency, in contradiction with free space law. 5 In version 1 and 2 of this report a range of 10 to 40 db was assessed. GPS/DME Report, Third Version Draft, B.Roturier 03/09/01 7

II.2.2. DME CW spurious radiation power From [6] and [7], the CW ERP in the whole DME band radiated by an airborne transmitter are allowed to be: ERPmax= - 77 dbw in 960-1015 and 1045-1215 Mhz ERPmax= - 87 dbw in 1030-1090 Mhz In the GPS/L5 frequency band, the -77 dbw figure has to be considered 6. II.2.3. Comparison and conclusion Using (9) and 77 dbw we obtain table 2 below: GPS/DME CW Interference Power Isolation 60 db -137 dbw 50 db -127 dbw 40 db -107 dbw 30 db -97 dbw Table 2. Spurious CW signal interference power versus aircraft isolation parameter Table 2 CW interference power may be compared to the threshold accepted for GPS/L1 C/A. In [4], paragraph B.3.7.2. CW Interference, table 3 (Table B.3.7-1.) is specified: Frequency range fi of the interference signal Interference thresholds for receivers used for precision approach phase of flight...... 1525 MHz<fi<=1565.42 MHz Linearly decreasing from -42 dbw to -150.5 dbw 1565.42 Mhz<fi<=1585.42 MHz -150.5 dbw 1585.42 MHz<fi<=1610 MHz Linearly increasing from -150.5 dbw to -60 dbw...... Table 3. CW-Interference thresholds for GPS and SBAS receivers Comparing table 2 and 3 it may be seen that if the signal L1 C/A was used instead of L5, the DME spurious CW interference power would exceed the allowed -150.5 dbw. Table 4 show that for a large isolation of 60 db, the spurious CW may exceed the L1 C/A threshold by 13.5 db while for a low 30 db isolation, which could occur on a small sized aircraft, the spurious CW may be 43.5 db above threshold. 6 In the second version of this report, the -60 dbw figure issued from now obsolete ED-48 Eurocae MOPS was considered. GPS/DME Report, Third Version Draft, B.Roturier 03/09/01 8

GPS/DME Isolation DME spurious CW level above L1 C/A CW interference threshold 60 db 13.5 db 50 db 23.5 db 40 db 33.5 db 30 db 43.5 db Table 4. Gap between DME spurious CW and assumed L5 CW interference threshold versus aircraft isolation parameter For the new signal this gap will be reduced by 6 db, due to a higher GPS/L5 power, and also probably by a larger processing gain due to a larger bandwidth, more spectrally efficient code. In [1] the processing gain of candidate GPS/L5 signal versus GPS/L1 C/A over CW interference is assessed at 12 db worst case. This lead to a global 18 db gain of L5 versus L1 in CW interference rejection leading to table 5. GPS/DME Isolation DME spurious CW level above assumed L5 CW interference threshold 60 db -4.5 db 50 db 5.5 db 40 db 15.5 db 30 db 25.5 db Table 5. Gap between DME spurious CW and assumed L5 CW interference threshold versus aircraft isolation parameter From the previous analysis, it may be seen in table 5 that the global gain of 18 db on CW interference which might occur with the new L5 signal could help to overcome the gap between DME allowed CW spurious radiation and CW interference threshold when large isolations are involved, however for lower isolation this DME CW spurious CW are still above the allowed threshold. It is likely that new signal processing techniques such as FFT or adaptive notch filters could be implemented into future GPS/L5 receivers. However it is not clear at this step whether these techniques could be able to completely annihilate the radiated CW interference, especially for lower isolations. Further studies should be dedicated to investigate how to handle this interference threat. GPS/DME Report, Third Version Draft, B.Roturier 03/09/01 9

III. Conclusion This report identifies two major issues concerning the compatibility between L5 GPS and DME: - Effect of the ground DME pulses at the L5 receiver when these pulses are coming from ground transponders inside and close to L5 band. It seems that only in the case of DME pulsed signals radiated by a small number of ground transmitters could these signals be compatible with an airborne GPS/L5 user. Operationally this would correspond to low altitude aircraft for which the radio-horizon line of sight blocking would limit the number of potential ground DME interferers. However, at higher flight levels, simulations over US have shown that the number of potential interferers is so important that L5 operation is not possible over large areas in presence of today DME allocation. The proposed solution in US is to remove from assignation a number of DME channel (from 10 to 20) inside L5 band. Due to frequency planning constraints, this solution seem very difficult to implement in Europe. If all DME/TACAN channels are to be conserved in Europe, then the only solution to obtain a full compatibility between DME/TACAN pulsed signal and GPS in all phases of flight would be that US adopt a smaller bandwidth signal so that the number of interfering DME be reduced. - Effect of the airborne CW spurious emissions within the DME band, regardless from the DME operating channels itself. This other potentially harmful interference comes from allowed CW radiation of airborne DME interrogators. Further studies should indicate if signal processing techniques are able to completely annihilate the effect of the DME radiated CW interference, especially for lower isolations. GPS/DME Report, Third Version Draft, B.Roturier 03/09/01 10

IV. Bibliography [1] A.J. Van Dieredonck, P. Reddan, Analysis of Proposed GPS signal Structures for Use in ARNS Bands, Institute Of Navigation National Technical Meeting, San Diego, Jan. 99. [2] ICAO Annex 10, vol.1 Attachment C, paragraph 7, Guidance material concerning DME. [3] Assessment of Radio Frequency Interference Relevant to the GNSS, RTCA paper N 386-96/TMC-242, 30 Dec 1996 [4] ICAO GNSS P 3 Draft SARPS,Montreal 12-23 April1999. [5] Communication with A. Renard, Sextant Avionique [6] ICAO Annex 10, vol.1, paragraph 3.5.5.1.3. Pulse shape and Spectrum [7] Minimum Operational Performances Requirements for Distance Measuring Equipment (DME/N and DME/P) Interrogators Operating Within the RF Range 960 to 1215 MHz (Airborne Equipment) EUROCAE ED-84, January 1987 [8] 'Preliminary report on DME interference on GPS third civil signal', GNSSP3 Montreal 12-23 April1999. WP 21. [9] 'Preliminary report on DME interference on GPS third civil signal', IUT WG 8D WP/8D306, Geneva, 19-28 April 1999. [10] Communication with J. L. Issler, CNES [11] Compatibility of a New RNSS Signal With Existing Systems in the 960-1215 MHz and 1215-1400 MHz Bands,, IUT WG 8D WP/8D274, Geneva, 19-28 April 1999. [12] ARINC 716 [13] R.S. Haendel, Antenna to antenna isolation of a general aviation aircraft, RTCA SC-172, WP 284, November 17 1998. [14] United Kingdom, Onboard compatibility analysis of DME (in the band 960-1215 MHz) and a proposed radionavigation satellite signal (1176+/-10 MHz) CEPT/ERC/CPG/PG3(99) 35. [15] Annex to Position on the request to allocate a future GPS frequency (L5) in the DME frequency band, AWOG/6, Paris, 18 to 20 May. [16] J.J. Spilker, A.J. Van Dieredonck, Proposed third civil GPS signal at 1176.45 MHz Inphase/Quadrature Codes at 10.23 MHz chip rate, May 11, 1999 Revision B6. [17] C. Hegarty, DME/TACAN Interference to GPS/L5, RTCA SC-159 WG1 May 13 1999, Washington DC. [18] C. Hegarty, T. Morrissey, P. Reddan, A.J. Van Dieredonck, Compatibility determination methodology for IGEB Third Civil Signal Implementation Steering group Rev. 3, RTCA SC-159 WG1 May 13 1999, Washington DC. [19] T. Kim, S. Ericson, GPS L5 Receiver post correlation S/No degradation caused by DME/TACAN, JTIDS, ARNS and RADAR,, RTCA SC-159 WG1 May 13 1999, Washington DC. GPS/DME Report, Third Version Draft, B.Roturier 03/09/01 11