RNSS Wide band and narrow band performance against Interference from DME/TACAN in the band MHz (Over Europe)

Transcription

1 Liaison Statement to GNSS-P (copy to CEPT/CPG/PT3) RNSS Wide band and narrow band performance against Interference from DME/TACAN in the band MHz (Over Europe) 1 Introduction : During the last ITU-R CPM meeting, the idea of superposing a narrow band (±1MHz) to a wide band (±12MHz) RNSS signal to achieve compatibility between RNSS European system and DME/TACAN in a part of the range MHz, was introduced. This liaison statement gives the results of studies which have been performed by CEPT/SE28 in order to have a better understanding of the interest of having narrow band signal in terms of DME/RNSS sharing together with additional simulation results which are felt of interest for ICAO GNSS-P. Simulated results are provided for narrow band and wide band in term of degradation of S/N. Five possible center frequencies were studied : MHz MHz MHz MHz MHz Attention should be paid to the section 3 of the documents where several comments are given on some of the key parameters and on some data which had been provided to GNSS-P by RTCA. 2 Description of the methodology used for the simulation: The methodology is described in annex 1. 3 Further comments on the criteria of interference The 5.8dB criteria (in acquisition mode). All simulations on interference from DME/TACAN to a RNSS receiver provided at ITU meeting were performed with the value of 5.8 db as a maximum possible degradation of S/N. The table 1 gives the way used to calculate this value which is based on the value of S/(No+(Io ) = 33.7 dbhz. However, this value of 33.7 dbhz is based on the maximum interference level for L1 WAAS of dbm/mhz (-116.5dBm/MHz to protect all receiver functions, including acquisition, adjusted by a 6 db margin [1][2]). Consequently, the 33.7 db contain already a margin and the maximum S/N degradation without margin is therefore 6.7 db instead of 5.8 db (table1). S/No criteria for the acquisition mode

2 RTCA documents are using the same S/No criteria for L1 narrow band signal and L5 wide band signal. However, it is understood that the cross-correlation properties of the L5 codes will be better than for the codes used in L1. This should normally be translated in a better S/No criteria for L5 acquisition mode. The 20% maximum of blanking time This criteria is mentioned on ITU-R WP 8D US input document and according to RTCA, is due to a possible degradation of performance of the Viterbi decoder in presence of higher duty cycle. However the need for this criteria and the relation between blanking time and Viterbi decoder performance is not clear The blanking Threshold level In RTCA methodology the value of dbw was used for the blanking which is 4.84 db higher than the maximum level of interference if there is no blanking (Io). This is not well understood having better performance regarding S/N degradation if the blanking threshold is equal to the maximum level of interference (Io). Is there a reason for that choice? The impact of the receiver temperature. In the table 1, it is also important to note that the S/N maximum degradation is based on a the receiver temperature which is K higher than the temperature for L1 receiver. This is due to the anticipated greater filter insertion loss resulting from the more stringent filtering requirement (RTCA meeting). However this assumption might not be necessary for receiver using narrow band filter with less stringent requirement (equal to L1) and the maximum S/N degradation without margin must be, in this case, 8.2 db. The noise for the wide band at which is narrower (±8MHz), could also be expected to be slightly lower than 626 K For wide and narrow band with higher requirement than in L1 the assumption still need to be clarified. The criteria for the tracking mode It could also be noted that the maximum degradation S/N without margin in traking mode (the most important mode in term of % of time compare to the acquisition mode) is 9 db. Margin for other sources of interference It is noted that, in accordance with the formula expressed in section of the annex 1, the S/N degradation due to DME is function of the noise level but also of the noise due to other source of interference. This means that the margin for other sources of interference is not simply equal to the margin calculated in section 4 and that it would depend on the nature of the other sources of interference.

3 Table 1 L1 WAAS L5 or a part of MHz Comments L5 or a part of MHz with same noise T as in L1 (no increase due to additional insertion loss (for narrow band signal) Comments Rcvd signal dbw Power Implementation db losses minimum antenna db gain Sky temperature K Receiver Excess K Higher for L5 (reason on Temp the RTCA doc: Greater filter insertion loss) Noise Floor dbw/hz Minimum S/No db-hz S/(No+Io) db-hz S/N degradationcriteria S/N degradation without margin 9 db Criteria 10.5 db S/(No+Io) with db-hz margin Safety margin on 5.50 db Io Io without margin dbw/hz Io with margin dbw/hz S/(No+Io) without margin S/(No+Io) with margin db-hz S/N degradation criteria 6.7 db ( ) Without margin db-hz used to define the criteria of 5.8 db with margin Safety margin on 6.47 db Io Io without margin dbw/hz Io with margin dbw/hz S/N degradation criteria = 8.2dB Without margin mode Tracking Mode Acquisition mode

4 4 Results 4.1 Acquisition mode Wide band RNSS signal Altitude and Frequency impact (aircraft at the horizontal position) For an aircraft flying at the horizontal position at Ft, the criteria of 6.7dB is met for the frequency and supposed for MHz. However the margin for interference from other interferer than DME/TACAN will be very low. At Ft the criteria is met for all frequencies. Altitude Fc= MHz Fc= MHz S/N Margin S/N Margin Degradation db Degradatio db n Ft db 0.03 Ft db 0.48 Ft db 1.1 Ft db 1.28 Ft db 2.25 Ft Fc= MHz S/N degradatio n Margin db Fc= MHz S/N degradatio n Margin db 6.58 db db db db Frequency impact (worst case : aircraft with a roll angle of 33 towards the worst direction) The worst frequency is MHz and this is due to the repartition of DME/TACAN ( figure in annexe 2). Frequency S/N Degradation Margin db MHz 8 (6.7-8) = MHz 6.87 ( ) = MHz (signal at 8MHz) 6.1 (6.7*-6.1) = 0.6 *note : the available margin might be higher due to lower filter insertion loss.

5 4.1.2 Narrow band RNSS signal The Results at Ft (worst case : aircraft with a roll angle of 33 towards the worst direction) show that the criteria is always met with large margin for interference from other interferer than DMETACAN. Two cases have been considered A narrow band filter with the same slope as the wideband filter (5.5 db/mhz) and the same noise temperature as in L1 (412.9 K, corresponding to a margin of 8.2 db) A filter with the same order of complexity than the wideband filter (55 db/mhz) and the same noise temperature L1 (626 K, corresponding to a margin of 6.7 db). Filter with a slope of 5.5 db/mhz Filter with a slope of 55 db/mhz Frequency S/N Degradation Margin db S/N Degradation Margin db MHz 3.17 db ( db ( )= )= MHz 2.8 db ( )= db ( )=4.55 Therefore, the solution consisting in a simple filter is better in terms of margin degradation due to DME. It can also be questioned whether any blanker feature would be really necessary for narrow band signal, since RNSS receiver would be intrinsically able to cope with high level short duration interference owing to saturation of the amplifier and automatic gain control. 4.2 Tracking mode In tracking mode the maximum tolerable S/N degradation is 9dB and the precedent section shows that the worst case for wide band is 8dB of degradation at the frequency MHz. In tracking mode the wide band receiver will then have a minimum of 1dB margin. In the case of frequency at MHz the margin will be 2.13 db. 5. Conclusion The simulation results above enable to determine, depending on the actual frequency, the relation between the operational altitudes and the available margins. It also shows the benefit of narrow band operation to increase the margin available at all altitudes. Some of the assumptions in the simulation, such as number of DME pair of pulses or DME antenna pattern, are particularly critical and could be refined at a later stage. These results will have to be validated by measurement of the receiver and/or simulation of the full receiving RF front end.

6 Civil Aviation community needs to study the : operational altitude at which the precision given by the wideband signal would be necessary margin to be kept for other interferences. References [1] Change 3 to RTCA/DO-229 Minimum Operational Performance Standards for Global Positioning System/wide area Augmentation system Airborne Equipment, RTCA Paper No /SCI59-781, March 31, [2] Draft new Recommendation ITU-R [RNSS.CHAR],Technical Characteristics of current and Prospective RNSS (Space to Earth) and ARNSS Receivers to be Considered in interference Studies in the band MHz, ITU-R Working Party 8D, document 8D/TEMP/145(Rev2) April 27, [3] ITU 8D/262, USA (bandwidth requirements for an additional RNSS allocation).

7 ANNEX 1 1 Description of the methodology. RNSS signal and receiver characteristics are taken from documents provided by the USA in RTCA meetings for the description of GPS L5. These models have not been fully validated through experience. 1.1 Interference scenario. In this document, we consider only the interference from ground based DME/TACAN, indicated in [1] as the main source of interference Ground based DME/TACAN characteristics. The region of the world having the densest DME concentration is Europe. An ICAO DME/TACAN data base which have the radiated power, frequency, latitude and longitude is used for the simulation. All present and planned DME/TACAN over Europe are taken into account. The DME/TACAN pulses coming from different emitters arrive randomly at the receiver. For simulation purposes, all DME/TACAN are set at 2700 pulse pairs per second. This value is a maximum for almost every DME, except for a few British DME which maximum will be 5400 pulse pairs, and for TACAN where it should rather be 3600 pulse pairs. In first approximation, the instantaneous received power of a perfectly Gaussian shaped DME/TACAN pulse pair separated by 12 micro second can be written as : p( t) = p( e -a t 2 + e -a ( t ) 2 ) Where : p is the received radiated power of the undesired signal from one DME/TACAN pulse. α= p(t) is set at 0 W when the receiver is blanked, taking into account 1µs of blanking recovery time. Typical DME/TACAN antenna gain pattern is shown in Figure1 and most (around 87.5 percent) of radiated power is contained within a 0.5MHz bandwidth centred on the 1MHz channels. The envelope used in the simulation is also shown in Figure 1.

8 O corresponds to the horizontal direction Figure 1 :Typical DME/TACAN antenna Gain vs. Elevation angle RNSS receivers characteristics assumptions. The receiver filter Response used in the simulation are: -For the wide band filter (5.5 db/mhz): (10 MHz,0 db) (22.7 Mhz,-70 db) Fc in MHz F in MHz -For the wide band filter (±8Mz) at frequency MHz (5.5 db/mhz): (8 MHz,0 db) (20.7 Mhz,-70 db) Fc in MHz F in MHz

9 -For the narrow band filter with the same complexity (same order and same temperature) than the wide band filter (55 db/mhz): (1 MHz,0 db) (2.27 Mhz,-70 db) Fc in MHz F in MHz -For the narrow band filter with less complexity (lower order, same temperature than L1 receiver and less expensive) (5.5 db/mhz): (1 MHz,0 db) (13.7 Mhz,-70 db) Fc in MHz F in MHz The RNSS receiver antenna gain in satellite direction is in worst case -4.5dB. For simulation purpose, only that value is considered. The RNSS receiver antenna gain in the direction of DME/TACAN is taken into consideration as follows : -The assumption used when the aircraft is flying in a horizontal position is 10dB for both LCP and RCP (antenna gain in lower hemisphere [2]) (figure 3). The assumption of a same gain for RCP and LCP has not been assessed by previous studies and must be considered with great caution.

10 RNSS receiver Gain at zenith 7dB 5 elevation Gain towards RNSS satellite : -4.5dB Gain towards DME: 10dB EARTH Figure 3 : RNSS receiver gain assumption (aircraft flying in a horizontal position) -The RNSS receiver antenna radiated pattern used for simulations with an aircraft having a roll angle of 33 is taken from 8D/142-E (equation 8): G(Φ) = x10-3 Φ 2.01 G(Φ) = Φ x10-4 Φ 2 for 0<Φ<80 for Φ>80 The gaseous attenuation and the diffraction attenuation are taken into account. All DME/TACAN ground transmitters are considered, up to the electromagnetic horizon (4130 h with h in m) RNSS receiver blanking threshold. The maximum tolerable interference power spectral density Io (blanking threshold/filter bandwidth-6db) can be determined using : No (noise floor -200dBW/Hz), Minimum S/N (33.7dB-Hz ), Minimum wanted received power signal S in 20MHZ =-160.5dBW (-154dBW (Received power) -2dB (implementation losses) 4.5dBic (antenna gain) ) N Io = ( N No) = (. S No) = 10.log(10 10 S ) = 194.3dBW / Hz

11 The Blanking Threshold (Io in 20MHz +6 db = Io + 73dB db) is therefore : = dBW. In some of our simulations, different values are studied for the blanking threshold. 1.2 Methodology to evaluate interference from ground based DME/TACAN to RNSS receivers Simulation. In order to evaluate the interference from ground based DME/TACAN to RNSS receivers, a simulation tool was created, taking into account the characteristics mentioned above. A simulation round consists in simulating the S/N and blanking percentage of an airborne receiver at a specified altitude, in each position over Europe sampled every 0,5 degrees of latitude and 0,5 degree of longitude. The tool provides the worst figure found for all these simulations Criteria of interference. - A maximum S/N degradation of: 6.7 db in case of receiver using wide band or narrow band filter having the same complexity (same order) than the wide band filter. 8.2 db in case of receiver using narrow band (±1MHz) filter having less complexity than the wide band filter Determination of S/N degradation. The S/N degradation margin is defined as the comparison between two values (RTCA meeting): The first one is the maximum tolerable S/N degradation, using N=Io+No where Io (in dbw/hz) is the maximum tolerable external interference power spectral 6 6 S No Io No Without erference S = = int with. max. tolerable. int erference density : 6.7dB The second one is the S/N degradation related to I 1 (in W) which is the simulated average received external interference (from DME/TACAN) power taking into account the blanking and the received filter:

12 S No Without.interference 6 1 No PDC. S.(1 PDCB ) B I 1 Withsimulated..interference = No 1 n n 6 No p( t) dt (1 PDC B) p S + + To(1 PDC ) B To i 1 To = To i= 1. = S.(1 PDC B) (1 PDCB 1 1 ( t) dt No ) 2 Where : n is the number of DME/TACAN in visibility (distance between DME/TACAN and RNSS receiver < horizon). i is one of the DME among the n in visibility. To is the Gaussian pulse pair period. PDC B is the percentage of time when the RNSS receiver is blocked. p(t) has been defined above, with p as follows : p= DME/TACAN radiated power + DME/TACAN antenna gain (θ) + RNSS antenna gain (ϕ) free space loss (f,d) + RNSS filter (f) - gaseous attenuation and the diffraction attenuation.

13 ANNEXE 2 Frequency distribution of the number of DME/TACAN in Europe as seen in a sliding window of +/ MHz

14 14