Single Entry Versus Aggregate MLS Interference Susceptibility Limits

1

ACP-WGF16/WP-14
/
International Civil Aviation Organization
WORKING PAPER / ACP-WGF16- /WP-14
04/12/06

AERONAUTICAL COMMUNICATIONS PANEL (ACP)

SIXTTEENTH MEETING OF WORKING GROUP F

Montreal , 11-15 December 2006

Single entry versus aggregate MLS interference susceptibility limits

(Prepared byAlain Delrieu)

SUMMARY
ICAO experts are studying the use of the 5 GHz bandbythe AMS/AMT, AM(R)S and AMS/AS applications and conditions under which the proposed services can share the MLS band.
This paper assesses suitable in-band interference susceptibility limits needed for MLS protection, for applicability to both single entry and aggregate sharing situations, taking into account existing services already defined in the R.R. and new services being considered under WRC’07 Agenda Item 1.5 and 1.6.

1.INTRODUCTION

During the latest ICAO/NSP meeting held in October, the level– 150 dBW/150 kHz was considered (ref NSPOct06 WGW WP20_MLS Interference issues _rev1)as the maximum aggregate power ensuring the protection of the MLSairborne receiver. This value represents an I/N ratio of – 14 dB as the appropriate criterion protectionagainst in-band MLSinterference, derived by considering that the allowable course motion noise (CMN) error limit is defined in probabilisticterms, i.e. a givenlimit such as0.1° for low desired MLS signalsnot to be exceeded more than 5% of time .

The analysis performed in that reference is based on anS / (N+I) approachwhere N+I is total noise, comprising both thermal noise andman-made noise to ensure MLS protection against all source of noises: thermal,MLS adjacent channel and external sources such asexisting FSS and RNSS or being considered for WRC’07 preparation, i.e. AMS/AMT and AM(R)S. This S / (N+I) approach has been proposed as a replacement to the previousICAO interference assessment methodology which was based on a signal to interference ratio (SIR)approach, and as such considered ill justified because interference related CMN errors cannot be realistically observed separately from that due to total noise; and furthermore that SIR approach might be too constraining for the introductionof the newaeronautical servicesby WRC’07 under agenda items 1.5 and 1.6.

Since then it was realized :

a)the SNR requirements numerical values derivation, reconfirmed underthat Amendment 81, indicate that the minimum desired MLS power density requirements defined in the SARPS are 0,5dB higher that needed,

b) This 0,5 dB margin can beused as part of the S(N+I)R methodology proposed in that reference,to derive a permissible I/N ratio of -9.1 dB, while still maintaining the usual 6dB margin in linewiththe aeronauticalflight safety objective,

c) there is a need to factor in the MLS interference limit analysis,the worst case 6 dB antenna gain ratio between undesired versus desired signals reception as customarily done for protection of systems equipped with omni directional antennas.

Accordingly this documentreappraises a suitable aggregate MLS interference limit with the objective to assist in the definition of MLS band sharingconditions with both existing, i.e. prior to WRC’07, and future services to be considered by the WRC’07.

2.CURRENT REGULATORY FRAMEWORK GOVERNING THE 5 GHZ ARNS BANDAND MLS INTERFERENCE PROTECTION LIMIT

The current regulatory framework comprises:

• R.R. footnote 5367, which grantsan AMS(R)S allocation, subjectto article 9.21,
• R.R. footnote 5.443B , which grants MLS protection against interference from RNSS operating space-to-earth in the adjacent band 5010-5030 MHz . In practise it is expected that a maximum of two RNSS systems are likely to be operated in that band , each emitting spurious into the MLS band, at a level not exceeding -124.5 dBW/m2ina 150 KHz band,
• R.R. footnote 5444which givesMLS an “super primary” status over all other use in the band 5030-5150 MHz ,
• R.R. footnote 5.444Awhich defines conditions under whichFSSground station earth-to-space feederlinksshare the MLS extension bandto ensure MLS protection,
• Recommendation ITURS.1342 , which describes a method for determining coordination distances between MLS stations operating in the band 5030-5091 MHz and FSS earth stations providing Earth-to-space feeder links in the band 5091-5150MHz;
• Recommendation ITU-R M. 1582, which describes a method for determining coordination distances between MLS stations operating in the band 5030-5150 MHz and RNSS Earth-to-space stations in the 5000-5 010 MHz band,

3.Interference Susceptibility LIMIT ISSUE: SiNGLE ENTRY vs AGGREGATE MLS INTERFERENCE SUCEPTIBILITY

From a close examination of the above it can be deducted:

a)Article5.443B, ITU-R Rec S 1342 and M 1582 all implicitly define the same in-band MLS interference susceptibility limit of -160 dBW/ 150 Khzto protect MLS from the consideredservice they each relate to, even thoughthey employ different approaches. Incidentally article 5.443B does so explicitlyby specifying an PFD limit of–124.5dB(W/m2) , whereas this value limit is to be obtained by inspectionof equation 2 of Annex 1 to bothquoted recommendations

b) Applying thethusidentified limit of -160 dBW/150 KHzin the same band, 5030-5091 MHz , the so-called MLS core band, tothreesources of interference, i.e. (a)FSS in the band 5091-5250 MHz, (b)two RNSS systems operating space-to-earth in the band 5010-5030 MHz and (c) RNSS earth-to-space in the band 5000-5010 MHz , implies thatin the eyes of the ITU-Rradio regulatorsthis limit is to be considered as asingle entrylimit per service, except for the S-to-E RNSS for which this limit is specified per system.

Note : For the purpose ofdefining conditions under which services FSS and RNSS would both have tocoordinate with MLS,the terms “out-of-band” and “inband” are relative to the MLS band 5 030-5 091 MHz. A further consideration forthe protection a givenMLS channel would be distinguish between co-channel interference and“out of channel” interference,witha possibility for the latter to fall “in-band”

ICAO work which culminated into input into ITU-R for WRC’97 preparation process and in particular to the establishment of ITU-R Rec S. 1342to ensure MLS protection against FSS , implicitly assumed at that timethat such a limit would need tobe interpreted in the aggregate senseand not as a single entry.

This ICAO understanding is not in line with the ITU-R radio regulators understanding as deducted from the review of the current regulatory context described in section 2 above . Additionally , if this value of -160 dBW/150 Khz was to be retained as an aggregate one, close examination of reference [ Doc 4A/Temp84-E , ITU-R study group working party 4A meeting in Rio de Janeiro, 26 Sept to 4 oct 96] reveals that this aggregatein-band limit is unnecessarily conservative by around 9 dB, with the risk of further constraining the introduction and use of new aeronautical services such as AM(R)S and AMS/AMT in the band 5091-50150 MHz .

4.Justification for a reviseDaggregate in-band MLS susceptibility limit of -151 dBW/150KHZ

a) The following figure illustrates the RF schematics of the airborne MLS receiverand the signal and noise levels referred to both the receiver inputs and the aircraft antenna output port , with an assumed 5 dB loss interconnecting cable in between:

Fig. 1 : Airborne MLSantenna receiver RF schematics

b) As explained in NSP Oct 06 WGW/WP 20, section 5, the derivation of the ITU-R Rec S.1342 -160 dBW/150 Khz limit figure is conjectured to have been established in the following manner :

• assume a worst case permissible CMN error of 0.005° (1 sigma) due to MLS external interference
• useMLS guidance material,Att. G , section2.6.1.2 formula to derive a signal to interfering noise ratio of 33 dB assuming150 Khz of IF bandwidth and a scanning 3°antenna beamwidthcase with high rate of angle scanning (see appendix of that paper for details)
• assume MLS equipped aircraft at edge of promulgated MLS coverage and receivingthe lowest specified power (flux) density of SARPS 3.11.4.10.1 for thatcase corresponding to an S_min at aircraft antenna of -86.5 dBm
• assume a+ 6 dB antenna gain towards the undesired interfering signal compared to an 0dBi towards the desired signal

Accordingly the sought interference MLS external system interference (ESI) limit, IESI,is:

IESI = -86.5 (S_min) – 6 (ant. gain) – 5 (cable loss in dB between antenna and receiver) -33 (SNR for 0.005°) = -130.5 dBm , or-160.5 dBW, rounded up to an -160 dBWvalue

However such a derivation is overly pessimistic:

i)the SNR of 33 dB is actually a signal to interference ratio (SIR), which doesnot exist inreality since: (1) thermal noiseis always present ,(2) the receiver is not able to process thereceived interference independently from the thermal noise background; the latter one iscalculated to be N= -106 dBm at airborneMLS antenna (see abovefigure 1) which is a much higherthan consideredESI. Accordingly an 33dBSIR leading to an interference levelof -130 dBm would be totally buried in the receiverthermal noise and as suchcannotbe justified as a separate interference limitdistinct for the noise-related one, since the CMN error is dominated by the -106 dBmthermal noiseand one cannot observe nor measure the ESI induced CMN error independently from the CMN error due to thermal noise .

ii)thereceived signal to noiseratio being evaluated at the aircraft antenna port, and considering that the S_min value of -86.5 dBm value, also quoted in Chapter 3Att. G table G-2 , already incorporates a 5 dB interconnecting cable loss value; thus subtracting another 5 dB in the above I_ESI equation is not justified as itwould signify adouble accountingof that 5 dB factor;moreover the desired signal to interference/noise ratio canbe evaluateddirectly at the antenna output port, since the 5 dB cable loss affects both the desiredand the interfering signals the same way.

c) Another consideration is that theSNR requirements numerical values derivation, which has been reconfirmed in SARPS 3.11.6.1.4 , under ICAO Annex 10 Amendment 81 ( coming into force this year),points to a discrepancy of 0,5 dBwith those ofthe MLS power (flux) density figures of “existing” SARPS 3.11.4.10.1 See Annex 1, table A-1 for derivations details and discrepancy identification. As the result, SARPS 3.11.4.10.1defined S_minvalues, are 0,5 dB higher than actually needed to meet the SNR requirementsthus reconfirmed by Amendment 81.

Note : although the heading of this SARPS3.11.6.1.4is “adjacent channel spurious response”the SNR values derivationtherein is strictly related to thethermal noise impact on CMN errors as explained by thereference NSP WGW_WP 48 ( St Petersburg, May 25 to June 4 2004) which was developed to validate the SARPS change brought bythe Annex 10 81st amendment . It was further validated experimentallythat these thermal noise related SNR specifications could also be used to protect desired MLS signalsfromthe spurious response due to adjacent channel emissions, as explained by thereference NSP WGW_WP 50 from the same St Petersburg meeting.

d) A more realisticinterference limit derivation can be preformed by employingthe signal-to-total noise-ratio,S(N+I)R asstated inref.[ NSPOct06 WGWWP20_MLS Interference issues ] andallowing the total noise to grow fromN in absence of interference, toN+I,such that the ratio (N+I)/Nincreases by 0.5 dB,as it can besimply demonstrated here-below .

Let us assume :

i) S_min the minimum MLSsignal levelsderived from 3.11.6.1.4 SNR requirements , table X2 in particular for low desired MLS signals,
ii) S_SARPS the min signal levels associated withSARPS 3.11.4.10.1 power density (PD) values .

As an illustrative example one can quotethe3° antenna beamwidth at the high ratescanningcase:the PFD value is -81 dBW/m2 corresponding to an S_SARPSsignal value at the aircraft antenna of -86.5 dBm,quoted inChapter 3 Att. G Table G-2 whereas the actual neededS_min valueis -87 dBm , as demonstrated inthe Annex 1 table 1 to this paper.

Accordingly :S_SARPS/S_min=10 0.5/10

Then : = 10 (SARPS 3.11.6.1.4 SNR)/10 = .

Hence :

therefore1+ I/N = 10 0.5/10 , i.e. I/N(dB) = -9.1 dB and with N = -106 dBm at aircraft antenna level the IESIlimit is -115.1 dBm to be approximated as -115 dBm (or -145 dBW)

Since allin-band interference are assumed to fallwithin the receiver IF bandwidth of 150 KHz , and furthermore as customarily done for safety applications, one should take into account that ESI can be received througha 6 dB higher MLS antenna gain than the 0 dBi assumed for the desired MLS signal reception,itis herein suggested to adopt the -121 dBm or -151 dBW as the aggregate in-band co-channelinterference limit, over 150 KHz

Note :the desired S_min levels , as defined in SARPS 3.11.4.10.1 power densitiesare consistent with Att. G Table G-2 “signal required at aircraft” figures, whichtable shows that a 6 dB (aeronautical) margin have been incorporated in those minimum signal levelsderivation. Asthe result the combination of the 6 dB margin on S_min and the 6 dBi assumptionforantenna gain towards multiple interference sources ,irrespective of their incoming directions, make the -151 dBW/150 KHzlimit still very conservative andin line with aeronautical safety objectives associated with ARNS/MLS[1].

5.Conclusion

WG F members are kindly invited:

• under existing ITU-R regulatory provisions to recognize that the -160 dBW/150 KHzlimit for MLS protection is a de facto single entry service limit for all services considered in the band 5030-5150 MHz except for RNSSauthorized to operate E-to-S in the band 5010-3030 MHz for which it is a-per-system limit,
• to concur that a realistic aggregate limit is -151 dBW/150KHz to be applicable to allservices considered in the ARNS 5030-5150 MHz band for MLS interference protection .

ANNEX 1: COMPARISON BETWEEN THE MINIMUM NEEDED MLS POWER DENSITIES USING ICAO AMENDMENT 81 SNR SPECIFICATIONS AND THOSE DEFINED IN SARPS 3.11.4.10.1

The minimum powerdensities values quotedin the following table arederived from the Amendment 81 SNR values usingthe methodology explained in NSP WGW Oct’06 WP 20 Annex 1, reproduced hereafter for convenience:

Table A-1 / Annex 10Amendment 81 "new "MLS min PD valuesvs establishedper SARPS 3.11.4.10.1
Item / DPSK / Az BW 1°low rate / Az BW 2° low rate / Az BW 3° low rate / Az BW 3°, High rate / El. BW 2° high rate / Comments
IF SNR (DB) required / 5 / 8,3 / 14,3 / 17,8 / 13 / 9,5 / High Rate
Noise power in 150 KHz (dbm) / -122 / -122 / -122 / -122 / -122 / -122 / Att G; Table G-2
Noise figure (dB) / 11 / 11 / 11 / 11 / 11 / 11 / Att G; Table G-2
Cable loss (dB) / 5 / 5 / 5 / 5 / 5 / 5 / Att G; Table G-2
Airboren Ant. Gain (dBi) / 0 / 0 / 0 / 0 / 0 / 0 / Att G; Table G-2
(Aeronautical) margin (dB) / 6 / 6 / 6 / 6 / 6 / 6 / Att G; Table G-2
Signalat a/c ant. Port (dBm) / -95 / -91,7 / -85,7 / -82,2 / -87 / -90,5 / Sum of above
dbW/m2 to dBm Conv. factor / 5,5 / 5,5 / 5,5 / 5,5 / 5,5 / 5,5 / -35,5(Omni area at 5 GHZ) -30 (dBW => dBm)
Req'd power density (dBW/m2) / -89,5 / -86,2 / -80,2 / -76,7 / -81,5 / -85 / Amendment 81 values
Existing SARPS(dBW/m2) / -89,5 / -85,7 / -79,7 / -76,2 / -81 / -84,5 / see3.11.4.10.1
Difference(dB) / 0 / -0,5 / -0,5 / -0,5 / -0,5 / -0,5

APPENDIX : MLSSNR requirementsderivation(extracts from NSP WGW (Oct’06) WP-20, Annex 1 )

a) CMN error limit formula:

Following guidance from the ICAO Annex 10 VOL 1 Chapter 3 and its Attachment as well as ITU-R4A/Temp/ 84-E [ITU-R study group working party 4A meeting in Rio de Janeiro, 26 Sept to 4 oct 96] the airborne MLSoutput to the aircraft flight controls is assessed bythe path following error (PFE) and control motion noise (CMN) error. The later one is measured an angular error. It isconsidered by Doc 4A/TEMP/84-E [ref.6],the most critical parameter, as “…it does affect pilot acceptance of the systemas well as mechanical wear in the actuators and control surfaces of the aircraft”. considered to better quantify the receiver performance can becalculated as follows :

(Eqn 1)

where:

 is the 3 dB beamwidth of the MLS scanning antenna; itsvaluesas quoted in tables G-1 and G-2of [ref. 2]arerespectively1°, 2° or 3° for the azimuth and 1° or 2° for the elevation scanning functions,

SNR is the signal to noise ratio at the output of the 26 KHz low-pass filter contained in the MLSreceiver processor,

g is the ratio where FSR is either equal to 39 Hz for high rate approach azimuth scanningor 13 Hz otherwise,

the noise bandwidth of thereceiver output filter is calculated , following Vol 1 Chapter 3 Att G. 2.6.1.2 guidance, as /2 times the 3 dB bandwidth of the single pole filter represented in fig G-11, which has a corner frequency of 10 radians/sec.; accordingly this Filter_noise_bandwith is 2.5 Hz

b) SNR requirements derivation

Equation 1 can be rewritten as :

(Eqn 2)

Usingthis equation this annexprovides the derivation of the SNR requirements found in MLS SARPS 3.11.6.1.4andAtt. Gtable G-2

The ITU-R4A/Temp/ 84-E in its annex 2 quotesan valuefor an error allowanceof 0.025° or less , assuming of 3° and a sampling rate of 39 Hz , as 26.5 dB ; note that the actual computation yields26,6 dB.Following that reference guidanceit is worth noting that the 0.025° error is considered as theone-sigma valuefor the error itself assumed to be normally (i.e. Gaussian) distributed with zero mean.

Equation 2 can be applied to derive changes in SNR as an SNR2/ SNR1 ratio as a function of changes in the other parameters:

(Eqn 3)

Accordinglythe followingSNR step-changes values can be calculated:

a)4.8 dB for a sampling rate change from 39to 13 Hz

b)-3.5 and -6 dB for going from respectively 3° to 2° and 2° to 1°

c)16.5 dB forchanging from0.1° (2sigma)to 0,015° (2 sigma)and
20 dB forchanging from 0.1°(2 sigma) to 0.01° (2sigma)

d)conversely an increase of 6 dB in SNR results in a reduction by ½

e)In addition , takinginto accountthe guidance of Att. G2.6.2.2 which defines the SNR specification in the receiver IF bandwidthof 150 Khz:
SNR (IF) = SNR(video) – 10 log = SNR(video) -7,6 dB

The following SNR specification values can subsequently bederived using equation 3and step-changes calculated above:

Table A-2 : MLS SNR incremental changes and SARPS specifications derivation

1) Low MLSsignalswith 0.1° CMN, 95%, requirement / Comments /source
Ground antenna Beam width in ° / 3.0 / 2.0 / 1.0 / as specified in the MLS SARPS (1)
SNR for 0,025°, 39 Hz and 3°BW, in dB / 26.6 / quoted in [Ref 6] (2)
SNR for 150 KHz IF / 19.0 / change (e):going from 26 to 150Khz (3)
SNRfor dθ (one-sigma) of 0,05° / 13.0 / change (d): from 0.025 to 0.05° (4)
SNR requirements per SARPS 3.11.6.1.4, table X2 , for CMN= 0,1° (2 sigma), in dB:
Approach Azimuth Guidance / 17.8 / 14.3 / 8.3 / changes (a):39 to13 Hz & (b):in(5)
High Rate approach az. guidance / 13.0 / 9.5 / 3.5 / changes (a):13 to39 Hz & (b) in(6)
Approach elevation guidance / N/A / 9.5 / 3.5 / no change : high rate assumed (7)
2°)High MLS signals with requirements of .015° in Az. and .01°( both 2 sigma)in El. , 95%,
SNR requirements per SARPS 3.11.6.1.4, table X1 , in dB:
Approach Az. Guidance (0.015° CMN) / 34.3 / 30.8 / 24.8 / changes (c):.1 to .015°& (b):in(8)
High Rate approach az. guidance (.015°) / 29.5 / 26.0 / 20.0 / changes (a):13 to 39 Hz & (b):in (9)
Approach Elevation guidance SNR (0.01°) / N/A / 29.5 / 23.5 / starting from line (7) :
changes (c): from .1 to ,01°& (b):in

Notes: 1°) It can easily be verified that these SNR computationsmatch those specified per the (new) SARPS 3.11.6.1.4 Tables X2 and X1, with at most0.1 dB discrepancies in line 8, col. 2° and line 9 col.1°

2°) There is an 0,5 dB discrepancy betweenthe SARPS 3.11.6.1.4 Tables X2 values for the 1°,2° and 3°cases and those of Att G table G-2, : 17.8 vs 18,3 dB for 3°, 14.3 vs 14.8 for 2°, 8.3 vs 8.8 for 1, suchthat the Att G SNR figures are systematically 0,5 dB more demanding thanthosecalculated above and specified in the “new” SARPS of Annex 10 81st amendment

[1]As an illustrative example, by taking the above 3° BW high-rate scanning case with 13 dB SNR and 0,5 dB higher min MLS power density than needed, one can calculate using equation 3 of the Appendix 1 to this paper that the CMN error varies between 0.47° and 0.94° in absence of ESI depending whether the 6dB aeronautical margin of Att. G, table G-2 is available totally- which results into a 6 dB higher SNR- or not at all, and in presence of ESI, in the range 0,05° to 0,1°