Definition of the MLS Aggregate In-Band Interference Protection Limit Based on Recent Analytical

MLS interference susceptibility analysis and testing Aug 2007

/
International Civil Aviation Organization
WORKING PAPER / ACP-WGF17- /WP15
17/09/2007

AERONAUTICAL COMMUNICATIONS PANEL (ACP)

SEVENTEENTH MEETING OF WORKING GROUP F

Nairobi 19-25 –September 2007

Agenda Items 1 & 4 related to WRC 20 07, and specifically A.I 1.5 and 1.6 :

Presented by A. DELRIEU, (DGAC/DSNA )
Prepared by A. DELRIEU with the assistance of C. DEHAYNAIN (DGAC/DSNA) , K. CHASSANIOL, ( AIRBUS) G. LEVI , THALES for the MLS tests

Definition of the MLS aggregate in-band interference protection limit based on recent analytical and test results

As part of the current WRC 2007 preparation , sharing studies of new spectrum allocations in the band 5091-5150 MHz under agenda item 1.5 and 1.6- for AMS aeronautical mobile telemetry, aeronautical security and ground AM(R)S applications, have prompted the need to reassess the in-band MLS interference protection value of -160 dBW/150 KHz, implicitly quoted in ITU-R Rec. S.1342 adopted by WRC 1997 and to investigate whether it should be taken as an aggregate or single entry value

The most note-worthy developments since WRC 1997 can be summarized as follows:

·  about four 4 years ago the need arose to better protect MLS against its own internal MLS emissions originating from adjacent channel transmissions. Studies and testing done within ICAO concluded on the need to apply signal to noise ratio (SNR) requirements to the adjacent channel “noise. On that occasion the MLS SNR requirements were reassessed. They have been found 0,5 dB more demanding than strictly required by the SARPS minimum signal specifications,

·  more recently as part of the current WRC 2007 preparation , sharing studies of new spectrum allocations in the band 5091-5150 MHz under agenda item 1.5 and 1.6- for AMS aeronautical mobile telemetry, aeronautical security and ground AM(R)S applications, have prompted the need to reassess this limit value and to investigate whether it should be taken as aggregate or single entry ;

·  within ICAO the current interference limit of -160 dBW/150 KHz, interpreted as an aggregate limit, has been subjected to extensive questioning. For some it appears too conservative, since it corresponds to a level some 24 dB below thermal noise, thus it is not detectable nor observable in practice; one administration presented a theoretical justification for relaxing this limit by about 10 dB, but the majority view has been to wait for test results as a prerequisite before acceptance could be considered.

This contribution brings forth the requested test results. It offers further analysis detailed in its annex. It shows that the current limit taken as aggregate can be relaxed by 15 dB to -145 dBW/150 KHz . This is obtained by revisiting assumptions behind MLS link budgets calculations and in particular the accounting of safety margins to ensure strict adherence to the usual 6 dB safety margin factor commonly used within ICAO and ITU for safety of life applications and eliminating double accounting or unjustified margins. . The tests were performed on ICAO-standard MLS receiver . They confirm the suitability limit of -145 dBW, at aircraft level, with an 8 dB margin , even in presence of MLS “noise” in the form of adjacent channel emissions being received in-band within the receiver 150 KHz bandwidth, including for a worst case ICAO compliant receiver with 11 dB noise figure.

Analysis and test report on the determination of a revised interference limit to protect MLS against in-band interference from sources other than aeronautical radio navigation systems

1.  Introduction and summary

The context within ICAO and ITU- R on this subject matter can be summarized as follows :

-  As part of the current WRC 2007 preparation , sharing studies of new spectrum allocations in the band 5091-5150 MHz under agenda item 1.5 and 1.6- for AMS aeronautical mobile telemetry, aeronautical security and ground AM(R)S applications, have prompted the need to reassess the in-band MLS interference protection value of -160 dBW/150 KHz, implicitly quoted in ITU-R Rec. S.1342 and to investigate whether it should be taken as an aggregate or single entry value.

-  About four 4 years ago the need arose to better protect MLS against its own internal MLS emissions originating from adjacent channel transmissions. Studies and testing done within ICAO concluded on the requirement to apply signal to noise ratio (SNR) requirements to the adjacent channel “noise”. The ICAO standards and practises (SARPS) were then revised accordingly. On that occasion the MLS SNR requirements were reassessed and have been found 0,5 dB more demanding than strictly required by the SARPS minimum signal specifications, which leaves room to trade signal power for additional noise, thermal or interfering noise, without having to change existing standards and equipment-units, on the ground and on-board aircraft

-  Within ICAO the current interference limit of -160 dBW/150 KHz has been subjected to extensive questioning, as being too conservative, since it corresponds to a level some 24 dB below thermal noise, i.e. not observable in practice. One civil aviation administration presented a theoretical justification for relaxing this limit by about 10 dB, but the majority view has been to wait for test results as a prerequisite before acceptance.

In the first part of this annex an analysis is herein proposed to show that the current limit can be relaxed by 15 dB to -145 dBW/150 KHz , by revisiting assumptions behind SNR and link budgets calculations and in particular the accounting of safety margins to ensure strict adherence to the usual 6 dB safety margin factor commonly used in ICAO and ITU fora, and eliminating double accounting or unjustified margins.

The second part reports on the testing done by one civil aviation administration in conjunction with national avionics and aircraft manufacturing industries to validate the above analytical finding. Test results have supported the acceptability of such a limit of -145 dBW, at aircraft level, with 8 dB margin . It is also shown that such a result is valid in presence of MLS “noise” in the form of received adjacent channel emissions, and for a worst case ICAO compliant receiver with 11 dB noise figure

2. Review of study work on MLS interference limit since WRC 1997

2.1 The ICAO approach at the time of WRC’97

In preparation to WRC 1997, ICAO submitted contributions to the ITU-R preparatory process aiming at defining susceptibility limits, to protect MLS from both in-and and out-band interference. This work culminated in the adoption of ITU-R Recommendation S.1342 [Ref 9] which implicitly defines two limits, for the purpose of determining coordination distances to ensure adequate separation from MSS feeder link earth stations. The first one of -91 dBW applies to the sum of out-of band emissions in the band 5091-5150 MHz received by a victim MLS receiver , the second one , of -160 dBW in 150 KHz is an in-band , relative to the MLS bandwidth, protection limit. Partial justifications for these are to be found in Ref. 7 and 8 for both. This paper however only addresses the latter and suggests to keep examination of the former as future work.

As part of the WRC 2007 preparatory on-going feasibility studies of MLS band sharing with considered new services such as AMS for aeronautical mobile telemetry and aeronautical security, as well as ground AM(R)S , under Agenda Item 1.5 and 1.6, considerable discussions took place within ICAO, CEPT/CPG and in ITU-R 8 meetings to understand how that latter in-band limit was established, since needed details are not documented in neither Ref. 7 nor 8 quoted above.

Central to those discussions has been the consideration on how the acceptable limit on one the key MLS performance parameters, designated as the Course Motion Noise (CMN) angle and chosen as the most sensitive to interference(see ref 7 , Annex 2, section 2.1), has been derived . While CMN does not contribute to deviation of the aircraft from its nominal flight path, it does affect pilot acceptance of the system as well as the mechanical wear in the aircraft actuators. The relationship between CMN – viewed as a noise-induced angular error by the aircraft flight controls - and the receiver signal to noise ratio (SNR) is quoted in The MLS GM (Guidance Material, Ref. 2), section 2.6 as :

(Eqn 1)

where:

-  is the beamwidth of the MLS scanning antenna, typically, 1,2 or 3°;

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

-  g is the ratio where FSR is either equal to 39 Hz for high rate approach azimuth and elevation scanning or 13 Hz otherwise; the noise bandwidth of the receiver output filter is calculated , following Ref. 2 guidance .

-  is interpreted as a one-sigma error associated with a 2 sigma angle limit not too be exceeded more than 5 % of the time. For low MLS received signals the 95 % limit is then 0,1°

These discussions in particular focussed on how to calculate the impact of an additional noise input N, such as due to an interference I , on CMN , i.e. what is the additional error , or (), resulting from =I, assumed to behave as broadband white Gaussian noise.

The ICAO approach used to establish the current MLS protection interference limit of – 160 dBW is conjectured , from Ref. 4, to have been carried out per the following steps:

i)  select an acceptable increase in CMN error. A criterion of 10 % of the allowed one-sigma 0,05° error defined in the MLS SARPS and GM has been selected, i.e. ≤ 0,005°

ii)  use this value in Eqn 1 by taking = 0,005° and compute the corresponding SNR , with = 3° and g as defined in MLS GM (see Ref.2,) which yields 40.6 dB; since this value is at the output of a 26 KHZ beam envelope filter , it needs to be adjusted for the receiver IF bandwidth of 150 Khz by subtracting 10 Log 150 /26 = 7.6 dB , thus SNR(IF) = 33 dB.

iii)  subtract this SNR to the minimum MLS signal corresponding to the = 3° beamwidth , i.e. -86.5 dBm (see table 1 herebelow) ; then subtract 6 dB to account for a worst case antenna gain towards the interference , which yields – 125.5 dBm; subtract an additional 5 dB to account for cable loss which finally gives -130.5 dBm , rounded to -130 dBm.

In retrospect that ICAO approach is questionable on several accounts :

1) mathematically , step ii) above calculates the interference SNR as the ratio S/, or S/I (which is signal-to-interference ratio, with = I the quantity to be determined) by “plugging” the value = 0,005° into Eqn 1 ,
i.e. 0,005 = , where SNR= SIR is the unknown.
This is erroneous as it can be easily seen by representing as a function of with C a constant function of and S, also considered as constant, i.e.

= . The differential is not equal to or as implied by step ii), but to which is a function of both the CMN error limit and the ration I/N

2) since step ii) above actually calculates a signal to interference ratio (SIR) with the erroneous relationship = it actually ignores the thermal noise N of the receiver; physically this is unacceptable since a radio receiver cannot extract a wanted signal from an interference background without thermal noise being present, i.e. an SIR cannot be considered in isolation from an S-to-(N+I) ratio, unless N becomes negligible compared to I, which is certainly not the case here.

3) from an radio-engineering standpoint, there is no reason in step iii) to subtract 5 dB for the cable loss since this step calculation refers to the aircraft antenna output , as is the chosen signal level of -86.5 dBm, i.e. at the point before the cable loss. Besides the use of a 6 dB margin to account for more antenna gain towards interference than the desired MLS, represents a double accounting since as indicated in the Table 1 herebelow there is already a 6 dB margin for aeronautical safety.

2.2 Protection against both MLS internal and extraneous interferences

Assessing the MLS receiver ability to perform in compliance with its specifications , including when subjected to interference, requires that both possible sources of interference, internal to MLS and extraneous, be taken into account.

About few years ago, ICAO on the advice of its navigation system panel experts amended the MLS standards and recommended practises (SARPS) to better define the MLS protection SNR requirements in the case where interference, N, comes from adjacent channels . The need for those SARPS enhancements is illustrated from the figure 2 extracted from Ref . 10

Fig. 1 MLS Adjacent Channel “noise” and the associated need for an SNR limit for desired channel protection

The MLS ground station successively transmits data preambles using a DPSK modulation on a nearly omni-directional antenna pattern , and scanning CW signals on narrow antenna beams of 1 to 3° beam-width; the purpose of the data preamble is to configure the receiver in different MLS function reception modes, such as runway azimuth and elevation angles determination. From an interference standpoint only the DPSK emissions on adjacent channels , with their broad spectrum exceeding the MLS channelling raster of 300 KHz, are considered to bring interference to the CW scanning beam reception.

The corresponding SNR requirement specifications, applicable to the 3rd adjacent channel and beyond, were recalculated using Eqn. 1 above and presented to ICAO/NSP panels under Ref. 11. After a test validation campaign, reported in Ref . 10, and after formal approval by ICAO, they have now been incorporated in the latest MLS SARPS issue (see ref 1) under Annex 10 amendment 81

2. 3 Consideration of interference as “coloured” versus “white” noise

The “noise” due to MLS adjacent channel emissions received in the receiver bandwidth is a combination of time-limited narrow-band spectral components of DPSK modulation and CW beam transmissions, both filtered by its front end selectivity filters, and is referred as “coloured”.