Development of frequency planning criteria for VDL Mode 4
AERONAUTICAL MOBILE COMMUNICATIONS PANEL
Working Group B
Montreal 13-17 January 2003
Development of Frequency Planning Criteria for VDL Mode 4
Status report since WGB/13
Presented by EUROCONTROL
| |
|SUMMARY |
| |
|This paper presents a summary of the activities carried out in support of WGB, by parties interested in VM4 issues following |
|the August 2002 meeting of WGB. |
| |
|Following the preliminary results of the EUROCONTROL testing of VM4 radio equipment, which were presented in the WG-B meeting |
|in August 2002, problems were identified in relation to the susceptibility of VDL Mode 4 in a ground scenario, when VDL mode 4|
|is the victim of interference (especially interference from DSB-AM). A potential solution was identified by WGB as being the |
|most likely to address the problem. The identified solution was to specify a more stringent reception performance for the VDL |
|Mode 4 radio. |
| |
|There were two teleconferences held during the period between September and October 2002 in relation to VDL Mode 4 |
|interference testing. |
| |
|A group of interested parties (involving VM4 equipment manufacturers, civil aviation administrations and EUROCONTROL) |
|investigated the feasibility and implications of this solution. Following some testing of the AM radio performance by DFS, the|
|effectiveness of the initially considered solution became questionable. |
| |
|The group had further discussions and a number of other recommendations are being considered. These concern the minimum |
|received signal strength and the distance between the interferer and the victim. |
| |
|There were no conclusive actions from this activity and the issue was left to be further considered and discussed in a future |
|WGB meeting. |
Contents
1 Background 2
1.1 ECTL Preliminary Testing Results 2
2 Activity since WGB/13 meeting 2
2.1 1st Teleconference - 30/09/02 3
2.2 2nd Teleconference - 25/010/02 3
3 Current Status and way forward 3
4 References 4
Background
The activities reported in this paper are in support of the definition of the VDL Mode 4 frequency planning criteria. A series of preliminary tests [1] were carried out in the Eurocontrol experimental facility in Brétigny, France. These tests included interference on VDL Mode 4 to and from other VHF digital and analogue systems in the same band, namely VDL Mode 2 and DSB-AM.
The tests consisted in adjacent channel rejection measurements with the aim to derive a preliminary set of protection distances compatible with the intended scenarios. The scenarios and the testing methodology are defined in the test plan agreed by WGB in March 2002 [3].
1 ECTL Preliminary Testing Results
The preliminary results indicated the possibility of feasible frequency planning for some of the scenarios considered and identified provisional frequency planning criteria for VDL Mode 4 as both interferer and victim for these scenarios. However in the case of the ground scenario the results showed that VDL Mode 4 is very susceptible to interference from other systems and particularly from DSB-AM, even at large frequency separations.
A number of solutions were envisioned to mitigate this issue and the specification of a stricter rejection requirement was accepted as being the most likely candidate solution, pending analyses and confirmation. These preliminary results were presented to WGB in the August 2002 meeting.
The WGB meeting requested the VDL Mode 4 manufacturers to evaluate the preliminary ECTL results and their implications and come with new material to WGB. EUROCONTROL proposed to organise a teleconference for interested parties to address the issue.
Activity since WGB/13 meeting
There were two teleconferences held during the period between September and October 2002 in relation to VDL Mode 4 interference testing.
A group of interested parties (involving VM4 equipment manufacturers, civil aviation administrations, EUROCONTROL and others) participated in these discussions.
1 1st Teleconference - 30/09/02
The first teleconference took place on 30th of September 2002. The main objective was to analyse the preliminary results and identify feasible solutions.
The information that was used as input for this teleconference is provided in Attachment A.
During this teleconference, the manufactures agreed that the draft proposal for ACR requirements presented by Eurocontrol in WGB as detailed in [2] seemed a feasible way forward. However a detailed feasibility analysis was required to confirm this position and to agree a way forward.
An important issue raised was the lack of explicit specifications for the AM transmitter mask (Adjacent Channel Protection) especially beyond the first adjacent channel, although it was recognised that most radios exceed the MOPS emission requirements. It was clear that a better understanding of the cause of the problem would be achieved by detailed knowledge of the specific characteristics of the DSB-AM signal.
Further information on the discussions during this teleconference are provided in Attachment B.
2 2nd Teleconference - 25/010/02
The second teleconference took place on 25th of October 2002. The objective was to provide the opportunity to confirm the approach that seemed feasible during the first teleconference.
A number of participants prepared input for this teleconference and this information can be found in Attachment C. In particular DFS carried testing of AM radios to provide a better understanding of their performance. The DFS data provide information for the performance up to the sixth adjacent channel for a set of typical DSB-AM radios.
The DFS DSB-AM test results indicate that due to the properties of the interfering DSB-AM signal, a stricter ACR requirement may not necessarily provide the solution. A significant outcome from the group deliberations was a proposal to increase the level of the minimum desired signal for the ground scenario from the present – 88dBm to –82dBm and to revisit the ground scenario distances (consider a minimum separation distance for the ground scenario of 300m instead of 210m). It is expected that if these two options were accepted, this would enable the development of feasible frequency planning criteria for VDL Mode 4. However the group agreed that it is important to provide operational justifications before accepting these two options and that further analysis is required (see section 3).
Further information on the discussions during this second teleconference is provided in Attachment D.
Current Status and way forward
Following the second teleconference meeting, DFS provided information on the performance of the radio in the form of link budget calculations assuming that the proposals are acceptable from an operational aspect. These calculations show how a minimum separation distance of 300m between a transmitting DSB-AM mobile and the victim VDL Mode 4 mobile for the critical ground scenario, would assure protection of a minimum -82dBm signal while satisfying the SARPs 60dB ACR requirement. On the other hand, the minimum protection distance required to protect a minimum -82dBm signal between a DSB-AM ground station and a VDL Mode 4 mobile victim remains unchanged at 750m. Additionally, the calculations extend the analysis to show that the proposed values also assure protection from VDL interfering signals on the ground. The results of the these ACR calculations are provided in Attachment E.1. For information on an initial approach discussing the operational suitability of the proposed requirements (also provided by DFS), see e-mail in attachment F. It is important to note that the –82dBm signal is being considered for the ground scenario only. This scenario only extends a few nautical miles thus for this required signal level more ground stations are not required.
The participating manufacturers are also investigating the feasibility of these latter proposals by considering the DFS DSB-AM emission test results. In this process, some of the proposals initially put forward by the manufacturers have been reviewed. In particular, Rockwell Collins and Telerad have updated their contributions (see Attachment E2). They confirm that given that the observed ACP performance (in the order of 70dBc) exceeds that specified by the ED-23B ACP requirements, the proposed values allow DSB-AM and VDL4 to operate alongside using 4 guard channels. This analysis indicates that suitable frequency planning criteria for operation of DSB-AM & VDL mode 4 in ground scenarios can be achieved only when considering measured ACP performance of real DSB-AM radios and not when considering the ED-23B ACP requirements. The Rockwell Collins/Telerad recommendations are thus to improve the ACP performance specified in ED-23B from 60dB up to 69dB.
Some radio manufacturers have confirmed the ability of their radios to comply with the proposed requirements. In particular, Telerad confirmed that their 9000 series receivers are capable of ACR in excess of 65dB on the second channel when considering both –88dBm and –82dBm as the minimum signal level. While confirming the ability of their radios to comply with the proposed requirements, Sectra added that radios satisfying an ACR requirement of 60dB on the second channel are capable of handling the proposed scenarios with margin provided the noise level on the victim channel is kept sufficiently low.
At the present time, and from the analyses made available so far, the DSB-AM ACP characteristic remains a primary concern. The suitability of the proposed changes in required signal level (-82dBm) and minimum protection distance (300m) to allow feasible frequency planning criteria remains the focal point of the discussions.
Further testing on VDL Mode 4 radios in a collective effort from industry need to complete and confirm the results available so far. New equipment is required to be made available by the manufacturers to perform this testing.
The issue needs further discussion in a future WGB meeting.
References
[1] VDL Mode 4 interference test results, AMCP/WGB-13 WP4, ICAO, August 2002.
[2] Discussion on VDL Mode 4 receiver rejection performance, AMCP/WGB-13 WP5, ICAO, August 2002.
[3] Plan for Frequency Testing of VDL Mode 4, AMCP/WGB-12 WP7, ICAO, March 2002.
[4] Proposed modifications to the Ground-to-Ground interference scenario, AMCP/WGB-11, ICAO, September 2001.
Telecon to discuss the
reception performance of the VDL Mode 4 radio
30/09/02 14:00 CET
Extract from ECTL paper to FMG meeting
…
1.2 Status of the definition of the VDL Mode 4 frequency implementation rules
In order to support the definition of frequency planning criteria for VDL Mode 4 in AMCP/WGB, EUROCONTROL has started testing VDL Mode 4 radios compliant with the VDL Mode 4 SARPs and MOPS. The tests were carried out in accordance with the methodology used previously in the definition of VDL Mode 2 frequency planning criteria and approved by AMCP/WGB.
This initial testing covered some, but not all the cases described in the test plan that is agreed for VDL Mode 4 in AMCP/WGB. Due to the limited time in which the VDL Mode 4 equipment was available to EUROCONTROL, there was not sufficient time to perform all the required testing. It is intended to repeat most of the testing and complete the required measurements to enable the definition of accurate frequency planning criteria. However, the testing that has been done is very important and provides useful information for what the finalised frequency planning criteria could be.
The tests carried out so far cover VDL Mode 4 interference to and from DSB-AM, VDL Mode 2 and VDL Mode 4, using three frequencies in the COM band, being 119MHz, 128 MHz and 136 MHz. The following table summarises what could be the frequency planning criteria (number of required guard band channels), if the results of this intial testing were confirmed by further testing. In relation to interference to voice mainly S/P testing was done, which is expected to be the most restricting. S+N/N and squelch lifting testing also needs to be performed. In summary, further work is needed to confirm these results, but also to complete pending investigations.
|Victim |Interferer |Number of guard bands |Scenario |
|VDL Mode 2 |VDL Mode 4 |1 |air and ground |
|DSB-AM |VDL Mode 4 |4 | |
| |VDL Mode 4 |1 | |
|VDL Mode 4 | | |Air |
| |VDL Mode 2 |0 | |
| |DSB-AM |2 | |
| |VDL Mode 4 |? | |
|VDL Mode 4 | | |Ground |
| |VDL Mode 2 |? | |
| |DSB-AM |? | |
Table 1: VDL Mode 4 guard band channels requirements (to be confirmed)
In Table 1, what is missing in order to address all the cases, is the guard band channel requirements for the ground scenario when VDL Mode 4 is the victim (highlighted in grey).
A significant finding in the EUROCONTROL test results is that a VDL Mode 4 radio compliant with SARPs and MOPS is susceptible to interference by voice DSB-AM transmissions even at large channel separations. To a lesser degree, the issue is also valid when VDL Mode 4 is interfered by VDL Mode 2 or VDL Mode 4. These draft results were discussed in AMCP/WGB and some potential solutions were identified during the meeting. The AMCP/WGB acknowledged that the most appropriate solution seems to be the improvement of the adjacent channel rejection requirements for VDL Mode 4. AMCP/WGB is soliciting input from the radio manufacturers about the implications and feasibility of specifying a more stringent reception performance for the VDL Mode 4 radio.
A paper describing the required improvements in the VDL Mode 4 reception to achieve acceptable frequency planning criteria was presented by EUROCONTROL as a starting point in the industry discussions towards finding a viable solution to the problem. EUROCONTROL is inviting input from industry on the issue, with the aim to achieve a consensus and make a final proposal. If there will be an agreement in this direction, then EUROCONTROL will plan further testing using VDL Mode 4 radios fulfilling the new requirements to support the work of AMCP/WGB.
AMCP/WGB also discussed the value of the S/P criterion to measure the interference of VDL Mode 4 on DSB-AM voice. The value of this criterion specifies the acceptable measured level of interference on voice and the value of S/P criterion is required to define the number of required guard band channels. Using recordings done by DFS of ATC text received by a DSB-AM radios interfered by VDL Mode 4, LFV arranged for the evaluation of the recordings by listening panels composed by operational groups (pilots and controllers). Based on the results of this evaluation, AMCP/WGB recommended to use the value of 12dB, while recognising that for ground station transmissions with heavy duty factors (e.g. TIS-B transmissions) special measures may have to be taken.
The EUROCONTROL results confirm that the channel loading assumptions are critical when considering frequency planning criteria. EUROCONTROL will analyse further and confirm these assumptions.
While the testing in EUROCONTROL is only addressing the operation in the COM band, additional testing is planned by STNA and DFS for the NAV band.
2. Creation of the VDL Band
2.1 Summary of the protection criteria between the VHF systems
| |Analogue ATS |VDL Mode 2 |VDL Mode 4 |ACARS |Analogue AOC |
|Analogue ATS |Usual analogue |1 (a) |4 (c) |Usual analogue |Usual analogue |
| |criteria apply | | |criteria apply |criteria apply |
|VDL Mode 2 | |1 |1 (b) |2 |1 |
|VDL Mode 4 | | |1 (c) |4 (c) |4 (c) |
|ACARS | | | |0 |0 |
|Analogue AOC | | | | |Usual analogue |
| | | | | |criteria apply |
Table 2: Number of 25 kHz guard band channel between systems in the AM(R)S VHF band
a: specific measure on ground scenario (up to 4 channels)
b: in light of preliminary results, seems to be an achievable target
c: upper limits for frequency management
It is possible to identify the upper limits of acceptable protection criteria for the VDL Mode 4 cases although further work is required to confirm the protection criteria. This can be done taking into account the results that have been obtained from the testing that has been already carried out and for which the experts have relative confidence (VDL Mode 2 Vs VDL Mode 4) for using them as credible working assumptions.
For other cases (Analogue Vs VDL Mode 4, and VDL Mode 4 Vs VDL Mode 4), target figures, which are the upper limits for any realistic frequency assignment have been set.
These values should be the guiding instructions to the VDL Mode 4 radio manufacturers who are currently considering improvements to the radio design (reception function).
This table is used to establish the proposed VDL frequency plan.
2.2 VDL Frequency Plan
The actions required for the creation of a VDL band need to be identified in consideration of a global frequency plan and of the different steps which compose the frequency plan which has to be built in order to allow the transition from one step to the next.
Considering the preliminary status of some implementation elements (e.g. some protection criteria), the proposed frequency plan shall not be considered as definitive, but has to be considered as a tool to be used for defining consistent actions for each step.
It has to be clearly understood that in case of evolution of some of the elements (e.g. protection criteria), the frequency plan will have to be re-assessed and adapted as required.
Frequency Plan Assumptions and Facts
The proposed frequency plan is established based on the following principles, most of them have been identified at the FMG 5:
• The protection criteria between systems are according section 2.1;
• The number of ECAC-wide channels for VDL Mode 2 at the ultimate step is 4;
• The number of ECAC-wide channels for VDL Mode 4 in the VHF AM(R)S band (COM Band) is 2 (with the assumption that 2 other channels will be allocated in the Nav band pending WRC 2003 conclusion and availability of the necessary spectrum);
• Sufficient ACARS capacity shall be provided to the service providers to operate properly. One ACARS channel will still be provided to ARINC. Considering the 2 ACARS SITA channels (131.525 and 131.725), SITA will operate a third channel up to the time sufficient capacity is provided by VDL (at least 3 VDL Mode 2 channels) to take the traffic over.
• The efficiency of the spectrum has to be maximised (minimisation of the wasted channels)
Target Plan
The “Step 1” of the frequency plan will lead to the creation of a protected VDL Mode 2 Common Signalling Channel (CSC) on 136.975 and of a non-protected VDL Mode 4 Global Signalling Channel (GSC) on 136.925 by mid 2003.
The other steps lead to the implementation of 4 protected VDL Mode 2 channels and 2 protected VDL Mode 4 channels.
…
End of extract from paper to FMG meeting
Other input received from Industry
MARCONI
Concerning the documents on the matter we have some doubts about the different performance of the VM4 receiver when the interferer is an AM DSB or VDL 2 transmitter: the measurement should be done using a suitably filtered signal generator (to avoid WB noise contribution). The results should be analysed in terms of co-channel and blocking and adjacent
channel performances.
We enforce the proposal to investigate the performance of the VDL4 demodulator in terms of voice AM modulated interferer instead of tone modulated one before studying modification to the SARPs (spectral masks or channel coding or else).
ADSI
With regard to the suggested change in the SARPs to improve the VDL/4 – ACI we feel that a tightened mask should be specified. Basically, we feel that restrictions on spectrum planning must be minimized and improved specifications that can speed-up the implementation of robust VDL/4 systems should be the desirable objective of any changes.
As these activities will ultimately require changes in the SARPs, we also think it is an appropriate time to discuss the reasons why GFSK is more sensitive than D8PSK to the DSB-AM interference. If it is because of a lack of FEC, it may be appropriate to review all of the physical layer specs while we are looking at the receive ACI performance. If you remember
the earlier TLAT analysis of VDL/4 to support ADS-B, our poor CCI performance versus a vastly superior UAT (original version) did not help us. Now with an even more robust UAT physical layer being proposed for standardization in the present MOPS, I believe we should consider how we can improve our physical layer while preserving the other benefits of
VLD/4. Certainly, today's technology provides a greater opportunity to build better and more economical R/Ts so we may want to capitalize on the opportunity.
On another point, we (ADSI) have noted previously, that we do not intend to, unless new information is shared, build equipment to operate down to 108 MHz. primarily because of FM immunity issues. Therefore, we suggest that future work address VDL/4 operations in the 112 - 136.975 MHz range.
ISSUES
- FM immunity standards. Have we finalized these specs?
- ACI protection standards (e.g., against a DSB-AM aggressor) We agree that we should take on reasonable technical solutions to minimize any problems related to VDL/4 deployment and frequency assignments. The fewer restrictions that we place on the frequency managers, the quicker VDL/4 deployment can occur.
- band of operation. We submitted a proposal to limit VDL/4 operations to 112-136.975 MHz and giving up operations in 108-112 MHz. I note that some of the papers that you submitted still talk about operations in the 108-112 MHz band. In order to satisfy FM immunity in 108-112 MHz, the preselector will require a much sharper roll-off. The sharper the roll-off (i.e., the smaller the transition band), the more poles are
required in the filter. Absent room-temperature superconductors, each pair of poles adds significant insertion loss. This insertion loss directly reduces the sensitivity of the receiver and the link budget of the system.
- physical layer performance. It is clear that the GFSK waveform without any FEC has hurt VDL/4's performance. This is seen in the TLAT reports (e.g., the old UAT waveform -- before they improved it -- had a CCI performance about 7-8 dB better than VDL/4) as the VDL/4 update rates were very poor at long range because the heavy capacity caused a high degree of
slot sharing (an improved CCI would have improved the receive update rates). Further, FEC would have improved the DSB-AM interference problems -- perhaps even enough to obviate the need for this telecon.
- validation schedule.
Rockwell Collins – SAAB
We are proposing the following road map in order to solve the issue, and participate efficiently in the definition of the performance requirements pertaining to Adjacent Channel Rejection.
#1: Analyze the test method and the results documented here above. An analysis paper will be submitted. Objective date is mid October.
#2: Review/discuss the results of our analysis with the AMCP group, in order to validate/freeze the requirements. Date to be planned end October.
#3: Assess analytically the technical feasibility of such requirements, according to state-of-the-art airborne receiver implementations (before mid November)
#4: Finalize the frequency planning criteria in time for the next AMCP group meeting in December
Teleconference discussing VDL4 receiver performance
30/09/02
Summary of discussion/actions
Participants:
|ADSI |Steve Friedman |
| |Prasad Nair |
|CNSS |Tommy Bergström |
|DFS |Armin Schlereth |
|EUROCONTROL |Philippe Renaud |
| |Bertrand Desperier |
| |Patrick Delhaise |
| |John Micallef |
| |Nikos Fistas |
|ICAO |Robert Witzen |
|LFV |Larry Johnsson |
|MARCONI |Paolo Maltese |
| |Andrea Berti |
|Rockwell Collins |Didier Marchetti |
|SAAB |Stephen Olson |
|SECTRA |Fritjof Qvigstad |
|SITA |Rodolphe Biet |
|TELERAD |Patrice Mariotte |
The teleconference was organised to discuss the performance of the VDL Mode 4 radio in the light of some initial testing performed by ECTL for the AMCP/WGB meeting in August 2002.
The initial testing indicates that the performance of the VM4 receiver may be significantly degraded when the desired signal is interfered by other systems.
In order to comply with the testing scenarios agreed in AMCP, an impractical number of guard band channels would be required.
The main points in the discussons in the telecon were:
• The reception performance of the VM4 radio is an issue that the industry (VM4 manufacturers) should take the lead.
• ECTL would be a facilitator to co-ordinate an agreement among the interested parties in support of the AMCP/WGB discussions to define frequency planning criteria for VM4.
• The victim and interferer loads are critical for the testing. The values that were used need to be reviewed and verified that are realistic and correct (providing a realistic worst case scenario). In addition the testing conditions need to be carefully considered (e.g. need to filter the wide band noise effect in the measurements)
• It was recommended that the testing should be done using a 100% load, to facilitate the testing procedure, but then the results will need to be translated to the realistic loads. In addition the issue of the repeatability of the testing was raised. It was recommended to use signal generators to generate the interferer waveform.
• There was discussion about the lack of explicit specifications for the AM transmitter mask, and consequently about the assumptions that need to be made to measure the impact of/to the DSB-AM radio. In particular the lack of specific requirements for DSB-AM beyond the first channel was mentioned. However, it was highlighted that it will not be feasible to change the requirements of existing systems. The emphasis needs to be put on what can be done with the specifications of the VM4 radio.
• The manufacturers seem to agree that the draft proposal specifying requirements for the “reception” performance of the VM4 radio is feasible. However this needs more consideration and evaluation. The VM4 radio manufacturers are invited to finalise the proposal and to agree in the way forward.
• There was discussion about some potential benefits of using specific frequency bands or parts of bands. However it was emphasised the need to provide frequency planning criteria, which would be applicable in general in the VHF COM band. The draft EUROCONTROL proposal to the ICAO EUR Frequency Management Group (FMG) for a frequency plan to implement VDL services (VM2 and VM4) and especially the upper limits in the number of guard bands it recommends, needs to be taken into account when considering the potential improvements for the VM4 radio.
• The unexpected difference in the spectral performance of the GFSK modulation (VM4) versus D8PSK (VM2), was discussed and the need to explain the differences was raised. The use (VM2) or lack (VM4) of a Forward Error Correction (FEC) scheme and the differences in assumed channel loading was mentioned as potential reasons.
• In relation to the scheduling, the tentative dates for the next WGB meeting in December 2002 was mentioned. None of the participating manufacturers was able to commit to provide any new equipment so that a new test campaign would be made. The telecon concluded that it is unlikely that there will be new testing results by the December timeframe. The need to do the testing correctly and with some sufficient time to evaluate all the required cases and analyse the results was emphasised.
• Due to the limitations of a teleconference procedure, it was agreed that participants with specific proposals will submit them in writing to the other participants to facilitate the discussion and exchange of views. A second teleconference will be held following this to evaluate if there is a possibility of an agreement among the VM4 interested parties for an action plan targeting a future meeting of WGB.
Actions:
|No |Action |Who |By when |
|1 |Participants to provide/cofirm by email their names and email |All |10/10/02 |
| |addresses | | |
|2 |Organise a teleconference for the 25th October 14:00 CET to |ECTL |24/10/02 |
| |progress the issues discussed. | | |
|3 |Provide written contributions for discussions in the |All |18/10/02 |
| |teleconference of Action 2 |In particular | |
| | |Rockwell-Collins/SAAB, SECTRA, | |
| | |ADSI, DFS and Marconi indicated | |
| | |that they will provide input. | |
VM4 radio receiver performance
Teleconference on 30/09/2002
Positions and technical suggestions
|EURO CONTROL |Summary of results and test method including assumptions. |
| |Industry to take the lead: Position of EUROCONTROL as facilitator to support the work. |
| |Focus must be on COM band. If VM4 is to be operated we have to minimise any interference with or by other |
| |equipment. |
|ADSI |Position with 112-118 MHz: Not used for VDL4 |
| |D8PSK does not show vulnerabilities of GFSK. Pointed out the need to reveal why D8PSK does not suffer the |
| |problems of GFSK. |
| |DSB-AM has no requirements beyond the first channel. The consensus at the time of drafting of the VDL2 and |
| |the 8.33KHz SARPs was that there was no reason to add requirements to existing system. |
|SECTRA |Lower frequencies in NAV band not required. |
| |Requirements should not be placed for things that are not going to be used. |
| |Rejection mask provided is fully obtainable but in order to comply with ACR requirements we need to adjust |
| |ICAO transmitter mask requirements. |
| |Using existing transmitter mask, not more that 58dB rejection is possible. |
|CNS |Tried to duplicate measurements, succeeded in certain parts but did not achieve the poor results reported. |
| |Results obtained indicate no need for 4 channels guard band but 1 channel is sufficient (in both directions)|
| |between DSB-AM and VM4. Thus a marginal difference between requirements and what is actually achieved. |
| |The test report should be able to give data necessary to duplicate the test; should supply part and serial |
| |numbers. |
|RC/SAAB |In the process of developing airborne radio. |
| |No tests have been done on the radio using the Ectl. Methods of frequency testing. |
| |Would like to participate in developing a test method by proposing a method to determine the mask. This |
| |method could be made available 2-3 weeks from now. |
| |The process should aim to freeze the requirements since this is not yet agreed. |
| |Agree to issue a paper that clarifies issues with this. |
| |Would like to see a simpler definition of spectral mask for easier and more repeatable testing. |
|DFS |60dB for 4th channel used with input of –87dBm. |
| |Propose to change requirement to 70dB and should work in range of –82dBm - –88dBm. |
|MARCONI |VM4 ground station is under development and will be ready for measurement tests by the end of the year. |
| |Identified that most critical issues are with airborne radio on the ground; |
| |With the actual VM4 radio used, the 4th channel was much worse than the 3rd (not the same with VM2). |
| |Unable to understand this observation as just a modulation problem. Hence propose the use of modulated voice|
| |instead of using a tone. |
|TELERAD |Some testing has been carried out on DSB-AM rejection. |
| |Results are positive: Obtained 68dB rejection at –88dBm interfering power (AM rejection). This provides good|
| |margin with respect to the present requirements. |
| |Tests conducted with 100% duty factor all the time to obtain high fidelity requirements. |
Teleconference to discuss the
reception performance of the VDL Mode 4 radio
25/10/02 14:00 CET
Radio Manufacturer and other Contributors Input
The following are various contributions received from participants as input to the follow-up teleconference on the issue of VDL Mode 4 receiver performance.
DFS contribution
regarding the requirements I would propose to have from the 4th adjacent channel onwards an ACR of at least 70 dB. This should hold for -88 dBm to -82 dBm desired signal input power.
This is in line with the WG-B criteria defined so far, meaning that
preliminary frequency planning criteria could stay.
We should avoid any further ongoing discussion on what could be possible. My advice would be to have SARPs chances with the above mentioned requirement presented to AMCP/8 then.
Regarding the proposed frequency planning You provided me with input, I have to sy, that this is not a good idea to inerleave VDL Mode 2 with Mode 4. What i would expect in order to be able to increase isolation is a subband for each of the VDLs. But in that context manufacturers should also give their input.
In response to the need to investigate closely the characteristics of the signal to be rejected, test have been carried out on a sample of three DSB_AM radios. The following tables summarize results of a measurements campaign on these radios considering their emission spectrum up to the 6th adjacent channel. Three radios have been tested:
1. Rockwell Collins 920 (25 kHz and 8,33 kHz Mode) Output power: 44dBm
2. Bendix 42a (25 kHz radio) Output Power: 44dBm
3. Dittel FSG90 (25 kHz and 8,33 kHz Mode) Output Power: 37dBm
Traces from the results are shown in the appendix to this paper. For each radio there are two traces. The upper one has the AM-signal at the centre. At the right top part of the figure the measured power values in a 25 kHz grid are outlined. e.g. for Dittel 8.33KHz mode Channel Power is -4.97 dBm corresponding to 37 dBm output power(as indicated above). The First adjacent channel up has -75,2 dBm and the first adj. channel down -75,83 dBm. For the second adj. channel up and down one can read -77,89 dBm and -76,20 dBm. and for the third adjcent channel up and down values are -79,96 dBm and -80,03 dBm.
The second figure just shifted the centre frequency to the left, in order to be able to measure up to the 6th adjacent channel, which corresponds to ALT2 Up with a value of -75,60 dBm. The centre power value of -75,95 dBm corresponds to the third adjacent channel up.
The noise floor of the analyser was at -85 dBm.
The results are summarised here below:
|Dittel FSG90 | | | | | | |
| | |8,33 kHz Mode | | | | |
|Adj. Channel # |0 |1 |2 |3 |4 |5 |6 |
|Output Power [dBm] |37 |-33,3 |-33,9 |-34 |-32,4 |-33 |-33,6 |
|dBc |0 |-70,3 |-70,9 |-71 |-69,4 |-70 |-70,6 |
| | | | | | | | |
| | |25 kHz Mode | | | | |
|Adj. Channel # |0 |1 |2 |3 |4 |5 |6 |
|Output Power [dBm] |37 |-35,2 |-35,3 |-33,6 |-32,6 |-32,6 |-33,9 |
|dBc |0 |-72,2 |-72,3 |-70,6 |-69,6 |-69,6 |-70,9 |
| | | | | | | | |
| | | | | | | | |
| | | | | | | | |
|RC920 | | | | | | | |
| | |8,33 kHz Mode | | | | |
|Adj. Channel # |0 |1 |2 |3 |4 |5 |6 |
|Output Power [dBm] |44 |-26,1 |-27,6 |-27,1 |-27 |-26,4 |-27,39 |
|dBc |0 |-70,1 |-71,6 |-71,1 |-71 |-70,4 |-71,39 |
| | | | | | | | |
| | |25 kHz Mode | | | | |
|Adj. Channel # |0 |1 |2 |3 |4 |5 |6 |
|Output Power [dBm] |44 |-25,8 |-28,8 |-27,4 |-26,4 |-26,8 |-26,5 |
|dBc |0 |-69,8 |-72,8 |-71,4 |-70,4 |-70,8 |-70,5 |
|Bendix42a | | | | | | |
| | |25 kHz Mode | | | | |
|Adj. Channel # |0 |1 |2 |3 |4 |5 |6 |
|Output Power [dBm] |44 |-17,6 |-26,2 |-26,2 |-25,5 |-24,6 |-25,8 |
|dBc |0 |-61,6 |-70,2 |-70,2 |-69,5 |-68,6 |-69,8 |
MARCONI contribution
As already pointed out during the telecon, we should separate the contributions to the poor ACR results of the VM4 receiver and the AM-DSB interference transmitters, before giving the GFSK modulation scheme full responsibility for it. In order to do this, a complete characterization of all radios involved, VM4 and AM-DSBs, in terms of receiver ACR, CoChannel and transmitter ACP masks, should be provided. The test report should clearly indicate the reference standards for each equipment (both airborne and ground) and the above measurements should follow the procedure indicated in the corresponding standard documents (e.g. for VM4 ground-based equipment, ETSI EN 301 842-1 clearly describes the test procedures for all physical layer specifications).
The use of signal generators as interference sources is implied by conducting the above measurements according to the standards' procedures. We also point out that the cochannel performances of the VM4 radio is to be reported, to better analyze the effect of interferent transmitter WB Noise.
The second step should then be the AM-DSP interference on VM4 test, conducted with a signal generator as AM-DSB interferent, not a commercial radio.
The ACR mask should be determined twice, the first time with a 1 kHz tone @ 30% depth as interferent signal, the second time with typical ATC voice communications as modulating signal for the AM.
We agree with the other manufacturers concerning the need for the test report to clearly describe the test environment and the parameters used, in order to allow for replication of the tests.
SECTRA contribution
Regarding ACR, some assumtions:
- GFSK (as in VDL4) has co-channel interferer rejection of 12 dB for 2% MER (VDL as on.channel interferer)- GFSK (as in VDL4) has a required SNR of 18 dB for 2% MER (white gaussian noise as on-channel interferer) - Side band transmitter noise has different properties from the on-channel signal. The side band noise is similar to white noise (or colored noise) and occupies the whole 25 kHz channel.
--- So lets assume that 17 dB D/U is required (U = interference from transmitter on some adjacent channel)
- A 100% channel signal and interferer load with 35% MER can be shown to
correspond (approx.) to a 2% channel signal and interferer load (as in VDL2 scenario) with 2% MER.
- Accepting 35% MER instead of 2% MER can give (approx.) 5 dB improvement on D/U figures above.
--- So we can accept 12 dB D/U (at least for the VDL2 intereferer scenario)
- The ICAO transmitter spectrum mask allows -28 dBm emission on second adjacent channel (-70 dBc).This means that we have an ACR of 58 dB (70 - 12 dB for the required D/U) at 35% MER
(*) 2:nd channel, VDL2 interferer: 53 dB @ 2% MER and 100% channel load As above we get 68 dB for the 4:th adjacent channel at 35% MER
(*) 4:th channel, VDL2 interferer: 63 dB @ 2% MER and 100% channel load)
Exectutive summary:
The figures above are 8 dB from what is desired for the frequency planning. However this is what is
possible to achieve with an interfering transmitter emitting wide-band noise (spurius emissions) according to the ICAO defined spectrum mask.
Ways to improve:
- Disregard the spectrum mask and assume that transmitters are better then required (assume a spectrum mask 8 dB more stringent then the existing one). this will work fine if most transmitters are so much better....
- Change the ICAO spectrum mask according to previus suggestion. (At least technical feasable...)
- Make some changes in the VDL4 protocol to improve reception. 8-10 dB improvement is achievable. (At least technical feasable...)
- Change the scenarios and thus the frequency planning criterias
Measurements:
- ACR measurements should be performed at 100% channel loads for victim and interferer.
- Verify the ACR using high-end signal generators used as interferers to ensure no on-channel
noise (D/U at least 20 dB)
- Verify that VDL4 equipment can handle 18 dB D/U using white noise generators.
FM-immunity:
If we believe there might be frequencies available close to 108 MHz it could be useful to specify the same requirements for VDL4 as for other nav systems (ILS, GBAs etc.). It is achievable to implement this. As it makes the VDL4 unit more complex it should only be considered if there comes some frequencies for VDL4 in that band.
ROCKWELL COLLINS and TELERAD contribution
Introduction
The purpose of this paper is to comment the paper from M. Desperier dealing with recommendations on VDL4 Adjacent Channel Rejection performance dated 26-29 august.
This paper establishes minimum ACR and the number of guard frequencies necessary to obtain the 210m goal between two aircrafts, considering any type of interferer (DSB-AM, VDL2, and VDL4).
This paper also asks for an improvement on the ED23B (DSB-AM MOPS), that is necessary to get consistency between VDL4 (or VDL2) system and DSB-AM transmitters characteristics.
ACR versus CCI :
Adjacent Channel Rejection (ACR) and Co-Channel Interference (CCI) are two different performances that are generally tested separately.
ACR definition :
The ACR is used to test the susceptibility of the receiver to a strong signal on a adjacent channel. Typically, the VDL4 MOPS (ED-108) ask for 40dB ACR on the 1st adjacent channel, and 60dB ACR on the 5th adjacent channel. The ACR is tested without any interferer signal transmitted within the pass band.
ED108 Note 1 page 76 clearly indicates that the test is performed without noise in the pass band:
Adjacent Channel rejection (test procedure)
The noise side band of the interfering signal must not interfere with the desired signal in the desired pass band….
CCI definition :
The CCI is used to test the susceptibility of a receiver to a interferer signal transmitted into the pass band. Typically, the MOPS VDL4 (ED-108) asks for 10dB CCI.
Combining ACR test and CCI :
The test performed par M.Desperrier consists in transmitting a interferer signal with a real transmitter. This test combines both ACR and CCI characteristics, the CCI being generated by the noise of the interferer transmitted into the pass band.
Although this test is near to the real life, it is not easily reproducible and its results may more depends on the unknown noise characteristics of the interferer, than on the ACR performances that are to be measured.
What we observed :
When performing the ACR test, if the pass band is not jammed by any interferer wide band noise, we could not notice any discrepancy when using different interferer modulations.
So we do not confirm that the type of modulation has an impact on the ACR performance.
We think that M.Deperier used different kind of interferer with different noise and spurious characteristics. And this leads to wrong conclusion, because the ACR test must be performed without any signal present in the pass band. We think that the results found by M.Desperier are more reflecting the CCI characteristics of the VDL4 receiver combined with the noise characteristics of the interferer, than the ACR characteristics of the VDL4 receiver.
When performing interference test, we must use a 100% load otherwise, otherwise the measurement is not stable and the results may not reflect the worst case of synchronous frames.
General case
When a interferer and a victim are transmitting in the same area, what are the performances necessary to avoid adjacent channel interferences?
The relative value of the spurious of the interferer must not be higher than the desired low signal :
|ACR(victim) + CCI(victim) < ACP(interferer) |(1) |
This relation means that it is not necessary to have a ACR(victim) greater than ACP(interferer) – CCI(victim) otherwise, the low desired signal is jammed by the spurious of the interferer.
When writing the absolute value of the spurious, we get :
Pmin(victim) – SPURIOUS(interferer) > CCI(victim)
Pmin(victim) - [ Pmax(interferer) – Attenuation(cable + antenna + space) – ACP(interferer) ] > CCI(victim)
|Pmin(victim) - Pmax(interferer) + Attenuation (cable + antenna + space) + ACP(interferer) > CCI(victim) |(2) |
Relations (1) and (2) can be illustrated as follows :[pic]
According to the following assumptions :
|Pmin(victim) = -88dBm |According to ED-108 |
|Attenuation(cable + antenna) = 2x7dB = 14dB |According to RTCA DO224 |
|Attenuation(space) = 60dB |Desired 210m between aircrafts |
|CCI(victim) = 10dB |According to ED-108 |
Relation (2) can be written :
|ACP(interferer) - Pmax(interferer) > 24 dBm |(3) |
In the worst case, we would have: ACP(interferer) - Pmax(interferer) = 24 dBm
And then (1) can be written
|ACR(victim) + CCI(victim) < 24dBm + Pmax(interferer) |(4) |
DSB-AM interferer against VDL4
What is the worst case of a DSB-AM interferer against VDL4 ?
(3) ACP(DSB-AM) - Pmax(DSB-AM) > 24 dBm
According to ED23B the maximum power of a DSB-AM transmitter is 44 dBm.
In this case, the relative ACP is 60dBc. So (3) can be written :
(3) 60dBc – 44dBm > 16dB
(3) is false, so the ACP of the DSB-AM must be improved.
The minimum ACP of the DSB-AM must be 68dBc (=16dB + 44dBm)
Remark :
This demonstration does not depend on the guard band, because the bad ACP of DSB-AM transmitter can generate spurious anywhere on the band. So even with 20 guard band channels, this problem is still present.
Conclusion :
To determine a number of guard channels, the ED23B must ask for a more stringent ACP performance
This ACP must be 68 dBc (instead of 60dBc) in the case of a 44dBm transmitter
This ACP must be 64 dBc (instead of 56dBc) in the case of a 40dBm transmitter
Improved DSB-AM interferer against VDL4
In the case where the ACP performance of the ED-23B is increased, what should be the ACR value of the VDL4 on the 5th channel ?
According to (1), the minimum ACR(VDL4) is
ACR(VDL4) < ACP(interferer) - CCI(VDL4)
ACR(VDL4) < 68 - 10
ACR(VDL4) < 58dB (on the 5th channel)
Conclusion :
If the ED23B ACP value is better than 68dBc (for 44dBm, or 64 for 40dBm) on the 5th channel, then 58dB ACR performance is enough to avoid interferences with VDL4.
VDL4 interferer against VDL4
What should be the ACR value considering VDL4 interferers ?
What is the number of guard channels in that case?
Assumptions
• The VDL4 transceiver output power is 43dBm
• The CCI(VDL4) is 10dB
• The channel load is 100%.
Demonstration:
1Guard channel : The VDL4 SARPS asks for an ACP of -28dBm on the 2nd channel.
The relative ACP is:
ACP(VDL4) = 43dBm-(-28dBm) = 71dB
According to (3) we must have : ACP(interferer) - Pmax(interferer) > 24 dBm
71dB – 43dBm > 24dBm
28dBm > 24dBm
We find that (3) is true, so 1 guard channel is enough.
According to (4) , the minimum ACR(VDL4) is then :
ACR(victim) + CCI(victim) ≤ 24dBm + Pmax(interferer)
ACR(VDL4) + 10dB ≤ 24dBm + 43dBm
ACR(VDL4) ≤ 57dB
Conclusion :
• 1 guard channel is enough to avoid interferences.
• With 1 guard channel, 57dB ACR performance is enough to avoid interferences at 210m.
VDL2 interferer against VDL4
What should be the ACR value considering VDL2 interferers ?
What is the number of guard channels in that case?
Assumptions
• The VDL4 transceiver output power is 42dBm (this is the only difference with VDL4 interferer)
• The CCI is 10dB
• The channel load is 100%.
Demonstration:
1Guard channel : The VDL4 SARPS asks for an ACP of -28dBm on the 2nd channel.
The relative ACP is:
ACP(VDL4) = 42dBm-(-28dBm) = 70dB
According to (3) we must have : ACP(interferer) - Pmax(interferer) > 24 dBm
71dB – 42dBm > 24dBm
28dBm > 24dBm
We find that (3) is true, so 1 guard channel is enough.
According to (4) , the maximum ACR(VDL4) is then :
ACR(victim) + CCI(victim) < 24dBm + Pmax(interferer)
ACR(VDL4) + 10dB < 24dBm + 42dBm
ACR(VDL4) < 56dB
Conclusion :
• 1 guard channel is enough to avoid interferences.
• With 1 guard channel, 56dB ACR performance is enough to avoid interferences at 210m.
Our recommendations
Test methods:
We recommend to distinguish the ACR test and the CCI test.
For the ACR test, we suggest to use the existing test method of the ED-108
FM 400HZ sine wave, +/-5.25Khz deviation. This spectrum was used for the VDL2 interference testing, for the following reasons :
• It is easy to generate and reproduce
• It looks like VDL2 spectrum
The VDL4 spectrum is slightly different, but not enough to justify a different interferer spectrum.
Improvement of ED23 ACP performance:
We recommend to improve the ACP performance of the ED-23B from 60dB up to 68dB for 44dBm output power, and 64dB for 40dBm output power, so that new DSB-AM transceiver will no more interfere with the low level VDL4 signals (-88dBm). Till this ED23 ACP characteristic is improved, then the goals of 210m, and -88dBm minimum power, cannot be achieved, whatever the number of guard channels is.
Note: this issue is NOT specific to VDL4, but also applies to VDL2.
ACR Value :
Assuming that the ED-23B ACP performance is improved as described above, then the ED-108 ACR performance should be 57dB on the 2nd channel (to cope with VDL2/VDL4 transmitters) and 58dB on the 5th channel (to cope with DSBAM transmitters), in order to avoid interferences when the 2 aircrafts are separated by 210m.
The current ED-108 value of 60dB at the 5th channel is more conservative than the above recommended value and can remain unchanged.
The only new ED-108 requirement that would allow a single guard channel from VDL2/VDL4 transmitters is the “57dB on the 2nd channel”. From the testing activity performed by Collins in Telerad on a Telerad receiver, we can already say that this performance is possible to achieve. But in order to later discuss the impacts of this new requirement with all the manufacturers, we suggest that they all test this performance with their equipment, and using the test method of the ED108.
Number of Guard channels
Assuming that the ED-23B ACP performance is improved as described above, and that the recommended ED-108 ACR new requirement is adopted, the number of guard channels is as follows :
• 1 guard channel for VDL4 and VDL2
• 4 guard channels for DSB-AM.
ADSI contribution
1) Interference source
I would like to suggest that one possible reason why VDL/2 did not have the ACI problems that VDL/4 had was the testing methodology, namely the interference source. If VDL/2 tested against a golden interferer (which caused no interference in the desired channel) while VDL/4 used actual DSB-AM radios (which caused interference in the desired channel), then the problem is simply an apples to oranges comparison. In fact, the use of a golden interferer which does not present any energy in the desired channel is a fine test for a qualification test (and possibly a MOPS test) where one is trying to determine the performance of the receiver front end as distinct from the demodulator. Since VDL/2 has MOPS tests defined, it is possible that the VDL/2 testing simply used the relevant MOPS tests without verifying whether the test is applicable.
However, a golden interferer is not suitable for use as a system test
designed to inform real world analysis of the actual performance of a VDL/4 system operating near a DSB-AM channel. In tests that I performed 10 years ago, some DSB-AM radios were emitting short bursts of broadband noise at a level of -10 dBm more than one MHz from the carrier frequency (these were key-on and key-off transients). Even at 1 nmi range, the undesired signal is momentarily stronger than the desired signal. Using this radio as the undesired emitter isn't testing the receiver design (since no VDL/4 receiver can operate down to 0 dB SNR). FEC might provide the necessary time diversity (there exist other radios emitting lower levels of noise for longer periods of time would defeat any FEC scheme we might adopt), and system designs using space diversity might provide suitable filtering, but in general all modulation schemes won't operate properly.
Testing with a golden interferer should be adopted if VDL/2 insists on using a golden inteferer (after all, why should we penalize ourselves), but we should admit both to ourselves and to the frequency managers that these results do not provide data that will be operationally meaningful.
2) Interference model
We agree that a 100% load should be used, but again only if it is
harmonized across all of the VDL modes. In fact, we believe that the exact same test procedures should be used for all of the VDL modes. If the VDL/2 testing has been completed and will not be revisited, then we should (at a minimum) perform their tests for VDL/4. If we feel that other tests are important, these can be done as well, but it is important that apples to apples comparisons be possible. [If the tests that ECTL has done to date were not identical to what has been done on VDL/2, then I would suggest that ECTL perform these tests on the existing radios by December to provide the committee with a better understanding and ability to compare the VDL modes. If, as I suspect, VDL/4 passes these tests, then we can provide
WG/B with a better understanding of the system problems going into AMCP/8.]
Eurocontrol contribution
AMCP WGB has defined different scenarios close to operational interfering situation that could be encountered. The tests parameters and methods used for that purpose are not intend to be proposed to the manufacturers for their own tests. Then it is clear that for manufacturer’s testing, it should be for test simplification and reproducibility:
- 100% channel load.
- Use of a signal generator with specified tuning as interferer source.
When we are talking about ACR, as underlined by Rockwell Collins Telerad contribution both adjacent and “in band” channel interferences are considered.
I say “in band channel” because it is not really CCI. May be the term “total D/U (total desired to undesired power ratio)” is more appropriate than ACR.
210 meters corresponds in free space (including 3dB wire loss approved in WGB) to 64dB attenuation.
Now we have two separate 2 cases:
1. VDLx to VDL4
First as explained in Sectra contribution some MER equivalence can be made between pulsed and continuous mode. For VDL2 to VDL4 interference, it has been shown that 2% MER with the defined pulsed configuration was equivalent to 70%MER (65%) for Sectra (I think 35% is the message success rate). That gives a 5 dB margin for signal to interferer ratio.
The same methodology should be apply for VDL4 to VDL4 taking into account the slot distribution pattern.
VDL2 output power is 42dBm, attenuation 64dB, 5dB margin, 2nd ACP 70dBc :
Is it possible to achieve 2% MER with 100% channel load ?:
Absolute interferer power : 42 – 64 –5 = -27dBm (61 dB ACR)
2nd Side band channel power : -27 –70 = -97dBm (9 dB D/U)
Apparently not because 9dB is not enough. Then two possibilities appear:
- Use a minimum victim signal level of –82 dB (55 dB ACR and 15dB D/U)
- Use the 3rd channel instead of the 2nd (54dB ACR and 16 dB D/U)
2. AM to VDL4
AM channel load has to be continuous. It shouldn’t be necessary to test both a well known signal and a typical ATC speech. 1kHz tone with 30% modulation depth is close to real speech with 90% modulation peaks.
Furthermore, there is no AM transmitter mask specified as VDLs. Then it could be found an AM radio with bad characteristics that will jam VDL4 transmission whatever the channel separation is.
So what could be done is to find the spectrum characteristics of the most common radios. Unfortunately after a quick look of DFS measurements, the ACP on the fifth channel is around 70dBc, which is clearly not enough to lead to the required protection.
Measurements
I do agree with Sectra proposal adding CCI test with a VDL2 signal.
To answer some ADSI’s questions, it should be clear that there is no competition between VDL2 (which is not the topic here) and VDL4. Furthermore, the same test campaign was done more than one year ago (real interferer, many equipment under test) for VDL2 and these problems did not appear. We could discuss hours about VDL2/VDL4 comparison, (technologies and operational field) but it is not the question.
Appendix : Test Results on spectrum of DSB-AM radios (DFS)
Dittel FSG90: 8,33 kHz Mode; f = 128 MHz
Dittel FSG90: 25KHz Mode; f = 128MHz
Rockwell Collins 920: 8,33 kHz Mode; f = 128 MHz
Rockwell Collins 920: 25 kHz Mode; f = 128 MHz
Bendix 42a: 25 kHz Radio; f = 128 MHz
Teleconference discussing VDL4 receiver performance
25/10/02
Summary of discussion/actions
Participants:
|ADSI |Steve Friedman |
|CNSS |Tommy Bergström |
|DFS |Armin Schlereth |
|EUROCONTROL |Nikos Fistas |
| |Bertrand Desperier |
| |John Micallef |
|Rockwell Collins |Didier Marchetti |
| |Dominique Cretsey |
|LFV |Larry Johnsson |
|MARCONI |Paolo Maltese |
| |Andrea Berti |
|SAAB |Stephen Olson |
| |Marcus Gustafsson |
|SECTRA |Fritjof Qvigstad |
|SITA |Rodolphe Biet |
|TELERAD |Patrice Mariotte |
Not able to participate
|ICAO |Robert Witzen |
|STNA |Christophe Dehaynain |
The teleconference was organised to follow-up discussions from the previous teleconference on the performance of the VDL Mode 4 radio in the light of some further testing and considerations by industry.
Participants provided explanations of their contributions and proposals were made. Amongst the input presented were results on tests carried out on DSB-AM radios that provided an insight into the characteristics of DSB-AM signals. These results show that the ACP characteristics of DSB-AM signals are consistently within the 70dBc range for up to the 6th adjacent channel. It is thus clear that in the presence of an interfering DSB-AM signal at a distance of 210m the total interfering power at the victim input is in the same order as the –88dBm minimum desired signal resulting in an unsuitable desired-to-undesired signal ratio. In light of this it was proposed that a feasible way forward is to re-evaluate the validity of the scenario parameters and consider relaxing the 210m minimum separation distance requirement to 300m and consider –82dBm for the minimum desired signal level if this can be justified from an operational point. These proposed values would then ensure a suitable ACR for the critical ground scenario.
Main points also discussed in the teleconference were:
• A recommendation to change one MOPS ACR requirement for VDL Mode 4 to 57dB on the 2nd channel, as this was seen as achievable by present radios. It was agreed to check the consistency of this proposal with the results obtained from the DSB-AM spectrum testing.
• It was shown that the modulation scheme used is not the cause of any differences in ACR performance. It was suggested that a possible solution consists in improving the ACP performance of the ED-23 and alternative values were recommended. However it was concluded that this is not present a feasible way forward.
• It was identified that the conclusions of this group need to be brought to the attention of AMCP/WGB. Other potential issues such as discrepancies in definitions should also be addressed. There are tentative dates for the next WGB meeting in December 2002 and though it has been agreed that no testing is likely till then, interested parties should prepare material for WGB.
• As most participants did not have sufficient time to review the DFS it was agreed that they evaluate the feasibility of the current proposals in light of experimental results presented in the meeting. A future teleconference will be held if needed, and any participant should indicate if they feel the need to discuss any issues. Eurocontrol will arrange logistics.
Actions:
|No |Action |Who |By when |
|1 |Supply summary of meeting discussion and list of actions. |ECTL |asap |
|2 |Provide statement of proposed requirements of -82dBm minimum |DFS |31/10/02 |
| |desired signal and relaxation of minimum separation distance | | |
| |to 300m and contribute to their evaluation. | | |
|3 |To indicate whether the radios from the different |VM4 Radio manufacturers |01/11/01 |
| |manufacturers would be able to comply with the above | | |
| |requirements. | | |
|4 |Re-evaluate the feasibility of RC/TELERAD proposals in light |Rockwell Collins, TELERAD |08/11/01 |
| |of the DSB-AM radio results obtained by DFS. | | |
VM4 radio receiver performance
Teleconference on 25/10/2002
Summary of Main Points
|EURO CONTROL |The need to prepare material for AMCP/WGB. |
| |The need to agree in WGB and then evaluate the performance and develop/confirm planning criteria. |
| |Drew attention that LSCs may be allocated in addition to GSCs. Hence if we are able to reduce the number of |
| |guard bands the frequency planning criteria of VDL4 can be made a simpler process. |
|ADSI |The importance of using consistent test methods and channel loads. |
|SECTRA |The need to separate ACR measurement from noise measurement. SNR also has to be considered importantly. |
| |Noise in victim channel has the characteristics of white noise. |
| |Practical improvements to VDL4 system to gain some dB, for instance improvements in sync sequence, error |
| |corrections and parameters in modulation scheme. |
|SAAB |The main identified problem is in-band noise. |
|Rockwell Collins |Proposal of 57dB ACR on 2nd adjacent channel to protect VM4 from VM2/VM4 to maintain the 1 guard band |
| |requirement. This value may have to be reconsidered. |
| |Recognised the impracticality of modifying ED-23 for current equipment. |
| |Operational limitations for VDL in certain conditions; very difficult to push receiver performance (in CCI |
| |area) beyond what it is today. In this respect the reconsideration of the validity of the scenarios may be a|
| |way forward. |
|DFS |Proposal of an ACR of 70dB on the 4th adjacent channel seems not feasible in the light of the AM-DSB |
| |measurements. |
| |From operational point of view it should be possible to change the value of the minimum desired signal to |
| |–82dBm. |
| |CCI assumed to be 10dB but it is not sure if this may be taken as CCI value in the presence of white noise. |
| |Need to verify that radios are compliant with requirements at –82dBm power. |
|LFV |The need to report the findings of this group to AMCP/WGB in the form of a written report. This report |
| |should also consider issues such as the value for cable loss and the definition of ACR. |
|MARCONI |Expressed approval that better understanding of DSB-AM spectral properties must be attained in order to |
| |interpret ACR results correctly. |
|TELERAD |Pointed out the difference between the cable and antenna loss assumptions used in the paper and those used |
| |in ICAO (7dB per aircraft vs 3dB per aircraft respectively) as a likely point to consider in determining the|
| |feasibility of the proposal. |
| |The possibility to tweak other physical parameters such as the transmitter antenna gain and cable loss. |
ATTACHMENT E
E.1 Calculation of required ACR for Ground scenarios provided by DFS
The following tables provide calculations of the required ACR depending on assumed protection distances for the ground scenarios when VDL Mode 4 is the victim. The calculations are based on the following formula:
ACR = P_undes – P_des + 27.5 – 20lg (F ) + L_udes + G_undes – 20lg(distance)
[dB] [dBm] [dBm] [MHz] [dB] [dB] [m]
Legend for requirements illustrated in the following tables:
|Current |
|Proposed |
|Variable |
DSB-AM interferer
|F [MHz] |P_des [dBm] |P_undes [dBm]|interferer |L_undes [dB] |G_undes [dB] |distance [m] |ACR [dB] |
|119 |-82 |50 |GS |-3 |2 |50 |83,0 |
|119 |-82 |50 |GS |-3 |2 |100 |77,0 |
|119 |-82 |50 |GS |-3 |2 |210 |70,5 |
|119 |-82 |50 |GS |-3 |2 |300 |67,4 |
|119 |-82 |50 |GS |-3 |2 |600 |61,4 |
|119 |-82 |50 |GS |-3 |2 |750 |59,5 |
| | | | | | | | |
|119 |-88 |50 |GS |-3 |2 |50 |89,0 |
|119 |-88 |50 |GS |-3 |2 |100 |83,0 |
|119 |-88 |50 |GS |-3 |2 |210 |76,5 |
|119 |-88 |50 |GS |-3 |2 |300 |73,4 |
|119 |-88 |50 |GS |-3 |2 |600 |67,4 |
|119 |-88 |50 |GS |-3 |2 |750 |65,5 |
| | | | | | | | |
|119 |-82 |44 |A/C |-3 |0 |50 |75,0 |
|119 |-82 |44 |A/C |-3 |0 |100 |69,0 |
|119 |-82 |44 |A/C |-3 |0 |210 |62,5 |
|119 |-82 |44 |A/C |-3 |0 |300 |59,4 |
|119 |-82 |44 |A/C |-3 |0 |600 |53,4 |
|119 |-82 |44 |A/C |-3 |0 |750 |51,5 |
| | | | | | | | |
|119 |-88 |44 |A/C |-3 |0 |50 |81,0 |
|119 |-88 |44 |A/C |-3 |0 |100 |75,0 |
|119 |-88 |44 |A/C |-3 |0 |210 |68,5 |
|119 |-88 |44 |A/C |-3 |0 |300 |65,4 |
|119 |-88 |44 |A/C |-3 |0 |600 |59,4 |
|119 |-88 |44 |A/C |-3 |0 |750 |57,5 |
VDL Mode 2 interferer
|F [MHz] |P_des |P_undes [dBm]|interferer |L_undes [dB] |G_undes [dB] |distance [m] |ACR [dB] |
| |[dBm] | | | | | | |
|119 |-82 |42 |GS |-3 |2 |50 |75,0 |
|119 |-82 |42 |GS |-3 |2 |100 |69,0 |
|119 |-82 |42 |GS |-3 |2 |210 |62,5 |
|119 |-82 |42 |GS |-3 |2 |300 |59,4 |
|119 |-82 |42 |GS |-3 |2 |600 |53,4 |
|119 |-82 |42 |GS |-3 |2 |750 |51,5 |
| | | | | | | | |
|119 |-88 |42 |GS |-3 |2 |50 |81,0 |
|119 |-88 |42 |GS |-3 |2 |100 |75,0 |
|119 |-88 |42 |GS |-3 |2 |210 |68,5 |
|119 |-88 |42 |GS |-3 |2 |300 |65,4 |
|119 |-88 |42 |GS |-3 |2 |600 |59,4 |
|119 |-88 |42 |GS |-3 |2 |750 |57,5 |
| | | | | | | | |
|119 |-82 |42 |A/C |-3 |0 |50 |73,0 |
|119 |-82 |42 |A/C |-3 |0 |100 |67,0 |
|119 |-82 |42 |A/C |-3 |0 |210 |60,5 |
|119 |-82 |42 |A/C |-3 |0 |300 |57,4 |
|119 |-82 |42 |A/C |-3 |0 |600 |51,4 |
|119 |-82 |42 |A/C |-3 |0 |750 |49,5 |
| | | | | | | | |
|119 |-88 |42 |A/C |-3 |0 |50 |79,0 |
|119 |-88 |42 |A/C |-3 |0 |100 |73,0 |
|119 |-88 |42 |A/C |-3 |0 |210 |66,5 |
|119 |-88 |42 |A/C |-3 |0 |300 |63,4 |
|119 |-88 |42 |A/C |-3 |0 |600 |57,4 |
|119 |-88 |42 |A/C |-3 |0 |750 |55,5 |
VDL Mode 4 interferer
|F [MHz] |P_des [dBm] |P_undes [dBm] |interferer |L_undes [dB] |G_undes [dB] |distance [m] |ACR [dB] |
|119 |-82 |44 |GS |-3 |2 |50 |77,0 |
|119 |-82 |44 |GS |-3 |2 |100 |71,0 |
|119 |-82 |44 |GS |-3 |2 |210 |64,5 |
|119 |-82 |44 |GS |-3 |2 |300 |61,4 |
|119 |-82 |44 |GS |-3 |2 |600 |55,4 |
|119 |-82 |44 |GS |-3 |2 |750 |53,5 |
| | | | | | | | |
|119 |-88 |44 |GS |-3 |2 |50 |83,0 |
|119 |-88 |44 |GS |-3 |2 |100 |77,0 |
|119 |-88 |44 |GS |-3 |2 |210 |70,5 |
|119 |-88 |44 |GS |-3 |2 |300 |67,4 |
|119 |-88 |44 |GS |-3 |2 |600 |61,4 |
|119 |-88 |44 |GS |-3 |2 |750 |59,5 |
| | | | | | | | |
|119 |-82 |40 |A/C |-3 |0 |50 |71,0 |
|119 |-82 |40 |A/C |-3 |0 |100 |65,0 |
|119 |-82 |40 |A/C |-3 |0 |210 |58,5 |
|119 |-82 |40 |A/C |-3 |0 |300 |55,4 |
|119 |-82 |40 |A/C |-3 |0 |600 |49,4 |
|119 |-82 |40 |A/C |-3 |0 |750 |47,5 |
| | | | | | | | |
|119 |-88 |40 |A/C |-3 |0 |50 |77,0 |
|119 |-88 |40 |A/C |-3 |0 |100 |71,0 |
|119 |-88 |40 |A/C |-3 |0 |210 |64,5 |
|119 |-88 |40 |A/C |-3 |0 |300 |61,4 |
|119 |-88 |40 |A/C |-3 |0 |600 |55,4 |
|119 |-88 |40 |A/C |-3 |0 |750 |53,5 |
E.2 Reviewed position provided by Rockwell Collins and Telerad
Summary
This document was previously written to establish the minimum ACR and the number of guard frequencies necessary to obtain 210m goal between aircrafts considering any type of interferer (DSB-AM, VDL2, VDL4).
After the teleconference dated 25/10/02 , we have worked from the following new assumptions:
|Attenuation(cable + antenna) = 3 dB instead of 14dB (to reflect Armin Shleret’s mail dated 28/oct/2002) |
|Space attenuation = 300m space (127,5 MHz) = 64dB instead of 60dB |
|Pmin(victim) = -82dBm instead of -88dBm |
The tests performed by DFS show that the 3 DSB-AM radios that have been tested are exceeding their MOPS by 10dB for the ACP value. Therefore, it is clear that the results of the analysis differ greatly whether ED23B ACP specs or measured ACP on real transmitters are used for the analysis:
1. We confirm that using DSBAM ACP specifications, it is not possible to derive sound criterias for frequency planning
2. If the measured ACP figures are used, then it is possible to achieve the desired frequency planning criteria at the conditions that
• the reference signal (Pmin victim) is raised from –88dBm to –82dBm,
• the spacing requirement between A/C is relaxed from 210m to 300m
In this second event, 4 guards channels will be enough
Notes :
1. This paper only deals with A/C to A/C interferences only. The assumptions for Ground stations interferer against VDL4 victim leads to the same conclusion because the additional 6dB on the interferer power (50dBm instead of 44dBm) and the additional 2dB on the antenna+cable attenuation are exactly compensated by the -8dB resulting from the 750m distance instead of 300m distance
2. It is a question of judgement whether we decide to work from measured data rather than specs. Measurements have been made on a limited number of equipment (3 manufacturers), on single Serial numbers (how repeatable), with specific loads (how representative to A/C installation), only in laboratory temperature conditions...
3. The new assumptions above leads almost to the same conclusion than before, because the additional 11dB of the cable+antenna attenuation (14dB instead of 3dB) are compensated by the +4dB increase of the space attenuation (64dB instead of 60dB) and the +6dB increase of the Pmin(Victim) (-82dBm instead of -88dBm). So the demonstration remains almost unchanged.
4. The document was modified using the MicrosoftWord modification tools in order for the reader to easily see the changes.
Introduction
The purpose of this paper is to comment the paper from M. Desperier dealing with recommendations on VDL4 Adjacent Channel Rejection performance dated 26-29 august.
This paper establishes minimum ACR and the number of guard frequencies necessary to obtain the 210m 300m goal between two aircrafts, considering any type of interferer (DSB-AM, VDL2, and VDL4).
This paper also asks for an improvement on the ED23B (DSB-AM MOPS), that is necessary to get consistency between VDL4 (or VDL2) system and DSB-AM transmitters characteristics.
ACR versus CCI :
Adjacent Channel Rejection (ACR) and Co-Channel Interference (CCI) are two different performances that are generally tested separately.
ACR definition :
The ACR is used to test the susceptibility of the receiver to a strong signal on a adjacent channel. Typically, the VDL4 MOPS (ED-108) ask for 40dB ACR on the 1st adjacent channel, and 60dB ACR on the 5th adjacent channel. The ACR is tested without any interferer signal transmitted within the pass band.
ED108 Note 1 page 76 clearly indicates that the test is performed without noise in the pass band:
2. Adjacent Channel rejection (test procedure)
The noise side band of the interfering signal must not interfere with the desired signal in the desired pass band….
CCI definition :
The CCI is used to test the susceptibility of a receiver to a interferer signal transmitted into the pass band. Typically, the MOPS VDL4 (ED-108) asks for 10dB CCI.
Combining ACR test and CCI :
The test performed par M.Desperrier consists in transmitting a interferer signal with a real transmitter. This test combines both ACR and CCI characteristics, the CCI being generated by the noise of the interferer transmitted into the pass band.
Although this test is near to the real life, it is not easily reproducible and its results may more depends on the unknown noise characteristics of the interferer, than on the ACR performances that are to be measured.
What we observed :
When performing the ACR test, if the pass band is not jammed by any interferer wide band noise, we could not notice any discrepancy when using different interferer modulations.
So we do not confirm that the type of modulation has an impact on the ACR performance.
We think that M.Deperier used different kind of interferer with different noise and spurious characteristics. And this leads to wrong conclusion, because the ACR test must be performed without any signal present in the pass band. We think that the results found by M.Desperier are more reflecting the CCI characteristics of the VDL4 receiver combined with the noise characteristics of the interferer, than the ACR characteristics of the VDL4 receiver.
When performing interference test, we must use a 100% load otherwise, otherwise the measurement is not stable and the results may not reflect the worst case of synchronous frames.
General case
When a interferer and a victim are transmitting in the same area, what are the performances necessary to avoid adjacent channel interferences?
The relative value of the spurious of the interferer must not be higher than the desired low signal :
|ACR(victim) + CCI(victim) < ACP(interferer) |(1) |
This relation means that it is not necessary to have a ACR(victim) greater than ACP(interferer) – CCI(victim) otherwise, the low desired signal is jammed by the spurious of the interferer.
When writing the absolute value of the spurious, we get :
Pmin(victim) – SPURIOUS(interferer) > CCI(victim)
Pmin(victim) - [ Pmax(interferer) – Attenuation(cable + antenna + space) – ACP(interferer) ] > CCI(victim)
|Pmin(victim) - Pmax(interferer) + Attenuation (cable + antenna + space) + ACP(interferer) > CCI(victim) |(2) |
Relations (1) and (2) can be illustrated as follows :
[pic]
According to the following assumptions :
|Pmin(victim) = -882dBm |According to ED-108 says -88dBm |
|Attenuation(cable + antenna) = 2x7dB = 14 3 dB |According to RTCA DO224 |
|Attenuation(space 127,5 MHz) = 604dB |Desired 21300m between aircrafts |
|CCI(victim) = 10dB |According to ED-108 |
Relation (2) can be written :
|ACP(interferer) - Pmax(interferer) > 24 25 dBm |(3) |
In the worst case, we would have: ACP(interferer) - Pmax(interferer) = 24 25 dBm
And then (1) can be written
|ACR(victim) + CCI(victim) < 2425dBm + Pmax(interferer) |(4) |
DSB-AM interferer against VDL4
What is the worst case of a DSB-AM interferer against VDL4 ?
(3) ACP(DSB-AM) - Pmax(DSB-AM) > 2425 dBm
According to ED23B the maximum power of a DSB-AM transmitter is 44 dBm.
In this case, the relative ACP is 60dBc. So (3) can be written :
(3) 60dBc – 44dBm > 2425dB
(3) is false, so the ACP of the DSB-AM must be improved.
The minimum ACP of the DSB-AM must be 6869dBc (=2425dB + 44dBm)
Remark :
This demonstration does not depend on the guard band, because the bad ACP of DSB-AM transmitter can generate spurious anywhere on the band. So even with 20 guard band channels, this problem is still present.
Conclusion :
To determine a number of guard channels, the ED23B must ask for a more stringent ACP performance
This ACP must be 6869 dBc (instead of 60dBc) in the case of a 44dBm transmitter
This ACP must be 6465 dBc (instead of 56dBc) in the case of a 40dBm transmitter
Improved DSB-AM interferer against VDL4
In the case where the ACP performance of the ED-23B is increased, what should be the ACR value of the VDL4 on the 5th channel ?
According to (1), the minimum ACR(VDL4) is
ACR(VDL4) < ACP(interferer) - CCI(VDL4)
ACR(VDL4) < 6869 - 10
ACR(VDL4) < 5859 dB (on the 5th channel)
Conclusion :
If the ED23B ACP value is better than 6869dBc (for 44dBm, or 6465 for 40dBm) on the 5th channel, then 5859dB ACR performance is enough to avoid interferences with VDL4.
VDL4 interferer against VDL4
What should be the ACR value considering VDL4 interferers ?
What is the number of guard channels in that case?
Assumptions
• The VDL4 transceiver output power is 43dBm
• The CCI(VDL4) is 10dB
• The channel load is 100%.
Demonstration:
1Guard channel : The VDL4 SARPS asks for an ACP of -28dBm on the 2nd channel.
The relative ACP is:
ACP(VDL4) = 43dBm-(-28dBm) = 71dB
According to (3) we must have : ACP(interferer) - Pmax(interferer) > 2425 dBm
71dB – 43dBm > 2425dBm
28dBm > 2425dBm
We find that (3) is true, so 1 guard channel is enough.
According to (4) , the minimum ACR(VDL4) is then :
ACR(victim) + CCI(victim) ≤ 2425dBm + Pmax(interferer)
ACR(VDL4) + 10dB ≤ 2425dBm + 43dBm
ACR(VDL4) ≤ 57 58dB
Conclusion :
• 1 guard channel is enough to avoid interferences.
• With 1 guard channel, 57 58dB ACR performance is enough to avoid interferences at 210300m.
VDL2 interferer against VDL4
What should be the ACR value considering VDL2 interferers ?
What is the number of guard channels in that case?
Assumptions
• The VDL4 transceiver output power is 42dBm (this is the only difference with VDL4 interferer)
• The CCI is 10dB
• The channel load is 100%.
Demonstration:
1Guard channel : The VDL4 SARPS asks for an ACP of -28dBm on the 2nd channel.
The relative ACP is:
ACP(VDL4) = 42dBm-(-28dBm) = 70dB
According to (3) we must have : ACP(interferer) - Pmax(interferer) > 24 25 dBm
71dB – 42dBm > 24 25 dBm
28dBm > 24 25dBm
We find that (3) is true, so 1 guard channel is enough.
According to (4) , the maximum ACR(VDL4) is then :
ACR(victim) + CCI(victim) < 24 25 dBm + Pmax(interferer)
ACR(VDL4) + 10dB < 24 25 dBm + 42dBm
ACR(VDL4) < 56 57 dB
Conclusion :
• 1 guard channel is enough to avoid interferences.
• With 1 guard channel, 56 57 dB ACR performance is enough to avoid interferences at 210 300 m.
Our recommendations
Test methods:
We recommend to distinguish the ACR test and the CCI test.
For the ACR test, we suggest to use the existing test method of the ED-108
FM 400HZ sine wave, +/-5.25Khz deviation. This spectrum was used for the VDL2 interference testing, for the following reasons :
• It is easy to generate and reproduce
• It looks like VDL2 spectrum
The VDL4 spectrum is slightly different, but not enough to justify a different interferer spectrum.
Improvement of ED23 ACP performance:
We recommend to improve the ACP performance of the ED-23B from 60dB up to 6869 dB for 44dBm output power, and 64 65 dB for 40dBm output power, so that new DSB-AM transceiver will no more interfere with the low level VDL4 signals (-88 82 dBm). Till this ED23 ACP characteristic is improved, then the goals of 210 300m, and -88 82 dBm minimum power, cannot be achieved, whatever the number of guard channels is.
Note: this issue is NOT specific to VDL4, but also applies to VDL2.
ACR Value :
Assuming that the ED-23B ACP performance is improved as described above, then the ED-108 ACR performance should be 5758 dB on the 2nd channel (to cope with VDL2/VDL4 transmitters) and 58 60 dB on the 5th channel (to cope with DSBAM transmitters), in order to avoid interferences when the 2 aircrafts are separated by 210 300 m.
The current ED-108 value of 60dB at the 5th channel is more conservative than equal to the above recommended value and can remain unchanged.
The only new ED-108 requirement that would allow a single guard channel from VDL2/VDL4 transmitters is the “57 58 dB on the 2nd channel”. From the testing activity performed by Collins in Telerad on a Telerad receiver, we can already say that this performance is possible to achieve. But in order to later discuss the impacts of this new requirement with all the manufacturers, we suggest that they all test this performance with their equipment, and using the test method of the ED108.
Note : This ACR performance of 60dB does not depend on the minimum power (-82dBm or -88dBm). This can be understood from the above, and this was confirmed by testing in our laboratories.
Number of Guard channels
Assuming that the ED-23B ACP performance is improved as described above, and that the recommended ED-108 ACR new requirement is adopted, the number of guard channels is as follows :
• 1 guard channel for VDL4 and VDL2
• 4 guard channels for DSB-AM.
Comments on the tests perfomed by DFS on DSB-AM transmitters
According to the tests performed by DFS, the 3 DSB-AM radios exhibit a relative ACP of about 70dBc on the 5th channel. So according to our conclusion above (improvement of ED23ACP performance), 69dB would be enough to have 4 guard channels (with 300m, and 82dBm assumptions).
ATTACHMENT F
From: Armin.Schlereth@dfs.de
To: nikolaos.fistas@eurocontrol.be; nikolas.fistas@eurocontrol.be; bertrand.desperier@eurocontrol.int; RWitzen@icao.int; r.witzen@videotron.ca; MICALLEF John
Cc: Otto.Lindenau@dfs.de
Subject: Requirement of -82 dBm for VDL 4 – Rationale
Sent: Mon 28/10/2002 10:09
Dear all,
enclosed is an Excel sheet including the calculation for the required ACR dependent on assumed protection distances for Ground and Air scenarios discussed. The critical ones are the Ground scenarios. You will find all possibles scenarios dealing with type of interferer (DSB-AM, VDL2,3,4) on VDL Mode 4 victim. I guess that is the critical case we are discussing.
Let's take a look at the Ground Scenario with DSB-AM as interferer:
In case we have a DSB-AM Ground Station (GS) as interferer the required distance is 750 m. This is fulfilled for -82 dBm desired input power (value for ACR in the table: 59,5 dB).
In case we have an airborne DSB-AM interferer for -82 dBm desired input power and a protection distance of 300 m we have an ACR of 59,4 dB and so on.
So also the other cases show that an ACR of 60 dB is sufficient as long as -82 dBm desired input power value could be agreed on and that a separation distance of 300 m is also acceptable.
For that purpose we have to discuss the scenarios on the GROUND a little bit more in detail. For that 2 cases shall to be discussed for the DSB-AM interferer case.
Scenario with VDL 4 ADS-B:
Assuming that the desired input power coming from another aircraft is well above -82 dBm. This makes sense due to the fact that it could be assumed that aircraft within a range of 300 m have a much higher power value anyway. Therefore 300 m separation distance is ok. For those aircarft within line of sight the distance would be more than 50 nmi (about 60 nmi) to create -82 dBm desired input power value. So nobody of the mobile units on the ground really care about that data coming from A/C beyond 50 nmi range. Only problems might occur from other aircrafts on the ground , which are covered by buildings or other aircrafts. But taking into account the data of the WP enclosed [4], it is not likely that an aircarft closer than 300m
would create less than -82 dBm desired input power. Therefore: -82 dBm desired singal power with minimum range of 300m is acceptable.
Scenario with VDL 4 Com point-to-point:
Here at least the same requirements as for VDL Mode 2 should hold regarding the minimum desired power value (-82 dBm), because a VDL Mode 4 ground station transmits with 4 dB more power and we assume in addition 2 dB of antenna gain compared to a VDL 4 transmitting aircraft (So even 2 dB more than a VDL Mode 2 GS). Regarding the protection distance of an aircraft interferred by another aircraft with DSB-AM a minimum distance of 210 m is assumed for VDL Mode 2. From my point of view it doesnt make really a difference whether there is 210 m or 300 m of separation. If there would be an aircarft just behing You at the holding point and it is communicating via DSB-AM with a ground station the aircraft in front of it would not be
able to receive VDL Mode 4 or VDL Mode 2 data. So what makes really the difference? For that scenario the desired VDL Mode 4 power have to be 10 dB higher anyway. For VDL Mode 2 it is not much less. I am not in the position to postulate the 300m here, I think WG-B has to
decide on it. So as Robert Witzen is now chairing this group he should be involved in the discussion soon. But I think 300 m is also appropriate.
I would propose that WG-B should have a telecon as well to discuss the issue.
................
................
In order to avoid copyright disputes, this page is only a partial summary.
To fulfill the demand for quickly locating and searching documents.
It is intelligent file search solution for home and business.
Related download
- chapter v site environment supervision
- hankook tire media center press room europe cis
- development of frequency planning criteria for vdl mode 4
- cept erc recommendation 00 00 e city year
- tcg turkana county
- 1 总论 world bank
- external noise in residential buildings consultation
- review of heathrow s noise mitigation schemes
Related searches
- criteria for hypertrophy on ekg
- criteria for demand ischemia
- criteria for lvh
- budapest clinical criteria for crps
- budapest criteria for crps 1
- budapest criteria for crps pdf
- minimum voltage criteria for lvh
- ekg criteria for rvh
- criteria for crps
- budapest criteria for crps diagnosis
- budapest diagnostic criteria for crps
- criteria for fha home loan