VDL Mode 4 Airborne Architecture Study. Comments from the External Review Group

Transcription

1 VM4AAS VDL Mode 4 Airborne Architecture Study Comments from the External Review Group WP1 - Deliverable D1: Definitions, Assumptions and Baselines... 1 WP2 - Deliverable D2:Identification of Requirements (Joint D2.1 & D2.2) WP3 - Deliverable D3.1: Architecture Descriptions WP3 - Deliverable D3.2: Radio Frequency Interference Analysis WP4 - Deliverable D4: Implementation and Transition Considerations WP5 - Deliverable D5: Overall report i version 1.0

2 VM4AAS (Intentionally left blank) ii version 1.0

3 VM4AAS- Comments on Deliverable D1 WP1 - Deliverable D1: Definitions, Assumptions and Baselines COMMENTS AND SUGGESTIONS AND THEIR RESOLUTION No where Comment Resolution A Tbl 2 Fix Airborne enhanced surveillance per reference 7. Airborne enhanced surveillance in agreement with requirements. 1. I note that there are no assumptions relating to requirements for company communications or for redundancy for commercial purposes (continued despatch following a failure). 2. Similarly there are no assumptions relating to the need for concurrent use of any combination of equipment without unacceptable interference. 3. Is there a need to state assumptions relating to the number of flight crew presumably two? Airframe classes. It would be helpful to indicate the typical size range of the classes defined, as size is of more significance than weight for this study. I guess we are also making the assumption that construction is conventional aluminium alloy structure (affects ground planes for aerials). 5. For the retrofit study cases the basic avionics technology generation (e.g. analogue, digital, IMA) will also have significant influence on the architecture and cost. How will this be covered? Ref 3.3, the description for assumption 16 makes it clear that it is also assumed that the flight management system provides intent information, and is connected to the Modify discussion Assumption 7 to include "redundant voice capability may be used for company communications or backup of primary ATC voice communications. Pick up B. Tremans r/l/c allocations in discussion New assumption [9a] on concurrent operation. "Data Link, ADS, and both voice channels are assumed to operate simultaneously. In discussion of assumption indicate, this study will indicate acceptable/unacceptable interference and any cooperative actions required. For a Large aircraft, then, the baseline equipment includes dual voice, dual SSR surveillance, and a communications data link capability, all of which must operate simultaneously (Assumption 9). This communications equippage is sufficient to support a variety of flight crew configurations. No other assumptions are made about the actual configurations of the flight crew which may be used by the airline or aircraft operator. Added this text to 3.3.1: The definition of aircraft classes will be reviewed as the study progresses. Minor adjustments may be made to the class definitions should they be warranted by the efforts of WP 2 and WP 3. After discussion within the Honeywell and Eurocontrol teams, it was decided that by 2008 the overwhelming majority of aircraft will be mostly, if not entirely, digital interfaces. Therefore, the study will only consider "digital". IMA vs. non-ima is an architecture decision to be considered in the study. RNAV is a lesser capability than FMS, and so is identified separately. The fact that existing FMS may need extensive modification to allow 1 version 1.0

4 VM4AAS- Comments on Deliverable D1 No where Comment Resolution VDL Mode 4 equipment. Many current aircraft that have FMS systems would require considerable expenditure to meet this, so it should be stated. enhanced and intent-based surveillance is attributable to the ADS application, not to VDL Mode 4. That is, any ADS implementation (1090ES or UAT included) would require such modifications. Therefore, these modifications, while important, are over and above considerations related to the VDL Mode 4 architecture. Statement still open through WP3. FMS integration issues are common to ALL implementations of intent-based surveillance and not specific to VDL Mode Tables 3-5 VDL mode 3 is not mentioned, even though assumption 14 mentions it. Is this as intended? In the newly renumbered Assumption 15, the following text is included: It is important to note that VDL Mode 3 is mentioned only the context of inclusion in the basic ARINC 750-X characteristic. This study does not specifically consider VDL Mode 3 as an alternative for either voice or communications data link functionality. The study methodology could be extended in a straightforward manner to assess VDL Mode 3/VDL Mode 4 integration issues, but these issues are beyond the current scope of effort. 8. Where ILS is mentioned should it be clarified that this includes category III automatic landing? In the descriptions of the Equipment Baselines, the following text is inserted: 9. Page 11, S3 Table 1 Assumption 8: This assumption may, under certain operational scenarii, be valid, however it potentially implies significant operational constraints on the use of VDL Mode 4. If we are to achieve the most significant potential operational benefits of VDL Mode 4, then major changes to operational conditions must be expected, for example aircraft separations, degree of cockpit delegation etc. Since, we must assume, these operational conditions are not feasible with voice-based ATC (otherwise we should be implementing them today) reverting to voice-based ATC on loss of datalink must be regarded as a problem. To suddenly, but smoothly, switch from datalink-enabled ATC conditions to voice-based would imply that the datalink capability is redundant. To put it simply, the operational scenario may not be feasible with voice-based ATC. Another aspect of this issue is that of spare resources. If we have the capability to Large, Small, and Light aircraft are assumed to have ILS Category III, Category II, and Category I performance, respectively. Clarification in discussion of 8 that redundant voice is ultimately a safety, ie. risk mitigation issue. "Fall back" rather than "back up". 2 version 1.0

5 VM4AAS- Comments on Deliverable D1 No where Comment Resolution 10. Page 11, S3 Table Page 12, S Page 12, S3.1 switch to voice-based ATC, then that implies that, at least spectrum capacity, and additional ATC Controller resources are instantly available. The only was to achieve this is to keep them in reserve as unused spares. In neither case is this an economically attractive proposition. Assumption 14: Change to "Radios built to an ARINC 750-X Characteristic are available and provide Mode 2...", i.e. "are" instead of "will be". I believe the radios are available and certified, as are compatible CMU's. In Section 3.3, mention should be made also of ARINC 758, which is designed to accompany 750. text on Assumption 3: We need to be careful here, both with terms and with concepts. It is certainly correct that ADS-B will not replace ACAS as a collision avoidance safety net in the foreseeable future. That does not mean, however, that ADS-B does not play a role in the tactical domain, including conflict detection and resolution, ASAS and such like. ACAS is the last line of defence; if the other kit works, then ACAS doesn't kick in. text on Assumption 4: Second last line; change "if one of the links is ADS-B" to "if one of the links is Mode S". No change to current text. The important point in this paragraph is definition of the ARINC 750-X capability, ie what functions are available in the radio. Mode 3 is beyond the scope of this study. Changes to 758 will be considered as appropriate to the study, but not included in discussion at this time, or made as part of the assumption. Clarify assumption with "There is no intent to preclude enhancement of the ACAS function from using ADS-B derived information, but this is an ADS benefit, not a necessarily a VDLM4 benefit." Change ADS-B in discussion Mode S. 13. Page 13, S Page 13, S Page 13, S Page 14, S Page 14, S3.2.3 text on Assumption 9: Propose deleting the text "GBAS is not required" in the title; i.e. just call it "GBAS uplink is not a data link function". The two parts of the title are inconsistent. middle of the Assumption 9 text: Don't like the reference to "experts". Too vague, too subjective. Other "experts" might disagree. Better to provide an objective reference. Last sentence "...interference consideration...": If we accept that there may be an issue with GBAS interference, then we should not, a priori, rule out other solutions, such as uplink via VDL Mode 4, which could be a very convenient alternative at very low incremental cost. Propose retaining this option in some form. After discussion of Assumption 10; propose changing "will be implemented as described in [4,5]" to "will be implemented as described in [1,2,4,5]". SARPs restriction to Surveillance does not affect (significantly) the technical implementation. The SARPs are still, from a purely technical perspective, the defining documents on communications with VDL Mode 4. Last sentence: Text says "ATN-compliant data link transactions are performed via VDL Mode 2". The timeframe is critical here. It may well be that means other than Changed in main document. Remove "GBAS not required". The text has been reworded. In accordance with ICAO, GBAS is not supported by VM4. References are consistent in newly numbered Assumptions 11 and 12. The comment misinterprets the purpose of Section In this section, and in this Operational Configuration, the fundamental 3 version 1.0

6 VM4AAS- Comments on Deliverable D1 No where Comment Resolution Mode 2 are used in due course for ATN-compliant data link transactions (for example, because we find that VDL Mode 2 requires excessive spectrum, i.e. it is spectrum-inefficient). We propose either making it clear that the comment here refers to the immediate present only (during such time as very few aircraft actually use VDL Mode 2) or deleting. assumption is that VDL Mode 4 is not an acceptable means of providing communications data link functionality. In this case, some other data link must be assumed. In Europe, the only reasonable alternative assumption is VDL Mode 2. Just as this study does not seek to judge the merit of VDL Mode 4 for a given application, neither does it pass judgement on VDL Mode 2 performance or implementation. Architectures based on the assumption that VDL Mode 4 is an acceptable means of providing communication data link functionality are included in the Operational Configurations of and Page 19, S3.4.3,Ta bles Page 11, Table P14, S3.2.2,-3 We need to be very careful in the document as to whether we are talking about functions or boxes (i.e. LRU's). Up to this point in the document, we have, in my view been talking about functions. Tables 3-5 now look as if they're talking about boxes. The point is that several functions can be combined in one or a small number of LRU's, thereby enabling sharing. The concept of multimode COM/NAV receiver is potentially extremely important for this study. It may, for example, provide a solution to the multiple-antenna/multiple RF path problem (by effectively sharing antennas and RF paths between different VHF fnctions). Refer, for example, to presentation made to AMCP recently. Assumption 16: Are FMS, and RNAV capability equivalent? The set of assumptions is a little dificult to understand; even though I see what the concept is, they seem to have been stated in a way that does not make it very clear what the essence is. Example (from page 14) shows a large number of cross references, nested functions, etc, without substance matter. This makes it quite difficult to apprehend the statement of the assumption (see copy below). ===quote=== VDL Mode 4 for Data Link Communications (but not Surveillance) This configuration is defined by Assumption 10 and Assumption 12. Assumption 10 VDL Mode 4 is used for data link communication. See Section for comments regarding Assumption 10. The comments regarding Assumption 13 also apply to this operational configuration. Assumption 12 VDL Mode 4 is not accepted as a means of surveillance. Tables 3-5 are intended to show what is already installed on the aircraft prior to the addition of VDL4, that is, the BASELINE for the study. The most efficient means may well be to remove an ARINC750 radio and replace it or upgrade it to ARINC750-y or some integrated radio function. That is the subject of the study. Tables 3-5 are intended as the starting point, not the finishing point of the study. We are well aware of the AMCP presentation, and it is under consideration as a source for this study. Text added to Assumption 17: Upgrades of both FMS and RNAV capability may be required to support ADS-B operations is attributable to the ADS application, not to VDL Mode 4 in particular. That is, any ADS implementation (1090ES or UAT included) would require such modifications. The information contained within Section 3 has been extensively reorganized. 4 version 1.0

7 VM4AAS- Comments on Deliverable D1 No where Comment Resolution Assumption 12 is the configuration-specific implementation of Assumption VDL Mode 4 for Surveillance (but not data link Communications) This configuration is defined by Assumption 11 and Assumption 13. Assumption 11 VDL Mode 4 is used for surveillance. See Section for comments regarding Assumption 11. Assumption 13 VDL Mode 4 is not accepted as a means of data communications. Assumption 13 is the configuration-specific implementation of Assumption 6. Because VDL Mode 4 is not used for data link communication, and because Assumption 6 states that data link is used for ATS/ATM, and because Eurocontrol has endorsed VDL Mode 2 for data link applications, this study will assume that ATN-compliant data link transactions are performed via VDL Mode 2. ===end of quote=== 21. P12, S3.1 From assumption 3, 5, "...The use of ADS-B is not likely to replace ACAS for tactical warning and threat resolution in the near-to-mid-term future (2 to 10 years)." Change the wording in assumption 3 to be "ADS-B "will not replace ACAS for tactical warning..." This assumption appears to consider the possibility of ADS-B assuming ACAS functions, the only issue being "when". I take strong opposition against that and the recent accident over Germany hopefully accentuates this. We shall never mix, or confuse, or even bring close to one another a primary and a monitoring channel. This is the case in autopilot designs as well as in air traffic system architectures. I would prefer to see a much stronger text stating something like "... ACAS (TCAS) and ADS(-B) are dissimilar functions which have to generate and integrate their information in dissimilar ways and from dissimilar sensors (sources) and as such there is the requirement to not consider merging or fusing their data". The accident over Germany showed how seriously we should take this. If TCAS and ground services would have operated from the same data set, there would not have been a TCAS RA different from what the ground conflict probes signalled (or did not, in this case). The fact that a collision happened after all, now appears to have been caused by opposing commands between the TCAS and the ATC vectors. This points to a discussion on flight crew procedures as well, but in the context of this discussion, that only complicates the matter. 22. S In the VDL-4 Baseline cases (chptr 3.4.3), in the accompanying text, there is no mention of the concurrent VHF services on the aircraft (VHF Comm, VDL2, 3). Fortunately in the matrices (Table 3) they are. These VHF services are by far the complicating matter in VDL-4, not the co-existing Mode-S transponders, although they do relate but only from the surveillance data perspective. This comment is addressed by the response to comment version 1.0

8 VM4AAS- Comments on Deliverable D1 No where Comment Resolution 23. P6 A description of the different WPs should be included in the ES to set the scene Add a short description of the four major work packages to the Executive Summary. 24. P8 Move reference 5 after reference 2 The reference list has been updated and reorganized. 25. P10 Include and consider paper by SH submitted in AMCP/WGC meeting in May (WP32) Comment accommodated by response to # P10, S3 Remove reference to CFT Editorial 27. P13, S3.1 Assumption 9: Correct title to reflect Table 1 text Editorial 28. P3 and 4,T1, S3.1 Assumption 3: Need to state what are the current requirements about redundancy for the SSR Transponder. For what a/c do we need to have dual equipage? It is not very clear the distinction between equipment and function/application Current FAA and JAA standards require only a single transponder. AT aircraft typically carry dual transponders for the purpose of allowing dispatch in the event of a failure. Functions/applications/configuration are consistent based on the new definitions in Section P11, T1 Assumption 4: SSR TRX and 1090ES are not the same. Why the presence of 1090ES satisfies assumption P11, T1 Assumption 8: Maybe to split this assumption in two new ones: no redundancy for LINK2000+ type of applications and redundancy for the next set of applications 31. P11, T1 Assumption 9: Simplify the assumption to: GBAS uplink is not a data link function to be supported by VDL Mode P11, T1 Assumption 13: specify that applicability is for COM type of applications. What about datalink COM communications 33. P11, T1 In relation to assumption 14: We need to address the issue of VM3. In Europe, VM3 is not seriously considered. We can assume that it will not be used. Q: Does this assumption make the VM4AAS useless in the current thinking of US? Need to discuss on this. After discussion between Honeywell and Eurocontrol team members, Assumption 4 has been deleted. The old Assumption 8 has been split into two new separate assumptions. Done. The term communications data link (as opposed to surveillance data link and navigation data link) is used consistently throughout the document. In the newly renumbered Assumption 15, the following text is included: It is important to note that VDL Mode 3 is mentioned only the context of inclusion in the basic ARINC 750-X characteristic. This study does not specifically consider VDL Mode 3 as an alternative for either voice or communications data link functionality. The study methodology could be extended in a straightforward manner to assess VDL Mode 6 version 1.0

9 VM4AAS- Comments on Deliverable D1 No where Comment Resolution 3/VDL Mode 4 integration issues, but these issues are beyond the current scope of effort. 34. P11, T1 Assumption 16: high-end a/c is not a term used later in the a/c classes. Need for consistency 35. P11, T1 The ECTL policy is that VM2 will be the first datalink to support LINK2000+ applications in the first phase. Therefore, the use of VM2 to support these early applications is an assumption. Is this clear in the study. VM2 is the link for the 1 st phase in a timeline. For the second phase, it could be, VM2 or VM4 or both (we could make a link to applications) High end is defined in Section 1.2. Terminology is consistent throughout the document. Equipment Baselines 2R assumes that VDL Mode 2 is already installed. Operational Configurations 1 and 2 assume the use of VDL Mode 4 for communications, given the Equipment Baselines. Because it is a given that VDL Mode 4 is added, the rationale as to why VDL Mode 4 is added seems beyond the scope of the study. No action, closed. 36. P11, S3.1 1st paragraph: need definition of baselines Equipment Baseline and other terms are defined in Section P11, S3.1 So the purpose of section 3.1 is to justify (provide references) the assumptions made. Do we need to make this more explicit in the introduction? The information in Section 3 has been extensively reorganized. 38. P12, S3.1 Assumption 3: references to a/c classes should be consistent. The terminology has been made consistent throughout the document. 39. P13, S3.2 Need to agree on names : Surveillance and Communication datalink applications? The terms communication data link, surveillance data link, and navigation data link are defined in Section 1.2 and used consistently throughout the document. Surveillance and Communication datalink? (Need to refer maybe to C N S??) Actually in Section this has been implemented 40. P13, S3.2 Need a section in which the Operational Configurations will be clearly defined. In Section 3, they are referred to as study baselines Propose to put this section in the beginning together with a description of the scope of the study. The term Operational Configuration is defined in Section P14, S P14, S3.2.1 What is the meaning of Assumption 10 is the configuration-specific implementation of Assumtion 6? (also later) Can we make it clearer? What is the impact if instead of ATN IP is used in VM4? The material of Section 3 has been extensively reorganized, and the subject text has been removed. The impact will involve the traffic load on the VDL Mode 4 link. Accepted procedures exist (see DO-270 and/or DO-277) to relate user load to ATN load. Similar rules-of-thumb exist for IP. The 7 version 1.0

10 VM4AAS- Comments on Deliverable D1 No where Comment Resolution assumption for this study is ATN, therefore this may be a topic for future study. No change to the document. 43. P14, S3.2.1, S P17, S P18, S P18, S3.4.3 Need a section similar to the last paragraph of S3.2.3 about the use of VM2. Need maybe to discuss the applications Need more introduction: After the 1 st paragraph to explain the importance of retro-fit and forward-fit. Class 2R: suggest to mention that the 1 st paragraph is exactly the same as in Class 1R (it makes the reading more friendly) Class 2R: What is the differences of Classes 1R and 2R? I assume that VM2 is already on the plane when VM4 is installed in 2R. The first sentence of the 2 nd paragraph is not clear. The material of Section 3 has been extensively reorganized, and the subject text has been removed. New text is "The tables are divided into retrofit and forward fit cases. The term "retrofit" refers to aircraft that are already in operation with a given equipment list to which VDL Mode 4 capability is to be added. The term "forward fit" refers to new aircraft that are produced with the established capability plus VDL Mode 4 at the time of delivery." The first paragraph of Equipment Baseline 2R explicitly states the similarities with Equipment Baseline 1R. The first paragraph of Equipment Baseline 2R explicitly states the similarities with Equipment Baseline 1R. 47. Use of the word class : a/c classes, baseline cases Can we make the reading easier? The terminology has been defined in Section 1.2 and made consistent throughout the document. 48. P18, S3.4.3 Class 1F/2F: Let s call it with one name. Use of word class? Class 1F/2F is replaced with "Class F". 49. T3 and 4 Why are there some discrepancies between the two classes? In particular Voice 2 for light a/c, Tpdr#2 for small a/c and Mode C for small a/c After discussion within the Honeywell and Eurocontrol teams, it was decided that by 2008 the overwhelming majority of aircraft will be mostly, if not entirely, digital interfaces. Therefore, the study will only consider "digital". IMA vs. non-ima is an architecture decision to be considered in the study. 50. We are assuming one baseline class for surveillance. Is this sufficient For the current study, one baseline surveillance class is sufficient. 51. There is no stated assumption on AOC, and I believe there should be. Does the study assume that AOC will be done on current VHF, VDL-2, VDL-4, satcom, HFDL, or a mix? This is critical to any conclusions, and should be stated explicitly for all scenarios. SOLUTION: This seems to be a WP2.1 issue, i.e., a functional requirement on the data link to do AOC or to not do AOC. It does not appear to be an underlying assumption. The issue to be dealt with in WP2.1 and not here. 8 version 1.0

11 VM4AAS- Comments on Deliverable D1 No where Comment Resolution State the assumed AOC environment, including expected end-end support (or not) for VDL-4 AOC on the ground. Specifically, will a/c need to be able to operate simultaneously and/or switch between VDL-4 and VDL-2 / VDL-x (current ACARS) / something else? 52. Assumptions for the a/c systems above the comm should be more clearly spelled out. Assumption 15 states something on not considering surveillance display and fusion, but what is the situation for ATS (and AOC?) point-point data comm? If the study is limited to comm architecture only, then you would still expect to see e.g. a CMU or MU listed in the equipment fits to serve the ATS/AOC switching needs as you move from VDL-4 region to non-vdl-4 region. If it is expected to cover other areas above that, then it will impact other systems. For example, CDTIs are always assumed to be part of Baseline classes. Does this imply that the only means of ATS and/or AOC data comm interface for VDL-4 is a CDTI? The operational configurations have been made more clear. The study assumes VM4 for both COM and SUR. In relation to the issues for the equipment (CMU, ATSU, etc), this seems an issue for WP2.2 Architecture Requirements, and is not an underlying assumption. The issue will be examined WP2.2 SOLUTION: I'm not sure on this one, but I think a start to finding out how to solve it would be: a) State clearly that the study is looking only at comm architecture issues, if that is the case. State this up front. b) If comm architecture is the only part being studied, then at some point a list of interface points would need to be defined for the VDL-4 applications. This would at least allow you to identify impact on systems expected to switch to/from VDL-4 & VDL-2 by area, and e.g. switch between CDTI and other systems if point-point is expected to run through CDIT per existing VDL-4 prototypes (FYI, I am *NOT* advocating the latter configuration at all, just listing it as an example of what is missing). This list of interface points at least allows you to see what has NOT been studied. c) If the study is to go to more than comm, then more work will obviously need to be done and the tables in section will get expanded considerably, 53. This will be unpopular, but this is not a political study as I understand so I will raise it. I think a critical assumption, in light of the FAA decision (like that decision or not- I have no opinion), and the ADS-B activities in other parts of the world, is that the There is no requirement for dual transponder (i.e. dual surveillance) implementation. It was agreed within the team that ADS would not be a replacement for ACAS in the time frame of the study, and that ADS- 9 version 1.0

12 VM4AAS- Comments on Deliverable D1 No where Comment Resolution functions served by VDL-4 will be served by other systems elsewhere. This is partially stated in assumptions 3 and 4, but I believe they need to be much more explicit. If requirements are for e.g. two independent ADS-B systems, and only one of those is common (Mode S), then you need to have three ADS-B systems on an a/c that operates inside and outside VDL-4 areas. SOLUTION: State the assumption that one ADS-B system is going to be Mode S (I believe this is a safe assumption??), and that if another is required for dual redundancy that will be UAT in the US and VDL-4 in Europe, so a/c must carry both. only dispatch was, therefore, unlikely in that time frame. Given these agreements, there is no requirement for dual surveillance of any kind and we don't need the recommended solution. There is consensus opinion of all I have talked to that dual surveillance (i.e. dual transponders) are a practical matter related to dispatch efficiency, not safety or regulatory requirements. 54. S These paragraphs state the assumption that all ATS data comm will be ATN. I understand that that is the aim, but this is a technical study so should deal with the actuals not the intent. In light of the stated timeframe, I believe it is incorrect. Stated timeframe is 2008, and LINK has just stated that FANS will be on the migration path to ATN, so it will be also be in this operating environment in The overhead may be different but I think that this is not really an issue. The loading will not be that different in ATN or IP to make the architecture feasible or not. To accept the comment and the solution Moreover, 3.3 discusses AOA, so in itself states that the ATN assumption is not correct. Unless 3.3 is referring to AOC. In which case, back to my point 1! SOLUTION: Just delete the references to ATN. I do not understand its relevance to this work (though I might change my mind once my points 1 & 2 are resolved, since in fact HI may be relying on an unstated assumption that ATN will provide some of the switching and HMI services I was querying above - if so, they need to state that explicitly for both AOC and ATS). 10 version 1.0

13 VM4AAS - Comments on Deliverable D2 WP2 - Deliverable D2:Identification of Requirements (Joint D2.1 & D2.2) Comments and Suggestions and their Resolution No section Affected text & Comments Resolution "When VDL Mode 2 is available (see Equipment Baselines described in [40]) this study assumes that it is used for AOC communications. When VDL Mode 2 is not available in the Equipment Baseline, this study assumes that AOC communications are provided by VDL Mode 4." Comment: FAA build 1A and Eurocontrol Link envisage use of VDL mode 2 for ATC as well as AOC purposes. There is no doubt that VDL Mode 2 is intended for ATS communications as well as AOC. However, if we assume that both ATS and AOC communications are sent via VDL Mode 2, then any study of VDL Mode 4 for the communications function is moot. The study specifically requires that we look at VDL Mode 4 for comm, surv, and comm+surv. Therefore, when evaluating VDL Mode 4 for comm applications, we will assume that VDL Mode 4 does the ATS and possible AOC as well. This is not a VDL Mode 2 study. Recommended action; Clarify the wording to be "When VDL Mode 2 is available (see Equipment Baselines described in [40]) and VDL Mode 4 for communications data link is under study, this study assumes that Mode 2 is used for AOC communications and Mode 4 is used for ATS communications. When VDL Mode 2 is not available in the Equipment Baseline, this study assumes that AOC communications are provided by VDL Mode 4." "Based on that model, the average AOC message, including ATN overhead, is 264 octets." Comment: Should we be using average or worst case data? Earlier, in 4.2.1, the safety critical nature of the environment was identified. We do require all flights to operate safely, not just the average ones. Current AOC messages can run to several thousands of octets, and in the time frame that this study must consider new functions such as weather graphics need to be considered. It has also been observed that current ACARS performance is weather dependent - when the weather is bad the demand for data-linked weather reports increases greatly, as does the need for delay/diversion related company communications, and this can cause existing ACARS frequencies to saturate. This would not be a good time for a VDL mode 4 based air traffic control communications and surveillance systems to fall over. Should be expanded in traffic loading section, is a wiring section only. Remove the example. Expand in the traffic loading. In add that the CDL covers ATS and AOC. 1. The traffic loading is important for air-to-ground communications because it determines the transmitter activity and, therefore, the interference load. Interference is the single driving aircraft integration issue. 2. The statistical law of large numbers says that the performance is reasonably modeled by using the average. 3. The comment raises weather as an issue. Weather reports are UPLINK not DOWNLINK and do not affect aircraft transmitter loading. 4. ACARS is a 2400 bps link. Mode 2 is 31,500 bps and Mode 4 is 19,200 bps. This will substantially reduce the frequency loading. Several thousand octets of ACARS = many seconds. Several thousand octets of Mode2 or Mode 4 < 1 second. Resolution: Include an uplink traffic loading analysis based on 11 version 1.0

14 VM4AAS - Comments on Deliverable D2 No section Affected text & Comments Resolution MACONDO to establish the susceptibility interval The VHF band is permitted for use for "safety and regularity of flight", rather than just safety. If it were just safety, then AOC wouldn't be allowed either! "At this level, loss or corruption of the data is unacceptable." Comment: Corruption is likely to be more of a concern than loss (probably does not alter the argument) Integrity. One aspect of "Integrity" that is commonly applied, and which differentiates Integrity from RER, is the concept of fault reporting. A system that recognises an error has occurred and informs the user accordingly can have a very high integrity, even if the service is very poor. As long as the user knows something has gone wrong, he can try another medium and get his message through. I didn't see this interpretation of Integrity in the usage in Section There may be another case to consider here. VDL mode 4 protocols rely on selforganization (i.e. all participants playing by the rules) and a common time reference. There may therefore be common mode failure cases that affect all aircraft in the airspace concerned. The resulting widespread loss of communications and surveillance would be very much more serious than loss of functions on an individual aircraft, or the loss of individual functions. (6.3, certification issues, alludes to this). GPS is known to be vulnerable to jamming due to the very low signal strengths at which it operates, and this could be one possible source of a common mode failure. Similarly a rogue transmitter in the area could disrupt services for all (due for example to equipment failure or software bugs) Middle of page 26. I note footnote 5, however the figure of 1000 hours for MRBF seems very low days (as per Table 5) seems more realistic. I look forward to clarification Middle of page 26. The sentence beginning "The AMS(R)S MASPS[24] suggests that signal-in-space availability...". This analysis seems to leave out network congestion. In communications systems, network congestion is a major system failure mode. I'm not sure the treatment proposed here ("short short disruptions in service... should be considered as part of the mean transfer delay, and not as a loss of service") really captures the problem. Text will be modified. These words are taken directly from MACONDO. Study team agrees with the sentiment expressed. In fact, this is expressed in the performance parameters. The quantitative requirements on integrity (i.e. corruption of data) are much more severe than the requirements on continuity (i.e., the loss of data). The commenter has identified a continuity vs. an integrity issue. Integrity is based on the residual error rate (RER), that is, the errors that can neither be detected nor corrected. Therefore these errors are not recognized as errors and are passed to the user. The concept of fault reporting is associated with automatic repeat-request protocols, that is, automatic notification to the transmitter that a received message had errors. VM4 protocols can support ARQ capability. Reference to be made to results of study sponsored by EUROCONTROL to examine the timing associated issues hours may be low for an individual unit, but not for a system as a whole days (24,000 hours) seems very high for an individual unit, but may be acceptable for a redundant system with repair. Will be clarified in the updated report. The network is required to provide a certain mean and 95 th percentile transfer delay, including the effects of network congestion. If the network can no longer provide the QoS, it is no longer suitable for the given application. The AMS(R)S MASPS contains quite a bit of detail regarding congestion. Congestion is NOT a failure mode. It is, in the terminology of Required Communications Performance, a fault-free rare event. 12 version 1.0

15 VM4AAS - Comments on Deliverable D2 No section Affected text & Comments Resolution This aspect of comms. system performance makes for a very interesting comparison between VDLM2 and VDLM4. Both will saturate at some point, leading to congestion, but VDLM4 has certain potential advantages: a) it has a potentially higher saturation demand load, b) it is capable of recognising and reporting saturation leading to higher system integrity Middle of page 27. The sudden appearance of an architecture where VDLM4 is used for both COM and SUR is disturbing. The functional (non equipment-based) approach of Section 3 seems to have gone out the window! It seems a bit rash. This step is probably inappropriate for air transport aircraft, at least not without considerable safety analysis, and is likely to be hotly disputed. It might be cleaner (and less controversial) to avoid this assumption and (even implicitly) to discuss an architecture where VDL4 equipment is used for both, but with independent units, or to assume independent technologies as backup. The all-vdl4 solution can allocate a VDLM4 unit independently for each of COM and SUR, with each acting as backup to the other. The suggestion is good but the VM2 capacity study, and the saturation of VDL Mode 4 vs. VDL Mode 2 are Mode 4 vs. Mode 2 SYSTEM issues, and not part of this airborne architecture study of VM4. VDLM4 for COMM and SURV is a topic for this study as explicitly required by EUROCONTROL. The availability concern expressed in the deliverable will, of course, become an aircraft safety concern. An "All VDL4" solution is not acceptable, as a minimum of redundant independent DSB-AM operation is also required. See the basic assumptions in D1. A COMM and SURV VDL architecture using independent units as described is under consideration Sentence beginning "For the airborne architecture, continuity of service is a function of...". I'd propose adding network congestion as a failure mode. Given the numbers in Table 3, where one aircraft can consume 3.8% of channel capacity, this seems an obvious eventuality Comment #1: VDL Mode 4 ARINC Characteristic:- I believe that It is generally agreed within AEEC that it would be impractical to amend ARINC 750 to support VDL mode 4. The defined form factor is not suitable for the following reasons: - Need for multiple receivers - Need for multiple aerial connections - Need for GPS interface or internal GPS receiver (another aerial) - Need for additional aircraft systems interfaces (e.g. CDTI, navigation and autoflight data) Comment #2: New Standardization Efforts:- Status of VDL mode 4 MOPS is not mentioned Suggestion on further constraints to consider: Looking at VDL mode 4 in isolation, the option of integrating the necessary GPS time source into the VDLM4 transceiver looks attractive. It is worth considering that This study is interested in single-aircraft continuity of service. Loading is used only for interference analysis. Report notes that there are two options: include in 750 or develop new standard. No position is taken. Resolution Include comment#1 in discussion of 750. Point to potentially new standard for x. Add section about existing standard and on-going maintenance work. This is a valid comment for an AT class aircraft, but the GPS-VDLM4 integration might be an attractive alternative for Light and Small aircraft classes. 13 version 1.0

16 VM4AAS - Comments on Deliverable D2 No section Affected text & Comments Resolution GPS also features in the multimode receiver (ILS/MLS/GLS), and is being considered for integration in the enhanced ground proximity warning system unit, as Will be considered in WP3. well as existing as a separate system on some aircraft. Airframe space for all of these aerials is a concern, so rationalising this situation is a significant architectural issue. A GPS source in a VDL transceiver is unlikely to be certifiable for low visibility GPS landing system operations. This would lead us to seek to either re-use the GPS in the Response to General Comment: There is intentionally no detail about the data flow analysis. The current Voice communications are assumed to remain on the aircraft (see equipment baselines of D1). MMR for VDL mode 4 (subject to timing issues being satisfactory), or to seek other solutions such as a VDL mode 4 using a VHF aerial with a GPS aerial integrated into Will be consider in WP3. it, and using a single feeder for both? (There may be problems with this approach too). General comment: I do not feel that the data flow analysis or this section bring out the fact that most modern commercial transports have three VHF communications systems and that they are already fully utilised, with dispatch relief only achieved at the expense of loss of some of the lower priority communications needs. It therefore seems to me that addition of VDL mode 4 to this picture, with the necessary redundancy in the VDL mode 4 function, will require additional systems. This has architectural consequences Comment on timing issues: This section identifies the criticality of timing to VDL mode 4 and the desirability of GPS as a source. In terms of impact on the architecture it may also be worth noting that if a GPS sensor external to the radio is used the interface is critical because the timing accuracy required is well beyond the performance of the conventional ARINC 429 interface. The PS One pulse per second output may be adequate but is the length/capacitance/inductance of interface wiring critical in this case, and how vulnerable is it to EMI? Comment on Infrastructure issues: The likely dependency of VDL mode 4 on GPS/GNSS was identified in 7.1. The infrastructure issues related to GPS, which are a major reason why it is not accepted as a sole means of navigation outside the USA, may be relevant to this study Comment on table 8: No mention of the aerial system in the table. True enough. We'll keep this in mind as the study progresses. This is a system level issue that is outside the scope of the current study. A line feed is missing in the third row of items. It should read Interference Preliminary Antenna Placement. 14 version 1.0

17 VM4AAS- Comments on Deliverable D3.1 WP3 - Deliverable D3.1: Architecture Descriptions COMMENTS, SUGGESTIONS and RESOLUTION of COMMENTS No. where Comment Resolution 1. Section , 5.3, 6.2.3,6.3, 7.2.3, 7.3, 8.2.1, & 9 & elsewhere in Appendic es. Availability & Continuity calculations for Arch #10 & #10A are identical. Consider modifying architecture 10/10A to just include 10A with a single antenna output. In accordance with other comments and a common sense review after the paper was completed, architecture 10 will be deleted and architecture 10A will be retained. Therefore, this inconsistency will be resolved 2. Sections on GA category architectu re SCAA are working on an architecture for GA that is some how included in your study but needs some clarification. Candidate Architecture Diagram 10 includes one architecture with an antenna switch. This architecture reduces the total number of antennas to two. We suggest that this solution has its own Candidate Architecture Diagram. Our architecture, advantages and disadvantages presented below could give input to this Candidate Architecture Diagram. SCCA are planning to equip a number of GA aircrafts with a combined voice and data transceiver. See Figure 1 (in Appendix A of this document). This transceiver replaces one of the two existing voice radios and is only using the existing antennas. This gives a number of advantages and a number of disadvantages. The main advantage is that there is no need for extra antennas. It is a low cost architecture both regarding installations and equipment. It is also a god reason for the owner of a GA to invest in a transceiver that it includes a voice radio. This Architecture also has some disadvantages. The main problem is the priority between Voice and VDL Mode 4. Regarding the solution in Figure 1 SCAA has solved the priority as in Figure 2 (in Appendix A of this document). Agreed. In response to this comment, architecture 10 will be deleted and 10A retained. The SCAA spec will be referenced as representative of 10A. Text similar to the following has been included in multiple places throughout the report: As in other architectures, the new Mode 4-capable radio also has DSB- AM voice capabilities, as required in [9]. Reference [9] has been added, as follows: [9] Requirement Specification for a Combined Voice and CNS Airborne System for General Aviation Aircraft, SCAA_NUP_WP33_GA_XPDR_SPEC_1.0, Stockholm: SCAA, In addition, the following footnote has been added in Section 4.1 Reference [9] is a draft specification for a Mode 4-capable radio targeted at the GA marketplace. The general architecture is very 15 version 1.0

18 VM4AAS- Comments on Deliverable D3.1 No. where Comment Resolution For more detailed information please find attached a document [Ref 1] that describes the specification for the transceiver described above. [Ref 1] Requirement Specification for a Combined Voice and CNS Airborne System for General Aviation Aircraft. Doc Id: SCAA_NUP_WP33_GA_XPDR_SPEC_1.0 similar to that discussed in Architectures 10 and 11. [9] also contains more detailed information regarding priority access to the limited transmitter resources during periods of simultaneous DSB-AM "keydown" and VDL Mode 4 ADS-B operations. 3. Executive summary Existing text: As the VDL Mode 4 data link is expected to operate alongside other data links onboard an aircraft, it has to be shown that installation alongside these systems will not present any major problems. VDL4 is designed to support both communications and surveillance functions. The combination of these two functionalities may result in additional considerations." Suggest changing to support both communications and surveillance to to support both data communications and surveillance. Actually there are many references to communications where data communications is meant. It may be rather pedantic and verbose to change all of them but I think we need to be careful about passages being quoted out of context when the report is published. The Executive Summary has been reworded to more clearly reflect the CDL and SDL functions. The document has been searched for "communications" and clarifications made where appropriate in light of this comment. 4. Figure 1 Note that FMS interface to CMU is far from universally implemented, and that currently it would be highly unusual for the Left VHF to be linked to the CMU (accepting that ARINC 758 envisages this architecture). 5. Figure 1 There is currently no interface between the SSR transponder and the CMU (much as I believe that there should be!) The paragraph describing Figure 1 has been reworded as follows: Figure 1 shows the simplified generic baseline aircraft architecture used for this study. This architecture is based on detailed information provided by Boeing. It is presented as representative of that used in Large aircraft, but is by no means universally correct. Only those interconnections that will be discussed later in this report are included. No attempt has been made to illustrate every connection. For example, audio panels used to route conventional voice inputs and outputs are not shown. A Communications Management Unit is shown, although many current aircraft have a simpler ACARS management unit, instead. This particular communication routing has been changed. In addition, please see the response to Comment VDL Mode 4 for CDL Only The report states: Left and Right GNSS Receivers: Most modern aircraft already have receivers installed Agreement with the comment. The wording has been changed. In addition, specific retrofit discussions are included in new 16 version 1.0

19 VM4AAS- Comments on Deliverable D3.1 No. where Comment Resolution Comment: This rather depends on the definition of modern. It is probably true that most air transport aircraft currently leaving factories have GPS but I question whether most air transport aircraft currently in service have GPS. sections 3.1.2, 3.2.2, and VDL Mode 4 for CDL Only VDL Mode 4 for CDL and SDL From the report: Additional GNSS receiver(s) may be required if GNSS precision landing applications are authorized. Comment: It would be highly undesirable to have separate GNSS sensors for different applications. If an aircraft already equipped with GNSS were upgraded to use GNSS as a landing aid I would expect this to be done by replacing the existing sensors (probably with ARINC 755 MMR), rather than adding further sensors. While GNSS sensors themselves may be relatively inexpensive and light, the multiplicity of aerials would be a concern if a GPS card were added to every system interested in GPS data. This section seems to imply that the same radio could simultaneously operate on VDL 2 and VDL 4. Is this a result of the conclusion that interference is inevitable so could be managed on a time sharing basis, accepting that some loss of performance of both systems would result? Agreement with the comment. Wording has been clarified What was meant here was that a precision landing function will probably require 3 GNSS receivers (like 3 LOC or 3 GS), not that an entire duplicate string would be required.. The indication of simultaneous support for VDL Mode 2 and VDL Mode 4 is based on the initial assumptions for the study as expressed by ECTL. Therefore all of the recommended architectures would be capable of simultaneously operating on VDL Mode 2 and VDL Mode 4. This is not, strictly speaking, a result of the inevitability of interference. The only interference mitigation strategy supported by simultaneous M4/M2 operation is on the transmit side, where WP3.2 indicates that interference isn't a problem anyway. A radio (with multiple receivers) that simultaneously performs M2 and M4 has no more or less control over interference. 9. Table 1 ID1 has the disadvantage that there are a number of common mode failure cases for the VDLM4 capability (is this what is meant by No diversity for VDL Mode 4?). There are common failure modes, and these are accommodated in the availability, continuity, unavailability analyses. But these failure modes are not related to the "loss of diversity". In Architecture #1, the transmit and receive antennas for M4 are on the top and bottom (or, equivalently, the bottom ant top) of the aircraft. Because this architecture is incapable of receiving signals from both sides of the aircraft, it does not implement the diversity performance assumed by the TLAT report. Therefore, 17 version 1.0