Docstoc

appendix-b

Document Sample
appendix-b Powered By Docstoc
					                                    AMCP/5-WP/4
                                    APPENDIX B




           APPENDIX B

THE DEVELOPMENT OF DATA LINKS FOR
    SURVEILLANCE APPLICATIONS

            VERSION 2

           5 February 1998
AMCP/5-WP/4
APPENDIX B

                                                                          B-2

                                                         TABLE OF CONTENTS


1. ASSESSMENT OF SURVEILLANCE ALTERNATIVES ................................................................ B-3
   1.1 Interference resistance..................................................................................................................... B-3
   1.2 System availability .......................................................................................................................... B-5
   1.3 System integrity .............................................................................................................................. B-6
   1.4 Non-interference to other aeronautical systems .............................................................................. B-6
   1.5 Acquisition time.............................................................................................................................. B-7
   1.6 Independent validation of position .................................................................................................. B-8
   1.7 Maintenance of functional partitions between navigation and surveillance and common
           failure modes ............................................................................................................................. B-8
   1.8 Autonomous air-air operations ........................................................................................................ B-9
   1.9 Operational A/C traffic densities .................................................................................................. B-10
   1.10 Operational domain radius ....................................................................................................... B-10
   1.11 Received update rate (air-ground) ........................................................................................... B-10
   1.12 Received update rate (air-air)................................................................................................... B-10
   1.13 Barometric altitude accuracy ................................................................................................... B-11
   1.14 Geometric altitude accuracy..................................................................................................... B-11
   1.15 A/C call sign and category acquisition rate .............................................................................. B-11
   1.16 RF frequency ........................................................................................................................... B-11
   1.17 Antenna requirements .............................................................................................................. B-11
   1.18 Spectrum efficiency ................................................................................................................. B-12
   1.19 Support to related applications ................................................................................................ B-12
   1.20 Ease of transition ..................................................................................................................... B-13
   1.21 Minimal complexity................................................................................................................. B-13
   1.22 Impact of fundamental design issues ....................................................................................... B-14
           1.22.1 Frequency band ............................................................................................................. B-14
           1.22.2 Modulation .................................................................................................................... B-15
           1.22.3 Multiple access technique ............................................................................................. B-15
   1.23 Other factors ............................................................................................................................ B-15
           1.23.1 Maturity of technology .................................................................................................. B-15
           1.23.2 Transition ...................................................................................................................... B-16
           1.23.3 Interoperability .............................................................................................................. B-17
           1.23.4 Flexibility ...................................................................................................................... B-17
           1.23.5 Functionally independent .............................................................................................. B-17
           1.23.6 Non proprietary ............................................................................................................. B-18
           1.23.7 Robustness/fallback states ............................................................................................. B-18
           1.23.8 Future suitability ........................................................................................................... B-18
           1.23.9 Cost ............................................................................................................................... B-18

2. CONCLUSIONS ............................................................................................................................... B-19

3. LINK BUDGETS FOR MODE S, UAT AND VDL MODE 4 ......................................................... B-21

4. INTEGRITY ASSESSMENT FOR VDL MODE 4 .......................................................................... B-23
                                                                                      AMCP/5-WP/4
                                                                                      APPENDIX B

                                                    B-3



1.      ASSESSMENT OF SURVEILLANCE ALTERNATIVES


1.1             Interference resistance


The Mode S system has been designed to resist harmful interference, stemming from the shared use of the
band 960 - 1 215 MHz, for the applications it currently supports. This is primarily secondary surveillance
radar (SSR) surveillance with limited use of auxiliary applications based on Mode S specific services.
The Mode S-based automatic dependent surveillance — broadcast (ADS-B) system shares a channel with
SSR and identification friend or foe (IFF) systems. Simulation and analysis is required to investigate its
ability to support future applications such as ADS-B, advanced surface movement guidance and control
systems (A-SMGCS) and other auxiliary applications in future projected operational environments.
On-going simulations show that Mode S is expected to provide suitable interference resistance for the new
ADS-B and ASMGCS applications if certain assumptions are made regarding:


                a)   1 090 MHz receiver equipment characteristics;


                b)   infrastructure (Mode A/C SSR replacement with Mode S);


                c)   operational traffic densities; and


                d)   low-power transmit mode for surface operations.


However, due to self-interference, currently-fielded equipment is not expected to satisfy air-ground and
air-air ADS-B update rate requirements in the far-term traffic environment projected by ICAO, unless
existing Mode A/C SSR sensors are replaced with Mode S, or the ground system is in some other way
especially designed to address the peak loading conditions projected. An example is the LA Basin, where
upgraded avionics and Mode S ground stations/radars are required to meet the requirements. Sectorized
high-gain ground stations might be substituted for Mode S radars; however, certain update rate
requirements might not be satisfied with current equipment in this environment, and the fall-back mode of
independent SSRs would be lost (unless multiple ground stations were netted together with accurate
system time to allow for multilateration).


In the Mode S architecture, auxiliary applications such as traffic information service (TIS) and flight
information service (FIS) would share the 1 030 MHz channel with SSR, IFF, ACAS interrogations,
extended surveillance applications associated with Mode S, other Mode S specific services and possibly
ATN data link (if supported via this medium). It is not clear how these could all share a single channel
without unacceptable impact to the primary applications associated with SSR.
AMCP/5-WP/4
APPENDIX B

                                                   B-4



The VDL Mode 4 system can be operated in the very high frequency (VHF) navigation band or the VHF
communications band. The current operations concept assumes one GSC high in the communications
band and one GSC in the navigation band; however, candidate frequency assignments for the operational
system are expected to be identified during the validation phase. The concept of dual GSCs provides for
additional capacity as well as resistance to intermittent interference and fading. Local ADS-B channels
(i.e. to support terminal area and surface use, etc.) may be assigned subject to local spectrum planning
constraints and availability and reliability considerations. These local channels may be in the navigation
band or the communications band, subject to local and regional decisions and policy guidelines. There do
not appear to be any significant interference issues associated with operations in the navigation band.1
VDL Mode 4 operations in the communications band may be subject to harmful radio frequency
interference stemming from the keying of nearby transmitters used for air-ground communications
functions on a different channel but within the same band; however, this potential interference mode is
mitigated by the desirable co-channel interference performance (CCI) characteristics of VDL Mode 4 and
the short duration of these events (i.e. the wideband noise burst due to power ramp-up, for a typical
ARINC 716 radio, is on the order of microseconds). Unintentional interference is also unlikely to occur
due to human factors considerations (i.e. a pilot unintentionally tuning to a VDL Mode 4 channel will hear
digital transmissions, which will indicate that the channel is not useable for voice). Unintentional
interference from (and to) correctly-tuned VHF voice and data radios, including but not limited to other
VDL Mode 4 radios, can be avoided with suitable receiver filter characteristics (which can be specified in
the SARPs) and suitable spectrum planning. There has been no harmful interference reported in
association with on-going field trials and operational usage of prototype VDL Mode 4 avionics, however,
further validation is required to determine the minimum channel spacing in a typical and worst case
operational environment.


VDL Mode 4 is specially designed to accommodate and manage self-interference from multiple VDL
Mode 4 stations on a heavily-loaded channel. This is achieved with a combination of good CCI
performance, autonomous self-organizing TDMA (STDMA) protocols, and an option for channel resource
assignments from a ground station in the most dense airspace environments. These features allow VDL
Mode 4 to satisfy the ADS-B update rate performance requirements under the worst-case ICAO loading
scenarios.



1 D8PSK   has been examined for GBAS, and is found to be suitable for uplink applications.
 However, technical issues have been identified which are thought to be related to
 transmitter keying.      These are expected to be resolved, but may introduce a schedule
 delay regarding the use of D8PSK for airborne transmissions. GFSK has lower sidelobes
 than D8PSK in adjacent channels, and has a longer transmitter ramp-up period, and is
 therefore expected to incur lower risk. The validation effort may need to address the
 interplay of channel assignment criteria, antenna isolation and avionics technical
 characteristics with regard to receiver desensitization.
                                                                                         AMCP/5-WP/4
                                                                                         APPENDIX B

                                                     B-5

The universal access transceiver (UAT) design relies on short messages at a high channel data rate with a
modulation technique that has good CCI. As a consequence, it is believed to accommodate the expected
levels of self-interference in an operational environment (validation required with respect to self
interference, as well as interference into, and from, existing users of the band such as TACAN and DME).


The RA DLS system has not been sufficiently well characterized to assess its interference resistance.
Potentially, the shift to civilian technology and open, standardized systems suitable for civil aviation, will
have an impact on system performance. Until suitable spectrum is identified, it is not possible to assess
the impact of possible existing users on the RADLS system.


A desirable feature of future ADS-B systems may be the ability to achieve reliable performance between
aircraft operating on the surface. The VDL Mode 4 system appears to be superior to the other alternatives
in regard to this feature, as a consequence of its improved immunity against multipath and shadowing.


1.2             System availability


System availability considers avionics availability, ground system availability (for air-ground applications),
RF channel availability and availability of other auxiliary systems (e.g. global navigation satellite system
(GNSS)) that may be required to support the navigation/surveillance data link system.


All system alternatives can meet avionics hardware availability requirements that may be associated with
future navigation and surveillance data link systems. The Mode S system has already demonstrated that it
meets the availability requirements associated with the current surveillance system. This applies to the
Mode S transponder. A Mode S-based ADS-B system would also involve a 1 090 MHz receiver (i.e. for a
user that can receive ADS-B messages as well as transmit them).2 Because of the existence of 2 LRUs,


2 This   second LRU could be imagined as an ACAS transceiver; however, the technical
 performance requirements of the 1090 MHz receiver, needed for ADS-B in the far term,
 are somewhat different from the technical performance requirements of ACAS.                             The
 ADS-B 1090 MHz receiver needs a lower noise floor than an ACAS receiver, a reduction
 in minimum threshold level, a reduction in noise bandwidth, and enhanced reply
 processing.     The reduction in MTL impacts Mode S surveillance logic per RTCA/DO-185
 para. 2.2.8.9.1; the reduction in noise bandwidth may make the receiver unsuitable for
 reception of +/- 3 MHz transponders per RTCA/DO-181 para. 2.2.3.1; and the capture
 of targets at long range will reduce the probability of reception of targets at closer range.
 As a consequence of these issues, and considering the fact that current TCAS/ACAS users
 already have a TCAS/ACAS LRU which works fine, it may be more expedient for some or
AMCP/5-WP/4
APPENDIX B

                                                     B-6

the availability of the total airborne ADS-B system will be somewhat lower than the availability of the
Mode S transponder alone. Implementers will need to ensure that the needed availability is maintained,
especially for GA installations that lack redundant equipment.


All system alternatives can satisfy over-all ground system availability requirements. The Mode S SSR has
already demonstrated that it meets the availability requirements associated with the current surveillance
system. However, this is achieved at relatively high cost. All the ADS-B alternatives, including the
Mode S alternative, have the potential to achieve availability and coverage requirements for air-ground
applications at substantially lower cost relative to current SSR (in the case of Mode S, this assumes a
transition from SSR to omnidirectional or sectorized stations). This is due to the simple architecture and
technology of a radio-based communications system as opposed to a radar. For all ADS-B system
alternatives, ground station availability is nominally affected in part by complexity (e.g. a sectorized
ground station with multiple transceivers is more complex than an omni-directional system with a single
transceiver). However, these differences can be overcome with additional redundancy and site diversity
(if appropriate), and do not significantly affect the feasibility assessment at the architectural level.


RF channel availability was addressed in part in the discussion of interference resistance above. Each
system has a differing set of factors which affect RF channel availability; over-all, there is no clear
determination of superiority.


In terms of auxiliary systems needed to support the navigation/surveillance data link system, the
Mode S-based ADS-B system architecture is suitably independent from the various navigation sources that
can be used to develop position reports (although a failure of the navigation source would result in a loss of
ADS-B performance, ACAS functionality could continue with active interrogations, and ground
surveillance could continue if radar coverage was maintained in the airspace). The UAT and DLS
systems could potentially be designed to meet the necessary availability requirements.


The VDL Mode 4 architectural design achieves optimum performance when all stations have accurate
knowledge of their own location and are synchronized with UTC. However, fallback modes exist that
provide for graceful degradation. In low density airspace, achieving optimum system capacity is not
critical (i.e. the system will continue to operate, and be available, in the absence of positioning and time
information). In high density airspace, optimum performance can be guaranteed to an arbitrary level of
availability with suitable ground infrastructure. As a result, there is no inherent impact to VDL Mode 4
system availability due to its architecture. In addition, because of the low power requirement of VDL
Mode 4 avionics, these radios can be placed on the emergency power bus without significantly affecting
battery requirements and are, therefore, immune to onboard power failures.




 all users to install a dedicated (new) 1090 MHz receiver especially designed for ADS-B,
 rather than attempt to build a hybrid LRU that can handle ACAS as well as ADS-B
 functions in a single device.
                                                                                         AMCP/5-WP/4
                                                                                         APPENDIX B

                                                     B-7


The UAT architectural design achieves optimum performance when all stations are synchronized with
UTC. However, fallback modes exist that provide for graceful degradation. In low density airspace,
achieving optimum system capacity is not critical (i.e. the system will continue to operate, and be available,
in the absence of time information). In high density airspace, optimum performance can be guaranteed to
an arbitrary level of availability with suitable ground infrastructure. As a result, there is no inherent
impact to UAT system availability due to its architecture.


Data is not available to analyse the system availability of RADLS.


1.3             System integrity


All the systems have the potential to meet the ADS-B system integrity requirement for probability of
receiving a message with an undetected error = 10-7. Mode S and UAT achieve the necessary level of
performance with a 24-bit CRC, which achieves better than a 10-7 undetected error rate regardless of data
link performance. The VDL Mode 4 system is also expected to meet this requirement, using a
combination of a shorter CRC and independent cross-checks on transmitter ID and reported state vector
changes relative to previous messages (see Appendix B). The RADLS system is not sufficiently
characterized, but is also expected to be able to meet this requirement.


1.4             Non-interference to other aeronautical systems


The Mode S ADS-B system is expected to have a small but measurable impact on SSR performance
(probably less than a few per cent according to data reported in RTCA). A full transition from Mode A/C
SSR to Mode S could mitigate this issue, and should lead to an over-all performance level which is better
than currently achieved (at the cost of new SSR ground systems).


The impact of TIS and FIS activity on the 1 030 MHz channel has not been assessed relative to full
equipage scenarios, and could potentially involve a significant impact to SSR as well as ACAS
performance (e.g. long messages associated with TIS and FIS could interfere with SSR and ACAS activity
on the same channel). Different interference scenarios are likely to be significant depending on ground
infrastructure and media access (e.g. rotating radar versus sectorized or omni-directional base stations,
addressed services such as Mode S TIS versus TIS-B, etc.).


The VDL Mode 4 system has been in field trials and limited operational use for several years, and
prototype equipment does not appear to interfere with other aeronautical systems (validation required to
determine the minimum channel spacing for typical and worst case operating environments).
AMCP/5-WP/4
APPENDIX B

                                                     B-8

As the RA DLS system does not currently define the operational band, an analysis of interference to other
systems is premature.


1.5              Acquisition time


Acquisition time is defined as the time required, from initial power ON or initial access to a channel (if
relevant), until the first full ADS-B report is delivered to a receiving end system. This parameter depends
on the transmit and receive avionics assumed (each candidate system has several alternatives), channel
loading and ground infrastructure. Acquisition time requirements (1-30 seconds) are not well-defined at
this time.


For Mode S, a mobile user starts transmitting almost immediately following power ON, and there is only
one RF channel, but a full ADS-B report requires the reception of three different messages – a position
message, a velocity or rate message, and an ID message. The ID message is transmitted at one-tenth the
rate of the two others (i.e. 0.2 Hz as opposed to 2 Hz). The data content of the ID message is relatively
unchanging, and therefore does not affect nominal steady-state update interval, but acquisition is not
considered to have been achieved until it is received the first time. As a first approximation, acquisition
time in a given scenario may be considered to be ten times longer than the expected update interval
between ADS-B reports. For example, a scenario that involves an update rate of 12 seconds (98 per cent),
such as separation assurance and sequencing (air-to-air), might involve an acquisition time of 120 econds
(2 minutes) considering the requirement to receive the ID burst the first time. This can be reduced by
active interrogation. Similarly, for air-ground surveillance, acquisition time could be extended
unacceptably unless mitigated by active interrogation.


For VDL Mode 4, a mobile user normally collects data on the GSCs for 60 seconds prior to transmitting.
This is normally done following power ON, before the aircraft has moved or entered the movement area of
an aerodrome. Operations on the GSCs involve a 10 second update rate and an ID burst once per minute
(a ratio of 6:1). Since the one-shot delivery probability is typically better than 98 per cent, the acquisition
time would be less than 1 minute (this would typically occur while the aircraft is still at the gate). In an
high traffic density operational environment, a user entering controlled airspace, or initiating operations on
the airport surface, would typically be autotuned to a local channel by the ground automation. This does
not require any additional network entry time, as the autotuning can force the mobile user to transmit in
specifically designated slots. As with the Mode S alternative, active interrogations from mobile users or
ground systems can be used to mitigate the impact of an ID message that is lost on the first attempt, in an
operational scenario where this information is needed immediately. In the case of VDL Mode 4, the
interrogation and response can be performed in reserved slots, further enhancing the confidence with
which the ID information is delivered.


The UAT transmits a full state vector report on every message, and includes all auxiliary on-condition data
every other message. It should easily satisfy acquisition time requirements when they are finalized, in all
cases where update rate requirements are satisfied.
                                                                                      AMCP/5-WP/4
                                                                                      APPENDIX B

                                                   B-9


The RADLS system is not sufficiently characterized to define acquisition time with confidence. However,
it is presumed that it could be designed to support the eventual requirement.


1.6             Independent validation of position


[This section should be deleted, along with the associated entries in the table, as there is no requirement
for the navigation/surveillance data link, or even an ADS-B system, to independently validate the
positioning information being reported. This may be a ground surveillance requirement, but is not an
ADS-B requirement. The ground could verify this information with SSR data (if available),
multi-lateration, one-way or two-way ranging, or even voice confirmation from the pilot. RTCA seems to
have agreed at its latest SC-186 plenary that there was no requirement for the ADS-B system to provide a
built-in mechanism for independent validation of position.


On the other hand, if the Panel determines that it wishes to address such a presumed requirement, subject
to final adoption of SARPs and MASPS, the following text may be used.]


Two of the systems (Mode S and VDL Mode 4) have an ability to verify target range under normal
operational conditions. Each of these systems experiences a different set of measurement errors during
this process. The UAT cannot independently validate position or range. There is not sufficient
information regarding the DLS to determine its potential to independently validate position or range.


1.7             Maintenance of functional partitions between navigation and surveillance and
                common failure modes


All ADS-B systems are nominally dependent on the navigation system of an aircraft to determine position,
velocity, and possibly time (if this is needed and/or reported).


When a Mode S-based ADS-B system is operating in the coverage region of a Mode S SSR, 3D
surveillance benefits derive from accurate navigation on the aircraft and timely reporting via ADS-B. 3D
surveillance can be maintained, with reduced accuracy and degraded tracking performance, even if there is
a navigation system hardware failure on the aircraft or a regional navigation failure. This can be achieved
since the radar positioning and altitude data are independent of GNSS or other area navigation avionics
and systems. However, both the ADS-B and radar data would be lost if there was a Mode S transponder
failure, interruption of power, or inadvertent action by the pilot that caused the transponder to stop
radiating. If a 1 090 MHz receiver failed, then ADS-B data would be lost on the aircraft experiencing the
failure (ACAS data would be lost as well, if the two receiving functions were contained in a single LRU
and no backups were available). Range (but not bearing) can be determined in a backup mode by an
omni-directional Mode S ground station using an active interrogation (assuming that the target aircraft is
AMCP/5-WP/4
APPENDIX B

                                                      B-10

sufficiently close, and the traffic density is sufficiently low, such that acceptable link performance is
maintained). 2D positioning can be maintained with multilateration.


When a VDL Mode 4 airborne station is operating in high density domestic airspace (qualitatively similar
to airspaces with radar coverage today), the service provider can ensure an independent ability to navigate
and derive system time to the accuracy needed to maintain near-optimal data link system performance (as
well as continued navigation on the part of the aircraft, and ADS-B with degraded accuracy). This also
ensures continued surveillance without reliance on radar. One-way ranging and 2D multilateration are
also possible, but with multiple ground stations, secondary navigation and normal ADS-B reporting is
expected to be operationally superior to multilateration. ADS-B data would be lost if all VDL Mode 4
avionics failed. However, unlike the other schemes, VDL Mode 4 avionics can be placed on the
emergency power bus (power requirements are low). This eliminates a failure mode and may be an
important capability should critical applications such as station-keeping be implemented.


All ADS-B systems, regardless of technology, require a detailed safety analysis, and extensive field trials
and experience, prior to adoption as a replacement for current surveillance systems (however, ADS-B may
have immediate benefits in currently procedural airspace, and for other applications and operational
situations that are not directly related to ground surveillance, e.g. trajectory negotiation, in-trail climb, etc.)


1.8              Autonomous air-air operations


All candidate systems can support autonomous operations in all airspaces. The VDL Mode 4 and DLS
systems also offer the option for ground-controlled operations within line-of-sight of a suitable ground
station.


The Mode S system meets performance requirements for the en route domain. The VDL Mode 4 system
meets performance requirements when operating autonomously (or centrally-controlled) in the en route
domain. Further analysis is required to determine the ability of the UAT and DLS system to meet
performance requirements in the en route domain.


When operating autonomously (the only available mode for Mode S and UAT; a selected mode for VDL
Mode 4 and DLS), all four systems suffer a performance degradation in the highest density terminal
airspaces. The Mode S system can satisfy air-ground ADS-B update rate requirements if a sectorized
ground antenna with multiple high-performance receivers are available. The other systems would fail to
meet ICAO ADS-B update rate requirements when operating autonomously in this domain. The VDL
Mode 4 system would also fail to meet ADS-B update rate requirements for the surface domain in a
heavily congested area, if it were to be operated in an autonomous mode. The achievable performance for
UAT and DLS in the surface domain requires further study.
                                                                                         AMCP/5-WP/4
                                                                                         APPENDIX B

                                                    B-11

Autonomous operations for VDL Mode 4 are typically confined to the global signalling channels, and
autonomous reporting rates remain at nominal/default values (nominally once per 10 seconds). If
autonomous operations are employed in high density airspaces, slot sharing would reduce the surveillance
range achieved at a given level of reliability (qualitatively similar to the performance degradations suffered
by Mode S and UAT). VDL Mode 4 ground control is required to increase reporting rates in local
terminal areas and surface domains, and also to optimize slot sharing in extended high-density areas such
as the LA Basin. This is considered by the developers of the system to be acceptable, since ground
stations will exist in these areas to acquire the ADS-B surveillance data and support data communications.
Scheduling software is also required; this is not expected to seriously affect ground system deployment
costs.


1.9             Operational A/C traffic densities


The Mode S system can meet the operational aircraft traffic densities for ADS-B if Mode A/C sensors are
replaced with Mode S sensors, new 1 090 MHz receiving equipment is available (avionics as well as
ground equipment), and sectorized ADS-B receiving stations are developed/fielded for ground
surveillance. Currently used 1 090 MHz receiving equipment cannot satisfy the operational aircraft traffic
densities in an Mode A/C environment. The VDL Mode 4 system can meet operational aircraft traffic
densities for ADS-B in all airspaces, but requires multiple channels and ground control for the terminal
area and surface domains. The UAT is believed to satisfy ADS-B update rate requirements in the
worst-case operational aircraft traffic density environment. The DLS system can potentially meet these
requirements (further study is required).


1.10            Operational domain radius


All systems are expected to be able to meet the operational range requirements for navigation/surveillance
data links. However, the Mode S requires the development and use of new avionics hardware (6 dB better
sensitivity level), which is not in operational use at this time, to achieve an air-to-air range of 100 nmi.
Even with these upgrades, Mode S performance is marginal and technical risk remains in terms of receiver
technology, compatibility with SSR and ACAS, and over-all performance in a worst-case environment.
Appendix A provides a link budget for Mode S, UAT and VDL Mode 4, ignoring these issues of risk. All
systems can rely on directional antennas on the ground to extend range and discriminate against near-by
targets separated in azimuth from a desired user. All systems can also rely on multiple ground stations to
provide diversity reception. The optimum ground architecture depends on the system under consideration.


1.11            Received update rate (air-ground)


All systems can meet the received update rate requirements for kinematic data (position, velocity, etc.) with
sufficient ground infrastructure. In the case of Mode S, this may be a mixture of SSR, high-gain
sectorized ground stations and omni-directional ground stations. In the case of the other alternatives, this
may be a mixture of sectorized and omni-directional ground stations. The Mode S and VDL Mode 4
AMCP/5-WP/4
APPENDIX B

                                                   B-12

systems cannot meet the update rate requirements if these are interpreted to include ID bursts at the same
rate; however, there does not appear to be a technical requirement to repeat this essentially static
information at the same rate.


1.12            Received update rate (air-air)


The Mode S system can meet the received update rate requirements for kinematic data (position, velocity,
etc.) if Mode A/C SSRs are replaced with Mode S, and existing Mode S and TCAS avionics are replaced
or upgraded. Channel bandwidth requirements for the Mode S system are roughly 11 MHz for
currently-fielded equipment (accounting for avionics with +/- 3 MHz carrier frequency stability), and
4.4 MHz for the proposed enhanced 1 090 MHz receiver (assuming all transmit avionics are upgraded to
satisfy a tighter +/- 1 MHz carrier frequency stability requirement). The channel data rate for the Mode S
system is 1 Mbps. The VDL Mode 4 system can meet the received update rate requirements for kinematic
data if sufficient local ADS-B channels are provided (for the LA Basin, this is expected to be two terminal
area channels, one surface channel and one PRM channel for a total of four 25 kHz channels in addition to
the two GSCs proposed world-wide, or an aggregate of 150 kHz). The UAT is expected to also meet
requirements utilizing a 3 MHz channel (99 per cent of signal energy is contained within 1.25 MHz).
Further analysis is required to assess the ability of RADLS to meet requirements.


1.13            Barometric altitude accuracy


This requirement is accommodated by all alternatives. The RADLS can potentially support it (message
sets for all systems may require adjustment following ADS-B MASPS approval).


1.14            Geometric altitude accuracy


This requirement can be accommodated by all alternatives. The Mode S system can meet this tentative
requirement to an altitude of 400 000 feet with its current message set (it reports geometric altitude as a
delta from barometric altitude). The VDL Mode 4 system can meet this requirement to 66 000 feet with
its current message set. The UAT system can meet this requirement up to 100 000 feet with its current
message set. Geometric altitude is not currently reported by the DLS system, but could be added. All
systems could potentially accommodate extremely high altitude and transatmospheric vehicles with special
message sets.


1.15            A/C call sign and category acquisition rate


[delete these requirements, as they are already addressed]
                                                                                      AMCP/5-WP/4
                                                                                      APPENDIX B

                                                   B-13

1.16            RF frequency


The Mode S, UAT and VDL Mode 4 systems either operate or are proposed to operate within suitable
protected aeronautical radionavigation service (ARNS) bands. VDL Mode 4 is proposed to also operate
in the VHF communications band. As noted above, the current operations concept for VDL Mode 4
proposes a GSC at the upper end of the VHF communications band, and another GSC in the VHF
navigation band; however, candidate frequency assignments for the operational system are expected to be
identified during the validation phase.


Operations during the transition period can be on frequencies that are not uniform world-wide, although it
is certainly more ideal to operate with uniform allocations.


Candidate frequencies for the RA DLS system have not been identified.


1.17            Antenna requirements


The Mode S system for ADS-B is expected to require on the order of four antennas for a GA configuration
and six to eight antennas for transport category aircraft (in addition to those antennae required for
TCAS/ACAS). For GA, this involves dedicated top and bottom mounted antennas for a Mode S
transponder and dedicated top and bottom mounted antennas for a 1 090 MHz receiver (which may be an
enhanced ACAS transceiver). A transmit-only ADS-B user could avoid the second pair of antennas as
well as the 1 090 MHz receiver, but would fail to derive ADS-B benefits. Dedicated top and bottom
mounted antennas are required to provide for effectively omnidirectional antenna gain, which is needed to
ensure adequate operational range. For air transport, the additional antennas (and associated avionics) are
due to redundancy and sparing requirements, as well as the possible need to maintain two types of
1 090 MHz receivers:


                a)   standard TCAS/ACAS for collision avoidance; and


                b)   enhanced 1 090 MHz receivers for extended range ADS-B (but unsuited to ACAS as
                     a consequence of performance trade-offs within the receiver).


The remaining three systems (VDL Mode 4, UAT and RADLS) are expected to require two antennas.
The UAT system (similar to Mode S) splits its transmissions between the two antennae to mitigate the
effects of the antenna pattern lobing and airframe shadowing.


In the case of VDL Mode 4, a two-antenna configuration would typically involve one receiving antenna
and one transmitting antenna. These could potentially be shared with other VHF services (e.g. voice,
AMCP/5-WP/4
APPENDIX B

                                                   B-14

GBAS), but switching and fallback alternatives would have to be carefully assessed in consideration of the
performance requirements of the services being supported. A general aviation user could possibly share a
single antenna with an existing voice radio, giving up some marginal reliability performance for ADS-B
but preserving two-way capability for voice as well as ADS-B and data. A transport category user could
possibly require one or two additional antennas, in order to avoid any marginal loss in message delivery
performance, but the optimum configuration will depend on the applications supported, the antenna
isolation available, the avionics available and the cost trade-offs performed by the user.


1.18            Spectrum efficiency


The VDL Mode 4 system offers the greatest measure of spectrum efficiency in terms of bits per second per
Hz of channel bandwidth; values are 0.77 for VDL Mode 4 versus 0.5 for Mode S and 0.3 for UAT and
RADLS.


1.19            Support to related applications


All systems have the potential to support related applications. VDL Mode 4 offers the greatest potential
and flexibility since it is a channelized system (different applications can be separated onto different
channels, and indeed can even be separated into different subbands of the VHF spectrum) and is designed
to support communications, navigation and surveillance applications. Mode S can also support CNS
applications, but flexibility is limited since the highest priority usage of the 1 030/1 090 MHz spectrum is
radar surveillance. This limits the ability to support high data rate (or low priority) applications. The
UAT system can support ASMGCS, TIS-B and FIS-B.


1.20            Ease of transition


Mode S offers a transition path for service providers who have already installed Mode S. However,
delivery of full benefits from ADS-B requires users to shift to new higher-performance Mode S
transponders and completely new 1 090 MHz receivers (which may have to be installed in parallel with
existing ACAS avionics). New Mode S ground stations are still required in areas not currently equipped
with Mode S, and possibly even in some areas that are (i.e. additional stations will be required for ground
coverage, etc.). Existing Mode A/C sensors must also be replaced with Mode S sensors in high density
airspaces, in order to reduce FRUIT rates to levels that will allow the Mode S system to achieve required
update rate performance. This may involve significant costs and operational impacts for service providers
and other users of SSR technology (e.g. military, etc.) in some regions.


The VDL Mode 4 system requires new equipage by both service providers and users. However, this
equipage is relatively low-cost and suitable for multiple CNS applications. Users may achieve operational
benefit from related applications at an early stage with VDL Mode 4, which will create an incentive for
user equipage and thereby speed transition. Furthermore, to the extent that VDL Mode 4 offers the
                                                                                         AMCP/5-WP/4
                                                                                         APPENDIX B

                                                    B-15

potential for a limited backup to GNSS in domestic airspace, it may ease the transition to GNSS, which is
synergistic with ADS-B.


The UAT and RADLS options would require new equipage for both service providers and users. The
UAT and RADLS options could potentially support multiple applications. However, user demand has not
been assessed, and suitable candidate spectrum has not been identified for RADLS. It is not possible to
assess ease of transition until user demand can be projected, suitable spectrum identified (in the case of
RADLS), and the impact of current users of that spectrum has been assessed.


1.21            Minimal complexity


The alternative systems are characterized by different design choices which imply that where one system
might be relatively complex in a given area, it is relatively simple in another. For example, the Mode S
system operates on a single channel pair with random access. This offers simplicity of control and
minimizes the number of operational modes. However, the ground infrastructure requires the replacement
of Mode A/C interrogators with Mode S interrogators to reduce the number of Mode A/C replies and
preserve far-term system performance. Interrogators may have to be internetworked, and non-radar
ground stations are likely to require sectorized antennas and multiple equipment strings to achieve
necessary range. Antenna siting and coverage issues may be significant for the surface domain. ADS-B
users may require complex trackers (relative to other candidate systems) due to the design which splits
position and velocity data into different messages. Finally, there is still a mode switch required for users
on the surface (reducing the message generation rate relative to airborne users). If the operations concept
for Mode S extended squitter involves non-transmission from stationary vehicles on the ground, yet a third
mode is involved.


The VDL Mode 4 system has a relatively complex control architecture and frequency plan, but simplifies
the ground station design and eases antenna siting and coverage considerations on the ground. The most
significant issue is the frequency plan, which requires multiple frequencies in the highest density airspaces,
and requires careful validation to ensure that full situational awareness is maintained at all times by all
users.


The UAT may be the simplest of the systems under consideration as it is a single-channel system with only
a limited degree of channel organisation, and simplified message structure.


The RADLS may be the most complex system; however, its eventual civilian architecture has not been
characterized, and cannot be assessed at this time.


It is not clear which elements of complexity are most critical, nor is it clear what importance should be
associated with this parameter relative to other parameters. Regional differences may exist.
AMCP/5-WP/4
APPENDIX B

                                                   B-16


1.22            Impact of fundamental design issues


1.22.1          Frequency band


The proposed bands of operation yield various advantages and disadvantages to the alternative systems.
Mode S and UAT operate in spectrum near 1 000 MHz that is channelized for wide bandwidth. This
leads to designs that favour short, high burst rate, unscheduled transmissions on a single channel. This
minimizes the complexity of system control, but increases the cost and complexity of the receiving and
transmitting equipment, requires users to demodulate and process more transmissions than are strictly
required to satisfy operational requirements, leads to more complex ground infrastructure architecture
(i.e. more ground stations, sectorized ground stations, more concerns due to shadowing and multipath,
etc.), limits the maximum range achievable with cost-effective hardware, and introduces other complexities
due to the low probability of message receipt on heavily-loaded channels. The Mode S system shares
spectrum with SSR, IFF and ACAS. This limits performance and reduces the ability to support auxiliary
applications on the same frequency(ies) (1 030 MHz/1 090 MHz). The UAT is also expected to share
spectrum with existing applications in some regions, so spectrum planning and re-assignment would be
required to provide a single world-wide allocation. The design of UAT (single channel, 1 Mbps burst
rate, random access) limits the auxiliary applications to ASMGCS, FIS-B and TIS-B. Spectrum resources
for civil aviation use of the RADLS system have not been identified.


The VDL Mode 4 operates in spectrum near 100 MHz that is channelized for 25 kHz, 50 kHz and 100 kHz
systems (VDL Mode 4 uses multiple 25 kHz channels). This increases the complexity of system control,
but use of VHF reduces the complexity of ground infrastructure (fewer, more simple ground stations),
simplifies the design of transmitting and receiving equipment, reduces transmit power requirements, allows
greater transmission range with cost-effective hardware, and allows users to process a subset of messages
which have been organized onto particular channels of interest (e.g. surface channel, terminal area
channels, etc.). The VDL Mode 4 system is proposed to operate in the VHF navigation and
communications bands, on multiple channels allowing the support of auxiliary applications. It appears
that a large number of frequencies are available in the navigation band by virtue of the ability of VDL
Mode 4 systems to operate on 25 kHz channels (most current applications in the navigation band are
narrowband, but have been assigned on 50 kHz or 100 kHz centres). The communications band is
currently congested, but a limited number of frequencies may be available in the near term, and spectrum
congestion may ease in the future as existing 25 kHz voice transitions to 8.33 voice and data, and Mode 3
voice and data. It has also been suggested, by at least one State, that the upper part of the navigation band
be shared with communications.


1.22.2          Modulation


The Mode S modulation has been validated. All other systems rely on a modulation technique that would
require validation.
                                                                                         AMCP/5-WP/4
                                                                                         APPENDIX B

                                                    B-17


1.22.3          Multiple access technique


Both the Mode S and UAT systems utilize single wideband channels and unslotted random access
techniques. Conversely, the VDL Mode 4 system and RADLS utilize two or more narrow band channels
and a TDMA access frame, which can be managed in an autonomous or ground-directed mode. The
multiple access technique is expected to have a significant impact on over-all system performance;
however, it is difficult to isolate the performance impact as each candidate system has many other
parameters and protocols which may be equally significant from varying perspectives of performance, cost,
operability, flexibility, etc. As a result, side-by-side field trials may be the only way to clearly elucidate
comparative system performance.


1.23            Other factors


1.23.1          Maturity of technology


The Mode S data link protocol (i.e. the signal in space standard) represents the most mature element of any
of the proposed systems, as it has been in use for civil aviation for several years and is based on ICAO
SARPs. Mode S SSR is also fully mature, and can be used to receive ADS-B reports at a ground station.
However, airborne equipage for Mode S-based ADS-B requires new or upgraded Mode S transponders and
completely new 1 090 MHz receivers. These new receivers represent an area of technical risk, and an area
of potential incompatibility with SSR and ACAS requirements.


VDL Mode 4 is next in maturity due to its combined civil and military use, the favourable results of
simulations and field trials, the number of manufacturers with on-going production lines, and the
availability of draft SARPs.


RA LS is third as a consequence of its military experience, but also considering as negative factors the
extensive changes that would be required for civil use, the limited number of manufacturers involved, the
lack of draft SARPs, the lack of any simulation or analytic work addressing its ability to meet civil
requirements, and the lack of cost/logistics/operations concepts associated with the technology (for civil
aviation).


UAT is last as a consequence of its lack of operational experience, lack of existing production facilities
(laboratory prototypes only), lack of draft SARPs, the lack of any simulation or analytic work addressing
its ability to meet civil requirements, and lack of cost/logistics/operations concepts associated with the
technology (for civil aviation).


1.23.2          Transition
AMCP/5-WP/4
APPENDIX B

                                                    B-18



The ACAS system developed by the SICASP is presently, and will continue to be, the means of collision
avoidance. Therefore, the development of a separate ADS-B system which provides ACAS is not
necessary. Any potential ACAS system based on ADS-B would need careful safety assessment in view of
current systems, although the use of ADS-B derived information by future collision avoidance systems is
expected to be beneficial. However, it should be noted that an ADS-B system in itself, by providing
increased situational awareness, will decrease the frequency of encounters where a resolution advisory is
issued.


The most critical element of the transition strategy is developing recognisable benefits for the user
community. If benefits are clearly available, users will equip and the transition to end-state can occur in a
timely fashion. If benefits are not clearly available, equipage will be slow or non-existent and full
transition may never occur. A plan to generate benefits in the near to medium term is likely to be more
effective, in terms of transition, than the immediate availability of SARPs. Another factor is the current or
planned availability of systems to support field trials and testing. Mode S can rely on existing Mode S
SSRs in regions where they exist or are planned. VDL Mode 4 can similarly rely on existing and planned
ground networks. Over-all, it appears that the most critical element for timely implementation is a clear
delineation of user benefits.


VDL Mode 4 is expected to offer over-all lower equipage costs for the user community. Some users
already equipped with Mode S may be able to re-use their equipment under the Mode S alternative;
however, a large fraction of current Mode S users (possibly a majority) may require avionics changes even
for this alternative. All Mode S users who wish to receive ADS-B data (as opposed to merely transmitting
it) will require a new 1 090 MHz receiver that is not yet demonstrated or available.


It may also be possible to pass non-Mode S-derived ADS-B information to an ACAS system via an
existing or new baseband interface (e.g. through the FMS). This requires further study. The relative
costs of equipage will drive user acceptability (in combination with the perceived benefits). The
near-term costs for selected users, as a function of current equipage, are also relevant; however, they are
less significant in terms of long-term decision-making (i.e. decision-making should be based on long-term
costs and benefits to the entire community rather than short-term cost for a selected subset of users).


Regardless of the navigation/surveillance data link(s) selected — and multiple data links may in fact be
implemented over time — primary and secondary radar surveillance is expected to be retained for an
extended period until the availability, reliability and integrity of ADS-B has been demonstrated.
Decommissioning may tend to proceed with elimination of double-coverage (i.e. moving to single
coverage) to conserve operations and maintenance costs. Data links will be implemented independently as
benefits are perceived. There may be a mixture of data links in some regions, catering to different user
classes. It is not anticipated that retention of primary and secondary surveillance systems will have any
significant impact on the timely introduction of data links.
                                                                                       AMCP/5-WP/4
                                                                                       APPENDIX B

                                                   B-19

1.23.3          Interoperability


All ADS-B systems can pass information to ACAS logic with suitable baseband interfaces.
Interoperability can be achieved regardless of the data link alternative; however, a 1 090 MHz-based
solution eases the interface issue by passing RF signals directly to the ACAS 1 090 MHz receiver. A
1 090 MHz solution is slightly preferred per this factor.


1.23.4          Flexibility


The Mode S alternative offers some flexibility to introduce ADS-B in conjunction with existing SSR
surveillance systems. However, this requires as a minimum a software change and may also require a
hardware change with some or all user avionics systems. As this may be a cost and logistics issue,
flexibility may be impaired. With dedicated omni-directional or sectorized L-band receivers, this issue is
avoided but siting and coverage constraints must be carefully addressed (e.g. propagation characteristics of
RF signals at L-band are less desirable than at VHF). An independent ADS-B system lacks the advantage
of sharing hardware elements with an SSR, but can share facilities with any existing aeronautical
installation (e.g. VORs, ground radio facilities, etc.). A VHF-based system eases siting considerations due
to more preferred propagation characteristics, especially near the ground and on the airport surface.
Over-all, in areas where SSR is already deployed, the various competing systems have differing
characteristics which on balance indicate no system is preferred in general.


In regions where no SSR surveillance exists or is planned, flexibility is maximized by a system which
offers minimum installation cost and minimum constraints in terms of siting. In these areas, flexibility
favours a VHF-based solution.


Furthermore, a system based on VDL Mode 4 allows the introduction of advanced services even in regions
where the 1 030 - 1 090 MHz band is congested, and without concern for interference impact on SSR
performance.


1.23.5          Functionally independent


Aviation systems are typically developed to support one or another of the three functional areas of
communications, navigation and surveillance. From a safety standpoint, it is important that aviation
systems avoid a failure mode whereby loss of one of these three elements (i.e. communications, navigation
or surveillance) leads directly to the loss of another. All four systems satisfy this independence. The
Mode S system can preserve surveillance with SSR and multilateration if navigation is lost. However,
air-to-air surveillance based on ADS-B alone would be lost if navigation were compromized. The VDL
Mode 4 system typically relies on GNSS to optimize the use of the channel, but can operate in the
complete absence of positioning and time information with a loss of efficiency. This degradation is not
significant in remote airspace. In domestic airspace, sync bursts from ground-based VDL Mode 4 stations
AMCP/5-WP/4
APPENDIX B

                                                      B-20

(e.g. intra-system signalling) can be used to derive position and time such that no loss of efficiency is
incurred and air-to-air surveillance can be maintained. So this system too is functionally independent of
primary navigation system(s). The UAT can maintain surveillance from the ground via multilateration,
although air-air surveillance would be lost if navigation were compromized. No information is known
about the RADLS system; however, it is expected that it, too, can perform multilateration on the ground.


1.23.6           Non proprietary


All alternatives are equal with respect to this factor.


1.23.7           Robustness/fallback states


The Mode S system offers graceful degradation with increasing traffic load. However, it combines
communications and surveillance (and possibly navigation) applications in a single data link system with
possibly multiple LRUs. All Mode S applications with the exception of ADS-B reception and ACAS can
be provided with a single LRU — the Mode S transponder (this ignores sparing and redundancy
considerations). ADS-B reception and ACAS also require a second LRU (the 1 090 MHz receiver), and
ACAS also requires a 1 030 MHz transmitter (this could be packaged in the same LRU as the 1 090 MHz
receiver; however, there is a potential technical incompatibility between the characteristics of the
1 090 MHz receiver needed for ADS-B, and the characteristics needed for ACAS). Since a user seeking
full ADS-B capability requires at least two LRUs, there is a potential reliability issue which may impair
acceptability (or require even higher equipage cost due to the need for redundant equipment).


The VDL Mode 4 system, UAT and RADLS offer independence of ADS-B from SSR surveillance. A
full ADS-B system based on VDL Mode 4, UAT or RADLS would require only a single LRU.


The VDL Mode 4 system has been designed for graceful degradation in response to all currently identified
failure modes. There is also the possibility that aircraft operating with VDL Mode 4 could use this
equipment, in airspaces where sufficient synchronized transmissions are available, to backup GNSS in the
event of an isolated hardware failure on the aircraft, or a regional failure of GNSS itself. This requires
validation; but if proven, would minimize sparing requirements for the avionics and support the
introduction of GNSS world-wide.


Fallback modes for UAT have not been reviewed. RADLS is not sufficiently well characterized to
review.


1.23.8           Future suitability
                                                                                         AMCP/5-WP/4
                                                                                         APPENDIX B

                                                    B-21

Only the VDL Mode 4 alternative offers the potential to flexibly expand capacity to meet new/future
requirements by the addition of local narrow-band channels. The VDL Mode 4 is preferred per this
factor.


1.23.9          Cost


Although relative cost figures have not been considered within this analysis, general relative costs can be
compared for various user populations and ground infrastructure assumptions (costs are sensitive to these
assumptions). In regions with extensive Mode S infrastructure and user populations with Mode S and
ACAS avionics capable of upgrade via simple software changes, ADS-B via Mode S may be relatively
attractive and existing ACAS users could potentially receive the new ADS-B transmissions with relatively
minor cost impact. However, in at least some regions, inter alia the core of Europe, current Mode S and
ACAS users would require replacement of their existing avionics in order to achieve desired performance
in the projected environment. In such regions, even existing ACAS users might perceive lower costs with
a solution offering avionics at a lower unit cost. The potential need to add redundancy may be an
additional factor that affects costs for Mode S.
In regions where the user population cannot easily upgrade its avionics to handle Mode S extended squitter
(both transmit and receive), sunk costs in existing equipage may be discounted and cost comparisons may
be based on comparative hardware and installation costs. Currently-available cost data for the alternative
technologies indicates a cost advantage (lower cost) to VDL Mode 4.


In regions where there is not extensive Mode S infrastructure, ground infrastructure can be compared on
the basis of the ADS-B base stations associated with the alternative schemes, and costs may be considered
roughly comparable with perhaps a slight advantage accruing to the VHF alternative (VDL Mode 4). This
is due to the anticipated cost advantage of VHF equipment on a unit basis, as well as the fewer number of
base stations required to provide coverage above a given flight level (VHF versus L-band). This may be
particularly significant in large States or regions with limited Mode S infrastructure. Furthermore,
low-cost Mode S ground stations, suitable for ADS-B, are expected to be unsuited to other applications.
VDL Mode 4 ground stations potentially allow amortisation over multiple applications, potentially
reducing unit costs per application. Ground infrastructure for UAT may be relatively low-cost on a unit
basis (similar in cost to a VDL Mode 4 ground station); however, UAT ground stations would be required
in larger numbers due to line of sight and fading limitations on the surface and at low altitude. Ground
infrastructure for RADLS system cannot be assessed until appropriate spectrum resources are identified.


If sectorized base stations are considered, it seems likely that L-band systems relying on random access
channels will have a greater need for such sectorized base station systems than VDL Mode 4, which relies
on STDMA or ground station resource assignments to avoid message overlap in virtually all cases. To
first order, the installed cost of a sectorized base station with six sectors may be considered to be roughly
twice the installed cost of an omnidirectional station. This may provide an additional cost advantage to
VDL Mode 4.
AMCP/5-WP/4
APPENDIX B

                                                    B-22

Relative cost impacts of the various alternative technologies will depend significantly on the assumptions
made with regard to existing infrastructure, existing user equipage and planned migration path for all air
traffic services. Regional differences are likely to lead to different answers in different regions. There is
no clear preference, among the alternative technologies, on a global basis.



2.      CONCLUSIONS


As a result of this analysis, both Mode S extended squitter and VDL Mode 4 appear to meet, or to be
capable of meeting, all the specified requirements and desirable features for ADS-B, TIS and A-SMGCS.
There are differences between these two alternatives, but over-all neither one clearly dominates the other.
The UAT also shows promise, but there is less development work and analysis in regard to this system so
uncertainty is higher. The UAT and DLS lack maturity with regard to civil aviation applications, and in
the case of DLS, appropriate spectrum has not been defined.


Surveillance applications based on ADS-B will first be introduced in areas with low density traffic and in
areas which are lacking surveillance means. TIS will serve as an important application in areas with
ADS-B when not all airspace users are equipped with ADS-B capabilities and in areas where radars are
still in use. This will give airspace users and ATS service providers in most parts of the world early
benefits of the CSN/ATM concept.


All airspace users, ranging from air carriers to balloons, have to be able to install and use the ADS-B
technology. Several levels of implementation have to be provided but the core functionality must be
maintained. The selected ADS-B technology should also be flexible so that it can be adapted to the
specific requirements of various regions in the world.


Secondary radar is used today as the primary mean for surveillance. It is foreseen that the present systems
will be enhanced with MSSR and Mode S and that they will continue to be in use for a long time in areas
with very dense traffic. The radar function should not be jeopardized by added functions in its frequency
band. Ultimately, however, in at least some airspaces, redundant radars might be decommissioned if high
confidence can be achieved that ADS-B can serve as a primary surveillance means.


ACAS is an emergency collision avoidance system which should be kept independent from all other
systems. It should not be dependent on ADS-B or compete with ADS-B for resources (although it can be
enhanced by ADS-B data if certification issues can be resolved, its primary function should be separated
from the ADS-B system).


One of the unique features of ADS-B is the possibility to provide surveillance functions at very low
altitudes and on the ground. This is very useful in A-SMGCS applications and also for search and rescue
(SAR) operations which can benefit from surveillance capabilities close to the surface of ground and water.
                                                                                       AMCP/5-WP/4
                                                                                       APPENDIX B

                                                   B-23

 When considering these applications due account should be taken to minimize problems caused by
reflections and multipath.


Requirements for only a few applications are identified today. As operational experience is gained with
the new technology(ies), these requirements are likely to be changed. Also new applications with specific
requirements will emerge. Therefore it is essential that a very flexible data link is selected.


Mode S and VDL/4 each provide a core set of communication capabilities which can support a range of
CNS applications.


Mode S relies on the same avionics used for SSR surveillance and ACAS. This offers some cost
avoidance for certain users but limits flexibility and leads to competition for channel resources. The
sharing of avionics eases the integration of ADS-B data into ACAS, but introduces a dependency and
common failure mode. The use of the 1 090 MHz frequency may limit performance on the ground. It
appears that for most users, ADS-B supported via Mode S will involve new or replacement equipment and
high costs. Ground infrastructure may also require significant and costly upgrade. Technical risk
remains with regard to the performance of upgraded 1 090 MHz receivers, over-all system performance in
a dense airspace environment, and the operations concept and airborne architecture (i.e. which applications
are supported, and how are they supported on the aircraft? What is the performance of each application
individually, given the aggregate of all applications sharing the channel resource?).


VDL/4 relies on new VHF avionics which are not currently fielded by any broad class of users, but which
is low cost and potentially attractive to general aviation. There is no dependency or common failure mode
with SSR surveillance and ACAS, but users already equipped with ACAS would require a change to their
hardware (or possibly software) to provide an input path for ADS-B derived data. VDL/4 is flexible for a
range of CNS/ATM applications, and the VHF frequency has demonstrated good performance on the
ground. The applications addressed can cover all phases of flight from ―gate-to-gate‖ and can be available
for a variety of users. The high flexibility offers implementation options which can meet different
regional requirements. In a long perspective this will minimize the proliferation of data links. However,
it requires multiple frequencies to support ADS-B applications (and other applications) in a dense airspace
environment. This raises issues of control complexity, failure modes and spectrum availability. These
issues are believed to be solved or solvable; however, validation is required. In view of the potential
shortcomings and risks still associated with Mode S, and the attractive cost and flexibility features of VDL
Mode 4 (notwithstanding the issues which remain for validation), it is desirable to continue development
of VDL Mode 4 as a possible data link that could support navigation and surveillance applications (as well
as communications).


As the Mode S technology is adequately addressed by SICASP, this analysis has led to the following
recommendations for ICAO:
AMCP/5-WP/4
APPENDIX B

                                                     B-24

    the AMCP should continue to develop SARPs for a VDL Mode 4 as a response to the need for data
     link(s) to support navigation and surveillance applications, identified in Recommendation 6/3 by the
     SP/COM/OPS/95; and


    the AMCP should perform the work of validating VDL Mode 4 as a matter of urgency.


    a decision regarding the near-term and long-term choice of a world-wide standard for navigation and
     surveillance data links, supporting general applications, should be deferred until further data regarding
     the benefits of VDL Mode 4 can be generated as an output of the validation effort.



3.       LINK BUDGETS FOR MODE S, UAT AND VDL MODE 4


The link budget presented in Table 1 demonstrates the reliability of the transmission media for all of the
systems except the RADLS system, for which no information was provided. It should be noted that this
computation is a very conservative computation as it assumes the likely worst case antenna gains in the
transmit and receive aircraft, that both the transmitting and receiving stations require the full cable loss,
etc. In practice, one should expect to receive reliably and repeatedly from some percentage of aircraft
even beyond the range for which zero margin has been computed because of better antenna alignments and
cable installations.

In particular, the transmit antenna gains for the aircraft in the Mode S-based ADS-B system and for the
UAT system rely on transmissions from a top and bottom mounted antenna and the receive antenna gain
for the aircraft for these systems rely on receiving from both antennae and selecting the louder source
(simple space diversity). The transmit antenna gain is worse than the receive antenna gain because, in a
high density environment, the probability is that only one or two transmissions will be received without
interference. The aircraft antenna gain numbers are based on the 95th percentile. That is,
                                                                          AMCP/5-WP/4
                                                                          APPENDIX B

                                         B-25

    Table 1. Link Budget

                                       Air-to-air                    Air-to-ground
                             Mode S     VDL/4        UAT       Mode S    VDL/4       UAT
Transmit Power (Watts)          500          10          50       500        10          50
Transmit Power (dBm)           56.99         40       46.99      56.99       40       46.99
Tx Cable Loss (db)                -2            -3        -2        -2        -3          -2
Tx Antenna Gain (dBi)             -2            -4        -2        -2        -4          -2
Free space Path (nm)            100         100         100       200       200         200
Tx Frequency (MHz)             1090         118         960      1090       118         960
Path Loss (db)                -138.5      -119.3      -137.5    -144.5    -125.3      -143.5
Rx Antenna Gain (dBi)             -1            -4        -1       14         6            8
Rx Cable Loss (dB)                -2            -3        -2        -2        -3          -2
Rx Power (dBm)                 -88.5       -93.3       -97.5     -79.5     -89.3       -94.5


Kt0 (dBm/Hz)                   -174        -174        -174      -174      -174        -174
External Noise                    1          20            3        1        20            3
Rx Noise Figure (dB)              1          11            1        1         8            1
Eff System Noise (dB)           3.12      17.97         4.01      3.12    17.51         4.01
Receiver BW (kHz)              4400          16        1100      4400        16        1100
Receiver BW (dBHz)             66.43      42.04       60.41      66.43    42.04       60.41
Total External Noise (dBm)   -104.44    -113.99      -109.58   -104.44   -114.44     -109.58


SNR (dB)                       15.93        20.7      12.07      24.93    25.13       15.07
Lossless Es/N0 (dB)            15.93        20.7      12.07      24.93    25.13       15.07
Tx tolerance (dB)                 -1            -1        -1        -1        -1          -1
Rx tolerance (dB)                 -1            -1        -1        -1        -1          -1
Received Es/N0 (dB)            13.93        18.7      10.07      22.93    23.13       13.07
Bits/Symbol                       1             1          1        1         1            1
Received Eb/N0 (dB)            13.93        18.7      10.07      22.93    23.13       13.07
Required Eb/N0 (dB)              11          14            8       11        14            8


Fading Margin (dB)                0             0          0        0         0            0
System Margin (dB)              2.93         4.7        2.07     11.93      9.13        5.07
AMCP/5-WP/4
APPENDIX B

                                                     B-26

regardless of aircraft manoeuvring, nearly all potential receivers will receive at least this large (in fact,
many will receive a signal 3 or 4 dB greater). Also, the majority of potential receivers in the worst part of
the antenna pattern will either be directly above or below the aircraft (and thus the free space path loss will
be significantly reduced). The ground antenna is based on what is reasonable for an extended range
station (i.e. at VHF, the typical ground station will use an antenna with a gain of 2.15 dBi; however, an
extended range station will use a better antenna).


The receiver bandwidth is not the equivalent noise bandwidth. Thus, the computation of the noise will be
stronger than the actual noise based on the roll-off of the receive filter. Thus, all of the schemes have an
additional dB or so of margin (although the actual amount depends on the specific implementation for the
specific scheme). The noise figure for the Mode S scheme, in particular, requires validation of a new
receiver design with a 6 dB reduction in the noise figure relative to existing designs.


The transmitter and receiver tolerances specified are just an initial estimate of what can be expected, these
numbers, in particular, are subject to validation once hardware is available.


The required Eb/N0 is what is required to meet a BER of 10-5 after any coding is applied. Since, in a
dense environment, all of the schemes can successfully deliver (with 98 per cent probability) only a few
successful transmissions within the required update time, the loss due to bit errors must be small so that
this effect does not substantially reduce the delivery rate (once media access performance has been
considered).


Fading margin was not specified as we did not have the necessary data to include any reasonable numbers.


In the air-to-air situation, the VDL Mode 4 is preferred (primarily because of the reduced path loss at
VHF). In the air-to-ground situation, the Mode S is preferred (primarily because of the large receive
antenna gain). The VDL Mode 4 and UAT systems both have are capable of improving the link budget,
by increasing the power or the receive antenna gain. All three systems can improve the performance by
reducing the cable loss.



4.      INTEGRITY ASSESSMENT FOR VDL MODE 4


The probability of receiving a message with an undetected error may be decomposed as follows:


            Pr{undetected error} = Pr{receiving message with error}  Pr{error is undetected}
                                                                                         AMCP/5-WP/4
                                                                                         APPENDIX B

                                                    B-27

The first factor on the right hand side is a function of the RF channel performance, while the second factor
is a function of the performance of the error detection capability of the protocol. For the simple analysis
presented here, the impact of FEC coding is ignored (note: the body of the report assumes that VDL
Mode 4 uses GFSK, one of two physical layer waveforms included in the draft SARPs. As currently
proposed, the GFSK waveform does not include FEC coding. The alternative D8PSK waveform includes
FEC coding, but this waveform is not considered within the assessment of the report).


The CRC check proposed for VDL Mode 4 is a 16-bit CRC which will detect all error patterns of weight 
3, all error patterns of length less than 16, and all error patterns with an odd number of errors. For error
patterns equal in length or longer than 16 bits, with an even weight, this CRC fails to detect the presence of
an error with a probability of 2-16  1.5 x 10-5. In addition to this link layer integrity process, VDL Mode 4
will only mark sync/ADS-B data as ―validated‖ if it satisfies a kinematic cross-check with the prior
received report from the same station (i.e. new data is always reported, but it is marked ―unvalidated‖ if the
kinematic cross-check fails. This allows the external application to use the data, not use the data, or ask
for a one-shot retransmission as deemed appropriate by that application). The information content of the
sync/ADS-B burst comprises 21 octets = 168 bits, including the CRC. The goal is to demonstrate that no
more than 1 message in 10 million (107) is received and accepted with an undetected error.


The RF channel may be characterized as residing in one of two alternative states: 1) a nominal AWGN
state characterized by a low bit error rate (BER); and 2) a burst error state characterized by a short-term
increase in BER to relatively high levels — potentially as high as BER = 0.5 (pure random bits). Figure
X2-1 illustrates this model. The channel is usually in the AWGN state and occasionally in the BURST
state. For basic channel throughput performance to be maintained, the probability of being in the BURST
state must be relatively low (less than about 1 per cent; however, we will assume a value of 10 per cent for
this preliminary analysis). When in the AWGN state, the BER is typically very low (lower than 10-5).
For the purpose of analysis, it will be conservatively assumed that the BER = 10-4 for this state under all
conditions. The BURST state is usually a result of overlapping transmissions (although the underlying
cause is not significant from the standpoint of determining integrity). We are primarily concerned with
burst error events that exceed 16 symbol periods in length, as bursts of shorter duration will be detected
with 100 per cent reliability by the CRC. The BER under burst conditions can vary over a wide range,
from a low value of 10-4 to a high value of 0.5. Bursts due to overlapping transmissions in adjacent slots
(i.e. due to long-range messages or improper timing) will tend to be located at the beginning or end of the
victim message, and will tend to be shorter than 16 bits. Therefore, the most significant event for analysis
is intentional or unintentional sharing of a slot. This will induce a burst condition that extends essentially
over the entire message.


The probability of being in the AWGN state must be relatively high in order for the system to operate
AMCP/5-WP/4
APPENDIX B

                                                    B-28

(i.e. if the system were in the BURST state too frequently, little or no information would be delivered).
For this analysis, a conservative assumption may be that the system is in the AWGN state approximately
90 per cent of the time, and the BURST state approximately 10 per cent of the time (note: simulations and
field trials indicate that stations avoid slot sharing on more than 98 per cent of the slots, when the offered
load in terms of ADS-B/sync bursts is less than the channel capacity in terms of available slots). It should
be emphasized that the probability of residing in a state, as well as the transition probabilities and BER
distributions for each state, could be subject to refinement during the validation process. The AWGN and
BURST states must be considered separately (burst errors will be found to be more significant).


With respect to random errors, channel errors may be characterized by a bit error rate (BER) describing the
typical probability that a single bit is received in error. For reasonable data link performance, this BER is
normally desired to be on the order of 10-4 or 10-5, or lower, to ensure reasonable probability of message
receipt (at a BER of 10-4, the message loss rate due to random errors alone is almost 2 per cent for a
sync/ADS-B message. At a BER of 10-3, the message loss rate is on the order of 20 per cent and over-all
data link performance is badly corrupted or completely destroyed). From a practical standpoint,
BER > 10-4 may be considered a ―burst‖ event (albeit a weak one), and can be considered in conjunction
with other burst events below. For random errors excluding such burst events, the likelihood that a
message will contain an undetected error, based purely on the performance of the CRC, is on the order of
10-12, and no further analysis is required. This can be seen by calculating the probability of an error
pattern with weight = 4 (the most likely event that can cause a failure of the CRC), multiplying it by the
probability of a CRC failure, and ignoring the limitation that the error pattern have length  16:


Pr{integrity failure | AWGN}      = Pr{4 bit errors in span of 168}  Pr{CRC failure}
                                  = 168C5 (0.0001)4(0.9999)164  2-16(eq. 1)
                                   10-12


The over-all impact on integrity can then be determined by multiplying the conditional probability
determined in (eq. 1) by the probability of being in the AWGN state (which typically would be > 0.9).
The AWGN channel is clearly not an issue.


For burst events, the analysis is more complex since the probability of being in a BURST state is unknown,
the underlying BER during the burst event is unknown and the distribution of errors across the message is
unknown. However, in principle, the conditional probability of an undetected error can be decomposed as
follows:


Pr{integrity failure | BURST}


        = Pr{long burst} Pr{k bit errors | long burst}  Pr{CRC failure | long burst with k errors}
                 Pr{kinematic crosscheck failure | long burst with k errors, and CRC failure}       (eq. 2)
                                                                                         AMCP/5-WP/4
                                                                                         APPENDIX B

                                                      B-29

        < (0.1) Pr{k bit errors | long burst} (2-16) (0.2)                                             (eq. 3)



where we have arbitrarily assumed (for this preliminary analysis) that most bursts are short (typically at the
beginning or end of a message). The probability of k bit errors given a long burst is difficult to evaluate as
it depends on the precise channel conditions, but is clearly less than 1 for any value of k, and the sum of
these probabilities is also less than 1, across all values of k that could cause a CRC failure (k=4,6,8,...).
The factor of 2-16 is the strength of the CRC check. The factor of 0.2 is the kinematic crosscheck, based
on k=4. This is an overbound for k=4,6,8,...


Removing the conditioning on Pr{BURST}, we have a level of integrity conservatively evaluated as
1-3x10-8, satisfying the requirement. It is emphasized that the analytic assumptions require validation, but
were chosen conservatively. The probability of an undetected error is typically orders of magnitude below
10-7. For example, with a strong burst where the ID and position fields are essentially randomized, the
probability that the kinematic cross-check will fail to detect the presence of an error is less than
approximately 10-18. A single bit error in the ID field contributes a minimum additional 4 orders of
magnitude to integrity (the actual amount is greater since it is unlikely that the ID so formed matches an
aircraft at the indicated position). Conversely, for a weak burst where the underlying BER is on the order
of 10-3, there are not sufficient errors in the message (typically) to create a problem for the CRC.


Summary: VDL Mode 4 appears to satisfy the ADS-B integrity requirement for realistic operational
channels. This should be verified during validation.


In the event that additional integrity is required, it could be achieved in various ways that would not
contribute to increased message length. One alternative is to add a second CRC (or an extension of the
first CRC) overlaid with the transmitter ID. Another alternative is to include additional information in the
kinematic cross-check, such as velocity, vertical rate, etc. Each has advantages and disadvantages;
however, subject to validation, it appears at this time that the currently-achieved integrity is acceptable.




                                                   — END —

				
DOCUMENT INFO
Shared By:
Tags: appen, dix-b
Stats:
views:74
posted:12/7/2009
language:English
pages:29
Description: appendix-b