The Mode S Beacon Radar System

航空

游客，如果您要查看本帖隐藏内容请回复

航空

V.A. Orlando
The Mode S Beacon Radar System

Air traffic controllers rely on primary and secondary radars to locate and identifY aircraft. Secondary, or beacon, radars require aircraft to carry devices called transpon-ders that enhance surveillance echoes and provide data links. Airports currently use a secondary-radar system known as the Air Traffic Control Radar Beacon System (ATCRBS). However, ATCRBS has limitations in dense-traffic conditions, and the system's air-to-ground data link is limited. In response to these shortcomings, Lincoln Laboratoryhas developed the Mode SelectBeaconSystem(referred to as Mode S), a next-generation system that extensive laboratory and field testing has validated. In addition to significantsurveillance improvements, Mode S provides the general-purpose ground-air-ground data link necessary to support the future automation of air traffic control (ATC). The FederalAviationAdministration (FAA) is currentlyinstalling the system with initial operation scheduled for 1991.
Airports around the world currently use a type of secondary radar known in the United States as the Air Traffic Control Radar Beacon System (ATCRBS) [11. Because this radar sys-tem was developed more than 30 years ago, itis beginning to strain under today's increased levels ofair traffic.To replace ATCRBS, Lincoln Laboratory has developed the Mode Select Bea-con System (referred to as Mode S) [21, which is scheduledforinitial operation atU.S. airports in 1991.
Mode S (a combined beacon radar and ground-air-ground data-link system) is de-signed for the dense traffic environments ex-pectedin the future. To supP9rtthe automation
('... 1 J
ofairtrafficcontrol (ATC),ModeS boaststwoat-tributes: accurate surveillance even in dense traffic conditions. and reliable ground-air-ground communications capability.
Because Mode S is capable ofcommon-chan-nelinteroperationwith ATCRBS, the fonner can be installed over an extended transition period dUring which Mode-S systems will eventually replace ATCRBS ones. In fact, a major design reqUirementindevelopingModeS was toensure that the system could be implemented in an evolutionary manner. By the time Mode-S de-ployment begins in 1991. about 200.000 air-craftand 500ground-basedradarswill beusing ATCRBS. Mode S is designed to operate in this ATCRBS environment in a way that would permit a gradual transition to all-Mode-S operation.
Providing a high degree of compatibility be-tween Mode S and ATCRBS has achieved the capabilityforsuchatransition. ModeSusesthe same interrogation and reply frequencies as ATCRBS, and the signal formats have been cho-sen to pennit substantial commonality in hard-ware. Such compatibility will permit a smooth, economictransitioninwhich Mode-S radarswill provide surveillance of ATCRBS-equipped air-craft and Mode-S transponders will reply to ATCRBS radars.
This article begins with a description of ATCRBS. followed by a discourse on the system's limitations in regions of high traffic and sensor densities. Next, details of Mode S are presented with an emphasis on the im-provements provided by the monopulse direc-tion-finding techniques and the specific fea-tures provided by Mode-S surveillance and the Mode-S integral data link. A descrip-tion follows of the field measurements that were made to validate the Mode-S design. Fi-nally, this article comments on the current status ofMode-S implementation in the United States.
'l11e Lincoln Laboratory Journal, Volume 2, Number 3 (1989)
Air Traffic Control Radar Beacon System
Figure 1 schematically illustrates the opera-tion of the current ATCRBS. The ATCRBS an-tenna, which is typically mounted above the primary-radarantenna,hasa fan-beampattern with a horizontal beam width of 2° to 3°. The scan rate is 4.8 s for a sensor used at an airport terminal, and 10 s to 12 s elsewhere. Civil ATCRBS transponders accepttwo types ofinter-rogations. Mode A has an 8-f.Js p)-to-P3 spacing and elicits a 20.3-f.Js reply that contains one of 4,096 pilot-entered identity codes. Mode C has a 21-,us p)-to-P3 spacing and elicits a similar reply that contains the aircraft's barometric altitude referenced to standard atmospheric conditions. The purpose of the Pz pulse is described in the following section.
ATCRBS Transmit Sidelobe Suppression
At short ranges, an antenna's signal strength may be high enough that transponders are interrogated via leakage through the antenna sidelobes. For control of this phenomenon, air-craft in the antenna-sidelobe regions are pre-vented from replying by a technique known as transmit sidelobe suppression (SLSJ, illustrated in Fig. 2. In transmit SLS, the Pz pulse of the interrogation is transmitted from an omnidirec-tional antenna at a slightly higher power level than the interrogator power produced by the antenna sidelobes. Transponders are designed to reply only if the received PI pulse is greater than the received Pzpulse. Note that this condi-tion is not satisfied in the antenna-sidelobe regions.

/ Omnidirectional Antenna Pattern P2
,

Fig. 2-Sidelobe-suppression (SLS) operation.
ATCRBS Limitations
The current ATCRBS satisfies operational requirements in most airspace. The system, however, has the following limitations in regions of high traffic and sensor densities.
(1)  
Synchronous garbling (described later).

(2)  
Azimuth inaccuracy (described later).

(3)  
Fruit. Replies received from interroga-tions by neighboring sensors are called fruit. These unwanted replies are not synchronized with the local sensor's interrogations, and are thus received at random times. The presence offruit can interfere with the reception of a wanted reply. As a result, high fruit rates can produce a detectable decrease in per-formance. The use of high pulse-repeti-tion frequencies (PRF) for sliding-window detection contributes to this problem. (Sliding-window detection will be discussed in a following section.)

(4)  
Overinterrogation. In a region containing manysensors, a transponderwill receive a high rate of interrogations and SLSs. Consequently, the transponder may be unable to replywhen it receives an inter-rogation from the local sensor. As is the case with fruit, the use of high PRFs aggravates overinterrogation.

(5)  Aircraftidentification. In many regions of the world, the limit of 4,096 different Mode-A codes is insufficient.
The Lincoln Laboratory Journal. Volume 2. Number 3 (1989)
Orlando -The Mode S Beacon Radar System
Synchronous Garbling of ATCRBS Replies
Synchronous garbling occurs when two ATCRBS aircraft (shown asA and Bin Fig. 3) are within about 3° in azimuth from an ATCRBS radar and when their slant ranges (Le., their line-of-sight distances from the sensor) differ by less than 1.64 nautical miles (nmi). Under such conditions, the transmitted interrogation elicits replies from both transponders, and the replies overlap at the receiver. The overlap can lead to missing or incorrectly decoded replies, which results in a loss of information on the control-ler's display. The loss persists until the aircraft change their relative positions. Thus the reply overlap canlastfor many scans, hence the name synchronous garble. Note that the altitudes of the two aircraft do not have to be close for garbling to occur.
Azimuth Inaccuracy
Current ATCRBS sensors in the United States use sliding-window detection (Fig. 4), a technique that determines the azimuth of an aircraft by marking the center of the aircraft's run length. (A run length is a series oftranspon-der replies that are observed as the antenna beam of a sensor scans past an aircraft.) A leading-edge detector determines the beginning of a run length by detecting the presence of a minimum of m replies from n reply opportuni-ties. (Following an interrogation, the detector

Aircraft
Azimuth

Characteristics: High Pulse-Repetition Frequency, Susceptible to Azimuth Splits
Fig. 4-Sliding-window beam splitter.
waits a certain time, called the listeninginterval, for a reply to be issued.) A similar algorithm is used for trailing-edge detection. Once the lead-ing and trailing edges have been determined, the aircraft azimuth is calculated as the center of the run length with an offset to account for the bias that the edge detectors introduce.
To ensure accuracy, a sliding-window detec-tor requires a relatively small interval between successive replies. Typically, a PRF of approxi-mately 400 interrogations/s is used, which produces the 15 or more replies required for sliding-window beam splitting. A disadvantage of such a high PRF, however, is that it can interfere with the operation of neighboring sensors.
Another disadvantage of a sliding-window beam splitter is a susceptibility to azimuth splits, which occurwhen interference (e.g., from fruit) or blockage from a physical obstruction, such as a building, causes a loss of data in the centerofthe replyrun length. The loss results in the false declaration of a trailing edge followed by a leading edge, which leads to the erroneous declaration that there are two aircraft instead of one. To make matters worse, neither of the two target reports contains the correct azimuth of the one aircraft.
The Monopulse Technique
To address the many limitations ofATCRBS, Mode-S sensors use the monopulse method [3] in which only one reply is reqUired to determine an aircraft's azimuth. Monopulse azimuth de-termination requires an antenna with two types of beam patterns (Fig. 5):

(1)  
Sum beam (labeled Lin Fig. 5[a].) A sum beam is equivalent to the single main r beam of a nonmonopulse antenna.

(2)  
Difference beam (labeled,1 in Fig. 5[a].) A differencebeamiscomposed oftwo lobes with a null at the antenna boresight.

Areply received from a target that is an angle eoffboresight produces different signal ampli-tudes from the receivers associated with the sum and difference beams. The monopulse processor uses these amplitudes to calculate a return signal that is a function of ,1/L, Le., the ratio of the signal amplitudes in the difference and sum channels. The ,1/L value is then used to obtain efrom a graph of ,1/L versus e(Fig. 5). (The graph was derived with calibration of the sensor against a fixed transponder located near the sensor.)
Thus the use ofmonopulse makes it possible to estimate the azimuth for each reply. This capability prevents azimuth splits.

ATCRBS Monopulse Azimuth Determination
Monopulse makes surveillance of ATCRBS aircraft at very low interrogation rates possible. In theory. monopulse surveillance can be per-

,
formed with as little as one Mode-A and one Mode-C reply opportunity per scan. In practice. however. additional replies are needed to ensure correct Mode-A and Mode-C code reception and to suppress false alarms. The Mode-S sensor interrogates at a rate sufficient to elicit two replies for each ATCRBS mode within the ATCRBS antenna's 3-dB beamwidth of 2.4°. This capability leads to a PRF of approximately 100 pulses/so which is about one-quarter ofthe current interrogation rate ofATCRBS.
Monopulse Degarbling of ATCRBS Replies
A second benefit of monopulse is that it
Orlando -The Mode S Beacon Radar System

enables the degarbling of ATCRBS replies [4). Figure 6 shows two aircraft (labeled A and B) thatare simultaneouslyinthe mainbeamofand near the same slant range from a radar. The received signal data show an interleaved mix of code pulses from the two aircraft. Referencing the monopulse data enables the pulses to be identified easily.
In the example of Fig. 6. the pulses do not overlap; hence they could have been sorted into the correct replies if the pulse-timing data were used. However. in instances of pulse overlap that cannot be resolved by timing alone. mono-pulse degarbling can sort the pulses.

Mode-S Surveillance
The principal features ofMode-S surveillance [5) are as follows.
Selective addressing. Mode-S signal formats enable the selective interrogation of individual Mode-S transponders. More than 16 million

 

Reply
Reply B
n rLJl

Fig. 6-Monopulse degarbling ofATeRBS replies.
The Lincoln Laboratory Journal. Volume 2. Number 3 (1989)
addresses are available, enough for each aircraft in the world to have its own unique address.
Adaptive reinterrogation. Selective address-ing enables the reinterrogation of an aircraft when necessary. The reinterrogation can be performed without involving all of the other aircraft in the beam. This feature allows sched-ulingofa subsequentinterrogation ofan aircraft when the expected reply is not received. Reinter-rogation can significantly improve the probabil-ity of detecting an aircraft in a marginal signal condition due, for example, to any shielding of the aircraft's antenna as the vehicle banks during a tum.

Monopulse beam splitting. The development of monopulse was critical to Mode S. Selective addressing makes a sliding-window detector impractical because of the lack of channel time necessary to conducta selective interrogation of each aircraft 15 or more times.
Single surveillance interrogation andreplyper antenna scan. The use of monopulse, coupled with a more capable data format that provides altitude and the Mode-S address in a single reply, makes routine surveillance with one transaction (Le., one interrogation and reply) per scan possible.
All-call acquisition. An all-call interrogation elicits replies from all Mode-S aircraft that are not being selectively interrogated. By periodi-cally transmitting all-call interrogations, Mode-S sensors can obtain the addresses ofunknown aircraft.
Lockout. Once a Mode-S aircraft is acqUired via an all-call interrogation, the vehicle is in-structed to ignore (Le., to lock out) subsequent all-call interrogations. This lockout option re-duces the probability of synchronously garbled all-call replies.
Error detection and correction. The Mode-S data formats enable an extremely high degree of error detection. The system boasts a rate ofless than one undetected error in 108 messages (6). In addition to detection, error correction is pro-vided on the downlink (7).
Basic Mode-S Surveillance Interrogation and Reply Formats
Figure 7 shows the basicMode-S surveillance
INTERROGATION
All-Call

r
Discrete
REPLY
All-Call
Discrete

Fig. 7-Basic Mode-S surveillance formats.

formats [8, 9], which include the following. All-callinterrogation. This format contains the same Pj, P2 , and Ppulses that ATCRBS uses.
3

The additional pulse P4 labels this format as originating from a Mode-S sensor. When an ATCRBS transponder receives a Mode-S all-call interrogation, the transponder cannot detect the Ppulse. It therefore responds with the
4

appropriate Mode-AorMode-C reply, depending on the spacing of the Pj and P3 pulses. On the other hand, a Mode-S transponder will detect the Ppulse and, if it is not in a lockout state,
4

respond with an all-call reply. Thus one interro-gation can satisfybothATCRBS and Mode-S all-call requirements. Because of this feature. the Mode-S all-call interrogation format is also re-ferred to as the Mode A/C/S all-call. Note that a Mode-S transponder will never generate an ATCRBS reply to a Mode-S sensor's Mode-A/C/S all-call interrogation. This detail is im-portant since it ensures that a Mode-S aircraft will never be reported as both a Mode-S and an ATCRBS vehicle.
All-callreply. The reply ofa Mode-S transpon-der to a Mode-S all-call interrogation is com-posed largely of the aircraft's Mode-S address. which is used in subsequentselective interroga-tions of the vehicle.
Discrete interrogation. This format contains
The Lincoln Laboratory Journal. Volume 2. Number 3 (J 989)
Orlando -The Mode S Beacon Radar System
the Mode-S address ofthe aircraft for which the interrogation is intended, surveillance informa-tion, and communication-control information.
Discrete reply. The basic surveillance reply to a discrete interrogation contains the aircraft's altitude code and Mode-S address.
Mode-S Elimination ofSynchronous Garble
Selective addressing completely eliminates the problem of synchronous garble. As before, the two aircraftin Fig. 8 are atthe same azimuth and slant range. The sensor has knowledge of the azimuth and range ofeach aircraft from the previous scan, and schedules an interrogation to one oftheaircraft (A in Fig. 8). The sensorthen schedulesan interrogationofthe second aircraft such that both the interrogation and reply for aircraftBwill occur attimes thatdo notinterfere with the reception of the reply from aircraft A. The scheduling technique can be extended to cover three or more aircraft.
Synchronous garble can also be handled in the all-call acquisition mode. The technique, known as stochastic acquisition, uses a special all-call interrogation that instructs the different aircraft to reply with a predetermined probabil-ity ofless than one. The resulting random loss of replies ensures that a reply from exactly one of the aircraft contributing to the garble will be received after a few interrogations. Mter acqui-

A
B

Interrogation --"-..'---,"~:---------
Reply  --.l-.·~·-.L.-.:iI-.
Fig. 8-Mode-S elimination ofsynchronous garble.

The Lincoln Laboratory Journal. Volume 2. Number 3 (1989)
sition, an aircraft is locked out to further all-call interrogations. The process is repeated (and the probabilitymay be increased) until all aircraft in the garbling region are acquired. Take the sim-plest case. If two aircraft are contributing to a garbled condition, both vehicles could be in-structedto reply witha probabilityof0.5. Ifboth vehicles reply again or if neither of them does, the aircraft are reinterrogated until only one reply is received. The aircraft that replies is then discretelyinterrogated and subsequentlylocked out so that the other vehicle can then be easily acquired.
ATCRBS-to-Mode-S Compatibility
Compatibility between ATCRBS and Mode S is achieved through the following.
(1)  
Mode-S transponders will respond to ATCRBS interrogations. However, a conventional ATCRBS sensor will detect a Mode-S-equipped aircraft as an ATCRBS aircraft.

(2)  
ATCRBS-equipped aircraft will reply to Mode-S sensors.

(3)  
Mode S operates on the same I,030-MHz uplink and I,090-MHz downlink that are used by ATCRBS. The use of the same frequencies greatly simplifies the construction of Mode-S sensors and transponders because the same trans-mitters and receivers will be able to handle both ATCRBS and Mode-S transmissions.

(4)  
Mode-S waveforms were designed to prevent mutual interference with ATCRBS signals. This provision is de-scribed in greater detail in a following section.

 

ATCRBS Mode-S Time-Sharing
Through time-sharing, a Mode-S sensor can provide surveillance of both ATCRBS and Mode-S aircraft. Figure 9 shows the time line, drawn approximately to scale, of a typical Mode-S sensor. During the time of one beam dwell, the Mode-S sensor provides four Mode-A/C/S all-call periods. (A beam dwell is the
I+-Beam-Dwell Time---l /ATCRBS/Mode-s All-Call Periods

Time
'Mode-S Roll-Call Periods
Fig. 9-ATCRBS-Mode-S time sharing.
typical time, approximately 30 ms for a terminal sensor, that a beam is actually on an aircraft.) This scheduling provides the interrogation and listening intervals required by the mandatory two Mode-A and two Mode-C replies. All-call ac-quisition is also performed dUring this time.
Selective Mode-S interrogations are sched-uled during the Mode-S roll-call periods. Note that the use of monopulse for ATCRBS enables the sensor to devote most of the time line to Mode S.
Mode-S All-Call Interrogation
Figure 10 shows the all-call interrogation waveform that Mode S normally uses. As stated earlier, thewaveformiscomposedofthesamePI and P3 pulses used for ATCRBS interrogations. The P4 pulse identifies the interrogation as originating from a Mode-S sensor.
Ifthe Ppulse is 1.6 ).ls long, the interrogation
4
Mode A: 8.0 JJs Mode C: 21.0 JJs 2.0 JJs I' .I. ., 
Interrogation  ~~ 1-1 1-1 I-t 0.8 JJs 0.8 JJs *JJs 
I~.O JJ~I 
SLS Control  .-J'P~'L...JP;1'---------

Transmission I-f
0.8 JJs
* Mode-AIC/S All-Call: 1.6 JJs Mode-AlC-Only All-Call: 0.8 JJs
Fig. 1Q-Mode-Sall-call interrogation waveform.

is a Mode-A/C/S all-call format. which elicits Mode-A/C replies from ATCRBS transponders and Mode-S all-call replies from unlocked Mode-S transponders. If the P4 pulse is 0.8 ).lS long, the interrogation is a Mode-A/C all-call format, which elicits Mode-A/C replies from ATCRBS transponders and no reply from Mode-S transponders. Mode-S sensors use this waveform in connection with special Mode-S-only all-call interrogations. The Traffic Alert and Collision Avoidance System (TCAS) also uses the waveform for surveillance of ATCRBS aircraft. TCAS acquires Mode-S aircraft pas-sively by listening for all-call replies (known as squitters), which are spontaneously generated approximately once per second by all Mode-S transponders.
The Pz pulse is used for SLS of the Mode-A/C and Mode-A/C/S all-call interrogations in the manner described earlier for ATCRBS interrogations.
Mode-S Addressed Interrogation
A Mode-S addressed interrogation (Fig. 11) begins with a two-pulse preamble followed by a data block. The data block is encoded with differential phase-shiftkeying (DPSK) at a rate of 4 Mb/s. A logical 1 is encoded as a 1800 phase shift and a logical 0 as the absence of a phase shift. DPSK was selected because of its resis-tance to interference [10]. All data blocks begin with a sync phase reversal that establishes the timing for the remaining phase-reversal posi-tions. The data block is either 16.25).ls or 30.25 ).ls long, and provides for either 56 bits or 112 bits of data. respectively.
The two-pulse preamble is an important ele-ment in the reduction of mutual interference with ATCRBS. It is important that ATCRBS transponders notrespond to Mode-S interroga-tions. Ifthey did, such responseswould produce ATCRBS replies that would interfere with the reception of Mode-S replies. Moreover, the ATCRBS replies would be synchronized with the Mode-S reply and would therefore lead directly to synchronous garble. Tests with many ATCRBS transponders indicated that there is no practical waveform that is invisible to all
TIle Lincoln Laboratory Journal. Volume 2. Number 3 (1989)
Orlando -The Mode S Beacon Radar System

___S_L_S_c_o_n_tr_o_I_T_ra_n_s_m_is_s_io_n_---'Ae.-_
.
Differential Phase-Shift Keying (DPSK) Modulation

.
Data Rate: 4 Mb/s

 

Fig. 11-Mode-S interrogation waveform.
ATCRBS transponders. Consequently, the ap-proach that was adopted was to precede the Mode-S data blockwith a preamble consistingof aPj and P2 pulse of equal amplitude. When received by an ATCRBS transponder, these pulses put the transponder into a period of suppression for 35 )1s. The data block is sent dUring this interval and is therefore undetect-able by the ATCRBS transponder. Note that the length of the suppression interval dictated the maximum length of the Mode-S data block.
SLS is not required with a selectively ad-dressed Mode-S interrogation, since the interro-gation is transmitted while the addressed air-craft is in the main beam of the antenna. Also, other Mode-S aircraft in the antenna sidelobe regions that receive the interrogation will not reply, since the interrogation does not contain their Mode-S address.
On the other hand, the Mode-S-only all-call format requires SLS. The P2 pulse cannot be used for SLS because it is used in the Mode-S preamble to prevent ATCRBS replies to Mode-S interrogations. Thus SLS for Mode S is provided by an additional pulse, which is shown as P5 in Fig. 11. This pulse is transmitted in an omnipat-tern at the time of the sync phase reversal. The presence of P5 obliterates the sync phase rever-sal for aircraft in the sidelobes, an action that makes decoding the interrogation impossible.
Note that the Mode-S transponder does not require any SLS circuitry because the suppres-sion occurs through controlled interference of the sync phase reversal. However, since P5 must be transmitted at the same time as the data block, a second transmitter (of a much lower duty cycle than the main transmitter) is needed in the Mode-S sensor.
Mode-S Reply
A Mode-S reply (Fig. 12) begins with a four-pulse preamble followed by a data block. The data block is encoded with pulse-position modulation (PPM) at a rate of 1 Mb/s-the highest rate compatible with a low-cost imple-mentation for the Mode-S transponder. In PPM, alogical 1 isencodedasthepresenceofa0.5-)1s pulse in the first halfof the 1-)1s data interval of Fig. 12. Alogical 0 is encoded as the presence of a 0.5-)1s pulse in the second halfofthe interval. The data block is either 56 )1S or 112 )1s long, thus providing either 56 data bits or twice that amount.
PPM was chosen because it provides bit-interference detection. PPM detects interference ofa data-bit position by comparing the received energy in both halves of the data chip interval. The absence of interference results in received energy in only one half of the interval. Thus
The Lincoln Laboratory Journal. Volume 2. Number 3 (1989)
energy in both halves indicates that an interfer-ing pulse was received at the same time.
PPM was also chosen because it enhances monopulse performance. The data block in Fig. 12 contains either 56 bits or 112 bits, independ-ent ofthe data content. The constant number of pulses makes it possible for a monopulse azi-muth estimate for the PPM reply to be based uponalargenumber(e.g., 16 or32)ofindividual pulse azimuth measurements. This feature contrasts with the pulse-amplitude modulation ofATCRBS in which a message composed of all logical as does not contain any data pulses.
Mutual interference with ATCRBS is man-aged on the downlink with a preamble whose spacing is selected so that it is unlikely to be synthesized by ATCRBS replies. The preamble minimizes the possibility that the Mode-S reply processor will be busy handling ATCRBS fruit when the elicited Mode-S reply is received. A second technique for managing mutual interfer-ence is called burst error correction, which can error-correcttheeffectsofa singleATCRBSfruit reply received at the same time as the Mode-S reply.
Mode-S Data Link
TheselectiveaddressingofModeS providesa natural mechanism for a data link that would supply the capacity and performance required to support air traffic services. The communica-tions-control features of the Mode-S data link have been designed for compatibility with the Open Systems Interconnection Reference Model.
The link design provides for both ground-to-air and air-to-ground message transfers. Air-to-ground messages may be either pilotinitiated or ground initiated. The latter type enables the efficient reading of the technical information available on board an aircraft, e.g., information such as an aircraft's roll angle, which predicts the aircraft's tum.
Messages sent over the Mode-S link benefit from the high degreeoferrorprotection provided by the link's design. For example, the system acknowledges message delivery. (On the uplink, the receipt of a reply to an interrogation that contained a message constitutes the technical acknowledgment ofthat message.) In all critical applications, provision is made so that an air-craft crewcan acknowledge message receipt and acceptance.
Mode-S Data Formats
Mode-S formats (Fig. 13), which are either 56 bits or 112 bits long, all contain a 24-bit address and parity field [11]. The address and parity functions were combined into a single field to minimize the channel overhead. For an interro-gation, a Mode-S sensor generates a 24-bit parity field from the entire message (either 56

1.....4~---Preamble 8.0 Jls----i..~r--Data Siock 56 or 112 JlS 4
Sit ISit 11Sit 21Sit 31Sit 41 IN -1 Isit NI
J1Jl IUl :1:-0: -1-:0-:1-:0:1-:0:--:--r-:-1-:0-:1:-6: 60~51 --3:5 4:5 ---~8io---":""9~io""""""'~I-~1 ~~C2-2~1 ~I~~I ~I-
1.0 ~~
Time (J1s) 1 1 a 1 10
.
Pulse-Position Modulation (PPM)

.
Data Rate: 1 Mb/s

 

Fig. 12-Mode-S reply waveform.
Orlando -The Mode S Beacon Radar System

..

Fig. 13-Mode-S data formats. (a) Surveillance interrogation and reply. (b) Surveillance and communication interroga-tion and reply. (c) Communication interrogation and reply.
bits or 112bits long) and overlays the parityfield on the address field to form the address and parity field.
The transponder, when it receives an interro-gation, performs a complementary decoding process. Ifthe message has been received error free, the transponder will recover and process the intended Mode-S address from the address and parity field. On the other hand, one or more errors anywhere in the message will change the decoded Mode-S address. In this case, the trans-ponder will not accept the message, since it appears to be meant for another transponder.
Surveillance formats (Fig. 13[a]) contain sur-veillance and communication-control informa-tion. On the downlink, the formats also convey Mode-C altitude or Mode-A identity codes.
Surveillance and communication formats (Fig. 13) contain all of the fields of a surveil-lance format with an additional 56-bit mes-sage field, which permits simultaneous data-link and surveillance activities. The design of
The Lincoln Laboratory Journal, Volume 2, Number 3 (1989)
the Mode-Sdatalinkallows the connectionofup to four of the 56-bit message fields into a single message entity.
The communications formats (Fig. 13[c]) handle longer data-link messages. The use of communications formats provides greater link efficiency in two ways. First, since the message field is longer, fewer interrogations or replies are required for a given message. Second, the message transfers use the Extended Length Message (ELM) protocol. ELM permits a single reply to acknowledge up to 16 communication interrogations, which conserves channel time. A similar approach is used for downlink ELMs. The sixteen 80-bit message fields provide a message length of 1,280 bits; the design ofthe Mode-S data link allows the connection ofup to 32 of these messages.
Mode-S Data-Link Characteristics
The Mode-S data link offers characteristics
that are well suited to the needs of air traffic services. For instance, the association of the Mode-S data link with the surveillance function offers a number of operational benefits. Com-munication with an aircraft can be established based solely on surveillance detection. Thus a message can be sent to an otherwise unidenti-fied aircraft. This capability is important for safety services; e.g., warnings can be sent to an unidentified aircraft that is flying too low or that is heading into controlled airspace. Since the same address is used for both surveillance and communication, the possibility of sending a message to the wrong aircraft because of an error in cross-referencing the surveillance and communication identities is eliminated. A fur-ther operational benefit is that communications coverage is assured whenever surveillance coverage exists.
Because of the inherent characteristics of Mode-S sensors, the Mode-S link discourages both accidental and intentional jamming. With a Mode-S radar, coverage is restricted to the sensor's line of sight. This restriction limits not only the airspace in which the sensor can cover traffic, but also the area in which an interfering source can affect the sensor. A narrow antenna beam further limits the active area. Thus a single interfering source, if it is outside the sidelobe region of the sensor, will prevent operation in only a single antenna beamwidth. In the worst case, in which the interfering source is within the sidelobe region (i.e., within a distance of about five miles from the sensor), the source could prevent operation of that sen-sor in all directions. However, the interfering source would have little effect on the opera-tion of all other Mode-S sensors. Thus the distributed nature of Mode S makes it tolerant of interference.
Mode-S Experimental Facility
The Mode-S design was first validated at the Mode-S Experimental Facility (MODSEF), lo-cated at Lincoln Laboratory (Fig. 14). Initially used for link measurements and monopulse development, MODSEF was later upgraded to a fully functional Mode-S sensor. Validation at MODSEF, however, is not sufficient verification of Mode S because the site does not experience high traffic density, fruit, or ground reflections known as multipath.
Transportable Measurements Facility
The Transportable Measurements Facility (TMF) [12) was constructed for observation of the Mode-S sensor operation at FAAsites known to provide environmental difficulties. TMF (Fig. 15) includes its own antennas, tower, and equipment van that contains a transmitter, receiver, and digitizing and recording equip-ment. The antennas shown in the figure are an Airport Surveillance Radar-7 antenna with a monopulse beacon feed and a monopulse-capable antenna on loan from the United Kingdom. In operation, TMF transmits and then digitizes the received video pulses, which are recorded for subsequent playback and analysis at Lincoln Laboratory.
At most sites, the TMF was operated as close as possible to the existing FAA sensor in order to experience similar environmental con-ditions. Figure 15 shows TMF at Washing-ton's National Airport; the operational ASRis in the background. In two cases, new off-air-port sites were selected to determine their
.

-,

effect on sensor performance.
TMF's initial operation and testing took place at Lincoln Laboratory and Boston's Logan Inter-national Airport. The system was then trans-ported to Philadelphia and Washington to expe-rience the traffic and interference levels of the northeast corridor. Las Vegas, a location well known for its significant ground-bounce multi-path problem, was the next site. Final TMF measurements were made atLosAngeles, which suffers from the highest airport traffic and inter-ference levels in the United States.
Comparison of Automated Radar Terminal System with TMF
At each of the above sites, TMF data was recorded simultaneously with data from the existing FAA sensor. The data enable a direct comparison between the monopulse processing ofTMFand thecurrentAutomatedRadarTermi-nal System (ARTS) processing.
Figure 16 shows an example ofthis compari-son for data collected in Philadelphia for an area of 80 nmi by 80 nmi (13). Figure 16(a) shows ARTS data; Fig. 16(b) shows TMF data. Each pointrepresents an unsmoothed position report measured once each antennascan. The reduced measurement errors of theTMF data are readily apparent.
Orlando -The Mode S Beacon Radar System

Surveillance Performance Comparison
Table 1gives a quantitative comparison ofthe average performance of ARTS versus TMF for all of the TMF sites. The table confirms the greatly improved qualitative performance seen in Fig. 16.

The Lincoln Laboratory Journal. Volume 2. NlLmber 3 (J 989)
Table 1. Surveillance Performance Comparison
Arts  TMF Monopulse 
All  Crossing  All  Crossing 
Blip/Scan Azimuth Error (1 a) Range Error (1 a)  94.6% 0.16° 124 ft  86.9%  98.0% 0.04° 24 ft  96.6% 

In the table, the blip/scan ratio is the proba-bility that the system will generate a target report for a specific aircraft on a given scan. When all aircraft are considered, the blip/scan ratio is 94.6% for ARTS and 98.0% forTMF. The most significant difference in blip/scan per-formance is revealed when only crossing tracks are considered. Crossing tracks are cases in which aircraft are close enough to present a possible synchronous garbling problem. For this subset of aircraft, the blip/scan ratio for ARTS dropped to 86.9%, while the performance of TMF remained at 96.6%. This result clearly indicates thebenefitofmonopulse processingin resolving garbled replies.
Monopulse processing was also responsible for the TMF's substantially smaller measured-azimuth error of 10" = 0.04°. Furthermore, the 1-0" range error for TMF was more than five times less because of an improvement in mea-suring the time of arrival of replies.
Engineering-Model Sensors
A major step in validating the Mode-S design occurred in 1975 when the FAA awarded Texas Instruments a contract for the development of three engineering-model sensors. In 1977, Texas Instruments delivered the three sensors to the FAA Technical Center for extensive field evaluation. Figure 17isa drawingofoneofthose sensors.
Aircraft Reply and Interference Environment Simulator (ARIES)
The engineering-model sensors were built to demonstrate a sensor with the capacity to handle a maximum of400 aircraft in 360°, and apeakof50aircraftina sectorof11.5°. Capacity tests ofthese sensors could not be accomplished with real ATCRBS aircraft because an aircraft density ofsuch magnitude did not exist. even in the highest densities of Los Angeles. Further-more, only a small number ofMode-S transpon-ders were available.

Fig. 17-Mode-S engineering-model sensor.

.

Fig. 18-ARIES with the Mode-S engineering-model sensor.
Consequently, capacity testing of the en-gineering-model sensors was accomplished with a traffic simulator known as the Aircraft Reply and Interference Environment Simulator (ARIES) [14-16). ARIES (Fig. 18) is interfaced at the analog level with the front end of a sensor and thus exercises the entire sensor. notjustthe computer subsystem. In operation. ARIES lis-tens to interrogations from the engineering model, and then inserts signals into the front endatthe timethatthetransponderreplywould have been received from the real aircraft. ARIES also correctly simulates the monopulse signals according to the off-boresight angle ofthe simu-lated aircraft. This monopulse simulation is accurate enough to permit operation with a mix of simulated and real aircraft.

ARIES Capacity Testing
A principal objective of developing the engi-neering-modelsensorwas toverifYthat Mode-S-sensor algorithms could achieve the reqUired surveillance and communication capacity for Mode-S operation. A traffic model that repre-sented a future worst-case scenario for the Los Angeles Basin enabled capacity testing of ARIES.
Figure 19 shows a typical display of traffic

The Lincoln Laboratory Journal. Volume 2. Number 3 (J 989;
Orlando -The Mode S Beacon Radar System

that the engineering-model sensor processed during capacity testing. In the figure, a square indicates a Mode-S aircraft. and a circle an ATCRBS aircraft. The total traffic load is over 300 aircraft, most of which are contained in a 90°, 60-nmi sector.
Implementation
The FAA is procuring 137 dual-channel Mode-S sensors from ajoint venture comprised of Westinghouse Electric Corp. and Unisys Corp. The first operational implementation at a site is scheduled for 1991. To outfit all of its beacon-radar sites with Mode S. the FAA is currently considering an additional purchase of 259 Mode-S sensors.
The sensors are designed to provide a total communication data rate of 92.5 kbls for a target load of 700 aircraft. Thus the initial sys-tem of 137 sensors will have a total capacity of more than 12 Mb/s. The characteristics of the sensor determine maximum data-link transfer to a specific aircraft. The production version of the Mode-S sensor, which has a rotatingnarrow-beam antenna, can deliver up to 360 bls on the uplink and 313 bls on the downlink for a total eqUivalent simplex rate of 673 b/s. A next-generation sensor eqUipped with an electroni-cally scanned antenna could transfer data to aircraft at a rate as high as 5 kb/s.
Fig. 19-ARIES traffic plot.

Summary
Mode S, an evolutionary improvement ofthe current ATCRBS, provides enhanced surveil-lance performance through monopulse, discrete addressing, and error protection. In addition, Mode S includes an integral data link with unique benefits to ATC because of the link's association with the surveillance function and its resistance to interference.
Extensive field measurements and the devel-opment of an engineering-model sensor have
.

validated the Mode-S techniques. The FAA is currently implementing Mode S for operational use in the United States.
Acknowledgments
The development of the Mode-S beacon sys-tem required the efforts of many individuals at Lincoln Laboratory. This article is dedicated to them and to our sponsor, the FederalAviation Administration.
.

References
1.  
V.A. Orlando and P.R Drouilhet, "Mode S Beacon System: Func~ional Description," Project Report ATC-42D, Lincoln Laboratory (29 Aug. 1986), FAA/PM-86/19.

2.  
SelectionOrder:U.S.NationalStandardfortheIFFMark X (SIF) Air Traffic Control Radar Beacon System (ATCRBS) Characteristics, Dept. of Transportation/ Federal Aviation Administration Order 1010.51A (8 Mar. 1971).

3.  
D. KarpandM.L. Wood, "DABS Monopulse Summary," Project Report ATC-72, Lincoln Laboratory (4 Feb. 1977), FAA-RD-76-219.

4.  
J.L. Gertz, "The ATCRBS Mode of DABS," Project Re-port ATC-65, Lincoln Laboratory (31 Jan. 1977), FAA-RD-76-39.

5.  
E.J. Kelly, "DABS Channel Management," Project Report ATC-43, Lincoln Laboratory (8 Jan. 1975), FAA-RD-74-197.

6.  
J.T. Barrows, "DABS Uplink Coding," Project Report ATC-49, Lincoln Laboratory (25 July 1975), FAA-RD-74-62.

7.  
J.T. Barrows, "DABS Downlink Coding," Project Re-port ATC-48, Lincoln Laboratory (12 Sept. 1975), FAA-RD-75-61.

8.  
J.D. Welch and P.H. Robeck, "Proposed Technical Characteristics for the Discrete Address Beacon Sys-tem (DABS)," Project Report ATC-71 , Lincoln Laboratory (30 Sept. 1977), FAA-RD-77-143.

Orlando -The Mode S Beacon Radar System
9.  
Selection Order: U.S. National Aviation Standard for the Mode Select Beacon System (Mode S), Federal Aviation Administration Order 6365.1A (3 Jan. 1983).

10.  
T.J. Goblick, "DABS Modulation and Coding Design-A Summary," Project Report ATC-52, Lincoln Labora-tory (12 Mar. 1976), FAA-RD-75-93.

11.  
J .L. Gertz, "Fundamentals of Mode S Parity Coding," Project Report ATC-]] 7, Lincoln Laboratory (2 Apr. 1984), DOT/FAA/PM-83/6.

12  RR LaFrey, J.E. Laynor, RG. Nelson, and RG. Sand-holm, "The Transportable Measurements Facility (TMF) System Description," Project Report ATe-9] , Lincoln Laboratory (23 May 1980), FAA-RD-79-111.
13.  
W.1. Wells, "Verification of DABS Sensor Surveillance Performance (ATCRBS Mode) at Typical ASR Sites throughout CONUS," Project Report ATC-79, Lincoln Laboratory (20 Dec. 1977), FAA-RD-77-113.

14.  
M. Goon and D.A. Spencer, "The Aircraft Reply and Interference Environment Simulator (ARlES): Prin-ciples of Operation," ProjectReportATC-871, Lincoln Laboratory (22 Mar. 1979), FAA-RD-78-96.

15.  
M. Goon and D.A. Spencer, "The Aircraft Reply and Interference Environment Simulator (ARlES): Appen-dices to the Principles of Operation," Lincoln Labora-tory Project Report ATC-87 2. Lincoln Laboratory (22 Mar. 1979), FAA-RD-78-96.

16.  
M. Goon and D.A. Spencer, ''The Aircraft Reply and Interference Environment Simulator (ARlES): Pro-grammer's Manual," ProjectReportATe-873. Lincoln Laboratory (22 Mar. 1979). FAA-RD-78-96.

 

1he Lincoln Laboralory Journal. Volume 2, Number 3 (1989)
.

Orlando -The Mode S Beacon Radar System

l,
VINCENT A. ORLANDO is currently in charge of the System Design and Evalu-ation Group at Lincoln Laboratory. where his re-search has focused on air traffic safety and capacity. Beforejoining the Laboratory 17 years ago. Vince spent 1Oyears workingon the development of military command and control systems. He received a

B.S. in electrical engineering from the University ofCincin-nati, an M.S. in statistics from Stanford University. and a Ph.D. in systems and information science from Syracuse University. A member of Tau Beta Pi and Eta Kappa Nu. Vince was the recipient of a National Science Foundation Traineeship.

 

bocome

The Mode S Beacon Radar System