A VLSI to cost effectively implement
Aircraft Collision Avoidance System

56 bit Mode S single packet position and velocity data squitter

New ICAO Mode S downlink format number single packet transmission
$GPRMC 2D GNSS/Loran-C position and velocity data input
Mode 3C pressure encoder altitude data input

1090 MHz transponder FAA system design failure repair

Mode 3A & Mode 3C tolerant of pre- (P0) and post- (P4) pulse interference
And no more need for biennial "digital circuitry adjustments"

B. Keith Peshak: keith.peshak@gtwn.net

To work at all, the ATC "system" requires that all aircraft transponders work correctly all of the time. The FAA officially reports that 96% of transponders don't work right (http://www.gtwn.net/~keith.peshak/tn97_7.pdf), and has had the exact problem explained to them (http://www.gtwn.net/~keith.peshak/P4Problem.htm), and has had that explanation reinforced by their own employees who are also "radar" operators and pilots and aircraft owners (http://home.columbus.rr.com/lusch/talotta.html). There is a solution for the entire problem set, which has been reduced to a single VLSI "system-on-a-chip", which implements the necessary corrected transponder control. The chip has been productized to produce a complete transponder, in order to demonstrate that all problems are sufficiently addressed. This VLSI is intended for very low cost new design transponders, but also for the modification of existing transponders. Originally offered at $200, FAA fumbling of "certification" has raised the cost to $500. FAA outright refusal to test the solution with their specially constructed transponder test and evaluation unique equipment, developed for use in data gathering for their official report on the transponder problem magnitude, slows the certification effort to a standstill. The additional problem of the FAA not requiring its certified repair stations to possess test equipment to evaluate the proper complete operation of transponders under receipt of interrogations containing the P0 and P4 pulse interference complicates the identification or true problem magnitude for the aircraft operator. These would seem improper behavior for a public funded institution of government responsible for correct "system" operation.

Attempts by the FAA to covertly "fix" these defective ATCRBS transponder designs, through issuance of Airworthiness Directives to alter circuitry, have been less than successful (http://www.gtwn.net/~keith.peshak/King%20AD.htm). The FAA does not even understand the problem sufficiently adequately to address the fundamental design failure mechanism (http://www.gtwn.net/~keith.peshak/peshak20.htm). The reader is also referred to the article "The Terra Transponder Saga" by Vernon Barr, published in the U.S.Aviator, August 1994, pp23-26. The FAA has even moved actively to block even recognition of the problem - the reader is referred to http://www.avweb.com/oshkosh/osh99 (click on "Meet The Boss" session with FAA Administrator Garvey, and advance the pointer to 41:00:00).

Working equipment was demonstrated to the higher technical levels of the FAA and the avionics industry. The reader is referred to an audio tape of an AirVenture99 symposium where working prototype ATCRBS and AIS-P augmentation equipment and prototype collision avoidance detection equipment were presented, along with full engineering disclosure to the FAA engineering staff and to the avionics engineering community (available directly from EAA by contacting dcyeoman@juno.com and requesting "#128 TCAS Almost Free"). Peoples lives remain in constant needless danger.
 
 
 
 

The United States FAA has always admitted that it is unable to provide collision avoidance between all equipped aircraft through the ATC system of Secondary Surveillance Radar (SSR) operating on the 1030/1090 MHz pair, utilizing Air Traffic Control Remote Beacon System (ATCRBS) {pronounced "at-crabs"} transponders. One deficiency is equipment related, the other procedure related. The FAA is unable to correct either deficiency.

A "radar" on the ground interrogates on 1030 MHz an aircraft transponder, which replies on 1090 MHz. The name "transponder" is a contraction - transmitter responder. The "radar" shows the transponder at the ground 1090 MHz directional antenna Poynting vector, at the range timed for C/2 (out and back), at the altitude contained in the 3C transponder reply data, for the transponder name contained in the 3A reply data. 3A allows for 4096 aircraft specific identities, with some publicly reserved by international treaty for "man with a gun", "emergency", "I am not participating", "my radio is broken"; and various governments utilize some clandestine codes for communication ("this is my attack code for today") to the ground "radar". 3C allows for 100 foot altitude resolution. There is no distinction between the 3A and 3C reply, allowing data to be confused by the "radar" post processing computer. The reader is invited to analyze the case of the Aegis class Vincennes downing an Iranian commercial passenger airliner; specifically because of confusion between squawk and altitude reply. This could be easily prevented if one or more of the three unused pulses in the transponder packet format were used to distinguish the 3C reply from the 3A reply (guarantee proper data interpretation when there is more than one interrogator). Therefore, the VLSI "fix" for transponders was designed to include that capability.

There are two types of flight. Visual Flight Rules (VFR), where the pilot is responsible for collision avoidance and must remain clear of clouds, and IFR, where the pilot is responsible for collision avoidance but also unable to prevent a collision because he must follow instruction received over a radio (the actual "pilot in command" is the "radar controller"). The only FAA service for collision avoidance guarantees separation between only IFR aircraft from only IFR aircraft, and is unable at this time to maintain the separation guaranteed. The FAA procedure was never intended to guarantee separation of IFR from VFR, VFR from IFR, or VFR from VFR aircraft. Obviously, any collision is likely to precipitously bring down both aircraft involved. Perhaps not so obviously; the cause of accident would be determined to be "pilot error" - failure to "see and avoid". The FAA procedures, therefore, do not provide collision avoidance.

One issue is the processing insufficiency of FAA "radar" computers. Modern FAA "radar" sensors are networked, so target data from all "radars" is, theoretically, available for any display device. However, with the antiquated architecture of FAA computer systems, including the most updated versions now being installed, the target data overwhelms the display driver for any "radar" display device. To reduce the "radar" data to that which can be processed by the compute power of the most modern of the FAA computers, the FAA instituted "sort boxes". FAA "radar" does not show that which all antennas see. Rather, all displays for a given geographic area (sort box) come only from a specific "radar" antenna. The reader is referred to the article "Real Targets-Unreal Displays - The Inadvertent Suppression Of Critical Radar Data" by Thomas G. Lusch, FAA Air Traffic Control Specialist, published in the Journal of ATC, January-March 1992, pp29-33; also published in the Proceedings of the Sixth International Symposium on Aviation Psychology, Ohio State University, pages 460-465, and on the web (http://home.columbus.rr.com/lusch/rtud.html). The reader is referred to the web site on the general subject matter of insufficient processing of available "radar" data by that author (http://home.columbus.rr.com/lusch/). The FAA purposely does not even look at all available "radar" data.

Another issue is that "radar" only works when there is a line of sight between the ground and aircraft antennas. The FAA does not update the "sort box" selection for a geographic area when a high rise hotel goes up near an airport, blocking visibility to a large part of the sky from a given "radar". The same is true even when a "radar" is moved because the airport on which it sits was closed. The use of sort boxes prevent "radar" from working adequately - a vestige from past technology limits.

The final equipment problem is FAA self inflicted. In recent years, the FAA has changed the requirements of the ATCRBS transponder based equipment, without changing the equipment in the aircraft or the Technical Standards Order (TSO) specification for the equipment in the aircraft. It changed the operation of the "radar" to issue false transponder suppression instructions. As a result, aircraft do not appear on "radar". In attempts to solve the problem set introduced by changing the interrogation of the transponder, each time, the FAA has exacerbated the problem magnitude, and enlarged the problem geographic area. The FAA equipment, therefore, cannot provide collision avoidance.

Recognizing the system design deficiencies, but ignoring that a solution exists, the FAA is presently seeking a replacement "radar" system. Throw everything away and start over with something entirely different. This tasks the world with a monumental cost, even if such a working replacement system existed. The FAA favored alternative is called ADS-B and costs, presently, ~$300,000 per aircraft, according to the equipment manufacturer. The mode S standard is altered by ICAO to allow squitter packets (responses sent to no interrogation). Unfortunately, the ADS-B alternative to "radar" cannot cooperatively operate with the ATCRBS system without much costly additional Traffic Information Service (TIS) ground equipment, necessary to provide the squitters for all aircraft not ADS-B equipped. The ADS-B alternative poses serious dangers of risk from terrorists, providing all information necessary to target a specific aircraft without complicated sensory apparatus in a flying weapon (http://www.airsport-corp.com/adsb2.htm). The ADS-B alternative causes data link failure problems to ATCRBS because of the monumental amount of additional transmitted packet traffic necessary on 1090 MHz to get a single message through, according to published analysis by FAA contractor Mitre.

The VLSI is a transponder controller that solves all of the ATCRBS transponder problems listed in the official FAA report. The VLSI fixes all ATCRBS issues, but also augments with ADS-B system design objectives, but also fixes ADS-B deficiencies. Usage is presently blocked by the FAA. Even without the needed augmentation, the VLSI continues to operate in ATCRBS mode without costly periodic "adjustments", presently required by the FAA of all transponders. This VLSI fixes ATCRBS and augments with all of the necessary advantages offered by ADS-B technology, and does so totally without the cost and without the limitations of ADS-B technology. We call this augmentation AIS-P, and it is offered to the world aviation community totally without charge or obligation (http://www.gtwn.net/~keith.peshak/T3.gif).
 
 
 
 

We have always had the problem of "the Los Angeles Basin" with ATCRBS. FAA transponder "radar" just does not work reliably there, because you just can not see where all of the IFR aircraft are all of the time. The problem is traced to too many interrogators there, and too many aircraft in flight there. A bandwidth problem on both 1030 MHz and 1090 MHz pre-exists, since the 1960s, but was limited geographically to "the Los Angeles Basin" (hereinafter, the LAB).

To attempt to solve this insufficient bandwidth problem, the FAA added the Traffic Collision Avoidance System (TCAS), requiring only the major long haul airline passenger aircraft to carry the equipment. At ~$225,000 per airplane to equip, TCAS is its own transponder interrogating "radar". As one would suspect, the result was that the "radar" invisibility problem additionally, then, came to every large hub airport when the sky is busy with TCAS equipped aircraft - just when you need the "system" most. The problem is referred to as N**2, though probably more accurately closer to N**3, where the calculation is the magnitude of the invisibility problem, and N is the number of "radar" interrogators (now inclusive of many aircraft) raised to the third power. Adding long haul commercial TCAS to the "system" drastically increases the number of interrogators, when all of the airplanes in the area get in line for the "five o'clock slam dunk" (maximum runway burst usage period at the hub airport).

Incredibly, the FAA solution to the migration of invisibility out of the LAB was to require that all short haul airline passenger aircraft additionally carry the TCAS equipment! That drastically increased N further, and the problem spread to every major non-hub airport during busy landing and takeoff cycles. The problem came to Austin, which is a non-federal control tower (a backwater of the ATC "system"). Astonishingly, the FAA reaction to the increased problem magnitude is to now propose that box carriers (tending toward all commercial traffic) should carry TCAS! The FAA does not have the technical understanding to deal effectively with the fundamental system design failure mechanism observed.
 
 
 
 

The FAA, then, tried to solve the problem by adding a "radar" post processing algorithm called "coast mode" to the targets displayed on the "radar" screen used by the controllers on the ground. This, they believed, helped to solve the sort box problem and the data link saturation problem, together. This adds the issue of what to do to "control" an aircraft when it is given verbal instruction to turn to avoid collision, but its track does not turn on the "radar" screen presentation. Those of us who fly are treated to entertaining return acknowledgments from the radio that sound something like: "(Airline-Name), (Airline-Flight-Number), uh, sir, uh, do you really want another right 90 for a course reversal?", said with some degree of urgency by the pilot, when all the aircraft are lined up for landing on a runway, and he is being turned about on a one way road. Coast mode assumes unaccelerated flight, until such time as the "radar" antenna again sees the transponder, the computer has time to process it, TCAS is not interfering with it, and ADS-B is not obscuring it. When the FAA "tested" ADS-B, they accomplished the "test" while concurrently grounding all aircraft in the area of the test. Propeller driven aircraft are only observed by the FAA to travel faster than the speed of sound on FAA "radar" displays - eventually, the "radar" target will "jump" to its real position, far away from where it was immediately previously coasting.

Every pilot with an instrument endorsement seems to agree - the collision avoidance system is extremely deficient. Personally, I have lost count of the number of near misses under 50 feet while both aircraft were operating under positive control.

The FAA believes that the "system" is so deficient that it must be totally replaced. We disagree - use of the VLSI to fix the ATCRBS transponder problem will suddenly cause transponders to become visible on "radar" systems. However, since change is in the air, why not do it right? The FAA current best favor "radar" replacement is called ADS-B, shown at AirVenture August 1999 (http://www.eaa.org), with data presented. I had the opportunity to sit in the cockpit and operate the ADS-B system. At the present time, according to II-Morrow engineers, the prototype equipment for an aircraft costs $300,000. At the present time, according to Mitre, the data link system capacity is such that the 1090 MHz data link starts to precipitously fail at about 300 ADS-B equipped aircraft within 80 NMi of eachother. That makes inoperable ATCRBS - all aircraft targets revert to coast mode. If that were not bad enough, there is a further terrorist exposure in ADS-B. Because all aircraft broadcast who and where they are and tactically where they are going, along with their complete and unique identity, one might be tempted to shoot at one that one doesn't like. Since the ADS-B data cannot fit in one packet, a multitude of packets are sprayed forth on 1090 MHz not in response to interrogation. That monumentally interferes with ATCRBS transmissions on the same frequency. The aircraft unique identity is transmitted with each individual packet to "tie the message together" for each aircraft position transmission. ADS-B broadcasts completely all of the information needed by the auto-pilot of a weapon system targeted against a specific aircraft, providing that target no means of escape. The homing system to drive the weapon auto-pilot is easily constructed utilizing readily available off the shelf components; with no need for a directional antenna, nor interrogating transmitter, nor large power supply. This leaves a large weight allowance for greater weapon range and speed and more powerful warhead. Airforce One and the Airforce and the Navy have already declined participation in the ADS-B "system" of the future, with corporate aviation expected to follow suit. That leaves ADS-B as incapable of providing collision avoidance between all aircraft, just as is present "radar".

One purpose of this paper is to provide proof that the "system" can be fixed to repair the FAA generated problems which result in "96% of transponders do not work right" (undisputed FAA official conclusion). That has been accomplished for ATCRBS, but the FAA has moved to block usage of the VLSI system-on-a-chip. That chip is also augmented to provide the additional ADS-B technology desired objectives, but we accomplished this augmentation without the ADS-B proven technology limitations. We also leave behind the ADS-B cost problem, and eliminate the need for the new ground equipment requirement to cooperate with ATCRBS (TIS) - saving hundreds of billions of dollars. This paper has introduced the availability of a custom VLSI for ATCRBS, that also includes a Eurofix (Loran-C with GNSS) or GNSS based AIS-P augmentation. The packet definition for the position-velocity AIS-P data transmission without need for aircraft identity is published for the intention of world adaptation. We request ICAO to issue one of the available downlink format numbers to this squitter packet definition, given the ICAO acceptance of the various ADS-B squitter packets into the mode S standard. We provide the ATCRBS VLSI for transponder automation, and the companion AIS-P receiver VLSI.

This augmentation to the transponder output provides all of the minimum collision avoidance requirements necessary for general aviation, and probably also commercial and military aviation. Without augmentation the VLSI repairs the problematic ATCRBS transponder reply to interrogations containing the P0 and P4 pulses. The equipage cost to restore being seen, and allow being seen better, and allow being seen anywhere, and allow all to be seen, has been reduced to about $500, the increase specifically because of the FAA blocking "certification" (pricing decision belongs to Bill Barton at Monarch-Air and Development, Inc.). The system capacity of AIS-P has been raised far beyond present ADS-B limits and also far beyond reasonable NAS capacity minimum requirements.
 
 
 
 

This paper has explained that we fix, by economical retrofit kit, all of the ATCRBS transponder problems identified by the FAA that cause aircraft to disappear from "radar". This paper then discussed that we add to this low-cost ATCRBS problem solution an augmentation to improve upon ATCRBS capability by providing a squitter of aircraft position and velocity data. This paper explained that a single chip VLSI has been developed to implement all of these system improvements, otherwise an extremely costly and sizable issue, cheaply and without redesign of the basic existing transponder. Finally, this paper here announces that there is available, to any world government having an "FAA" like regulatory facility, and to any interested avionics manufacturer, a demonstrator transponder. Available for loan for evaluation purposes, a productized transponder equipment totally compliant with TSO-C74c and all variants, including the additional unpublished ATCRBS requirements stated in the FAA report of transponder deficiencies. The AIS-P portion is compliant with the ICAO Mode S specification, utilizing a usurped downlink format number (11111). The product will remain compliant indefinitely without adjustment or attention, and is totally compatible with all current equipments world wide, both implemented and proposed. The product will be less costly to manufacture than the common transponder design of today, because of the level of integration.

Components removed from the Narco Avionics AT-150 transponder:

L403
C411 variable
C408
C412
C415
C416
C420
C421
C422
C423
C430
C432
C433
C435
C436
C437
C438
C439
C440
C441
C442
C445
C448
C449
C450
C451
R414 variable
R467 variable
R468 variable
R469 variable
R473 variable
R411
R418
R419
R435
R436
R437
R442
R459
R470
R471
R472
R474
R475
R476
R477
R478
R479
R480
R481
R482
R483
R484
R489
R490
R491
R492
R493
R494
R495
R496
R508
CR403
CR413
Q403
Q404
Q410
Q413
U402
U403
U404
U405
U406
U407
U408
U409
U411
U412
U413
U414
U415
U416
U417
U418
U419
U420

Actel 42MX24 PL84I IC
7414 hex schmitt inverter IC
clock oscillator IC
resistor
diode
capacitors (5)
 
 

Since a mode S downlink format number has yet to be assigned by ICAO (of the 32 there are some 6 available and unassigned), one was picked and "solder selected" at the VLSI pin input (changeable on the PCB).

With this modification to the common aircraft transponder, existing systems such as TCAS or TCAD or TLS or FAA "radar" will suddenly begin to work reliably. To avail themselves of "ADS-B objectives", these existing proximity detectors require only a simple software upgrade to receive the AIS-P packet. These systems can be dramatically cost reduced, because they can lose the 1030 MHz interrogation transmitter, and directional 1030 and/or 1090 MHz antennas, and the "fast processor", to display a complete "radar" picture in the cockpit. A VLSI to provide a low cost TCAS or TCAD or TLS or FAA "radar" receiver is in the works. Prototype will use a laptop PC running Delorme Street Atlas for "CDTI" traffic display (http://www.tapr.org/~kh2z/aprsplus/).

This system would be ideal for the introduction of ATC services into the third world countries, where cost is a consideration, and correct system operation is more necessary because a mechanism to command increased inter-plane delays does not exist. Systems such as the proposed Narco Avionics "Proximity Detector", for the general aviation aircraft, will cost mere thousands of dollars (about three), rather than hundreds of thousands of dollars (about three), and will be able to issue autonomous audio warnings, now received from ATC, though much more timely, and for the first time involving every aircraft, and for the first time anywhere in the world, and for the first time without ground support: "Traffic, three o'clock, four miles, two hundred high, twenty seconds (to impact if neither of you changes course)".

A technical solution to the P0 and P4 problems has been perfected and reduced to a custom VLSI. The FAA has reacted adversely, and refused to cooperate in the testing of the problem solution technology modification to existing aircraft transponders. The FAA has acted overtly to disallow the problem solution to be tested with the dedicated FAA means to determine ATCRBS transponder operation, used by the FAA to gather the evidence for its official report stating that 96% of transponders do not work completely correctly.

That technical solution has been augmented to allow, for the first time ever, aircraft collision avoidance 1. Between all aircraft, 2. Everywhere on the planet, 3. With no need for additional expensive ground equipment, 4. Without additional aircraft equipment cost, 5. Which is safe to use because it does not link each unique aircraft position and velocity to unique aircraft operator, 6. Which does not block ATCRBS from continuing to work. This is accomplished by transmission of 2D position and 3D velocity, received from GNSS or Loran-C sensor, and pressure altitude from the "blind pressure encoder" sensor. The FAA has not acted on this technology, thus effectively prohibiting the use of this technology in U.S. airspace.

This technology is not only safe and effective, not only non-interfering to any other existing or proposed technology, but legal under international treaty. All that is required to allow use of this technology is for ICAO to add a downlink format number assignment to the Mode S standard for the AIS-P packet data format. There are many downlink format number assignments available, and unused, and unrequested for use at the present time. The contact point at ICAO for this assignment is Jack Howell, Director Air Navigation Bureau, 999 University Street, Montreal, Quebec, Canada,H3C5H7 and he can be reached at 514-954-8176.

http://www.natca.org/natca/mediaandpublicrelations/pressreleases/engines.html
http://www.aeroelectric.com/articles/2_crises.pdf
http://www.gtwn.net/~keith.peshak/taillight.htm
http://mx.iki.rssi.ru/SFCSIC/english.html left frame to LINKS