John Hayhurst, Boeing Company 2/5/01
Norman Mineta, Department of Transportation
Jane Garvey, FAA
Mr. K. Koplin, JAA

I was recently excited to hear the Boeing Company announce an intent to invest in an Air Traffic Control system based on the Global Positioning System. Such a system is the obvious solution to our dire ATC problems.

Seven years ago, I wrote a paper proposing an extremely simple digital radio link intended for a GPS based ATC and Collision Avoidance system.

Since then, I have been concerned with the apparent direction the FAA is taking in this area. The FAA seems focused on a narrow band (9600 bps) system already implemented by the Scandinavians.  The NEAN (Northern European ADS Broadcast Network) approach uses a 25 KHz wide voice communication channel (135.975 MHz). The FAA approach is to share the 1090 MHz surveillance radar transponding channel with the same, low, 9600 bps, data rate.  I am sure that this rate is insufficient  for today’s metro traffic areas, let alone  future loads.

I can only guess that the Europeans and our own FAA feel that they must live within the allocated spectrum bandwidth they already have.

I believe that the future of the World’s Air Traffic Control is important enough to warrant allocating the proper bandwidth.

My proposal would require reallocating one of the 46 UHF television channels presently under utilized in the USA.  These channels fall between 500 to 800 MHz in the USA and Europe. This would provide a full 6 MHz of bandwidth for a system that would handle the most ambitious traffic projections for the next decades.

The NEAN and FAA approaches both use STDMA (Self-organising Time Division Multiple Access).  This is a network protocol that assigns unique time slots to prevent nodes from talking simultaneously.  It is not a simple protocol.  Time division multiplexing is made more complex because the propagation times between nodes is constantly changing (Fast aircraft moving in different directions over large distances).   The system  can take several minutes to organize and the update rate is once per minute.

I believe a collision avoidance system should come to life immediately at turn on and have an update rate of once per second.

My approach does not attempt prevent coincidences (simultaneous talkers).  It avoids coordination, self organizing, and arbitration all together, resulting in enormous simplification.

Each aircraft transmits 225 microsecond data packets, roughly once per second.  Since, all transmissions are in a single channel, occasionally, packets will arrive at a receiver overlapping in time.  Ordinarily, this interference would cause a loss of all the data.  If simple Frequency Modulation is used, the marvelous mechanism of capture applies, and only the data from the further aircraft will be lost.  The stronger data packet, from the closest aircraft, is received accurately.  This is as it should be.  It is this important point that makes FM so attractive.

My brief paper can be found in PDF and TEXT form at:
http://www.sonic.net/~ritk/fmdl.pdf
http://www.sonic.net/~ritk/fmdl.txt (no figures)

It provides some simple explanations of the mechanism of capture.
It presents statistical examples involving 1000 aircraft in a 50 mile radius.

I hope that you will pass this letter on to any who are involved in the future of Air Traffic Control.
 

R C Keiter
ritk@sonic.net
1 707 542 6053
fax 1 707 542 1281
2655 N Fitch Mtn Rd
Healdsburg CA 95448

 Home