Transmit Packet Definition & Format:

Management Summary:

There are 5 separate and distinct types of information we, as pilots, desire in the cockpit. In recent history the most important has been provided by Air Traffic Control (ATC), before that by the individual pilot looking out of the window.

For the vast overwhelming part of aviation, general aviation and sport aviation – 90% of all air traffic, this continues to be provided by looking out the window. The system of collision avoidance, called air traffic control, is both too expensive to allow participation in, and could not handle these numbers of traffic if they were forced to participate. Listening to the radio to have some guy tell you how and when to turn to avoid collision with another airplane is not an option.

In the remaining 10%, they are extremely busy in the cockpit of that big airplane, and they have very little windows, and the visibility is very restricted. Looking out the window is not an option.

The problem with ATC is that it does not inform the 10% about the existence of the 90%, because it cannot see them with the 1200 filter turned on and primary radar turned off. ATC also does not like to talk to littler airplanes, because they would become overwhelmed.

This is unfortunate for the 10% when a collision occurs. This is unfortunate for the 90% when a collision occurs. There is no collision avoidance system between all aircraft! Never has been. The collision avoidance system between the subset of aircraft that are allowed to participate in the collision avoidance system cannot handle all of that subset of aircraft available in places like the Los Angeles Basin, and it has been that way since the 1960s. Lost AirForce 1 several times, so far this year? Because, they said, the "physics of radar"? Now you have a brief definition of what that is!

This needs to get fixed! Here is how:

Objectives:

There are 5 separate and distinct types of information we, as pilots, desire in the cockpit:

1. Knowledge of where we are, displayed in such a way that we can figure it out. Currently served by charts and navaids, including VOR, DME, ADF, Loran-C, GPS, and still by looking out the window. Commerce Secretary Ron Brown could testify as to how well this can work, but he isn’t around anymore.
2. Knowledge of where significant permanent objects are. Permanent objects, such as terrain and P56 over the White House, don't move, so this is served quite well by paper and electronic charts. Coupled to an electronic pilot interface display would be nice. That technology is available, finally, but the FAA will fine you if they catch it in your 90% airplane (yoke mount for a hand held GPS can result in a "ticket" from the FSDO).
3. Knowledge of where are significant temporary, but slow moving objects. (Slow moving as in hours, days or months, not as in fixed wing aircraft.) These include such nasties as thunderstorms, ice, new tower construction, and temporary airspace restrictions. These are currently served by FSS weather briefers and NOTAMS. While this system is sorely inadequate, it is currently being addressed by various alphabet groups. Once a communications format is found, and a data link provided, this could be viable.
4. Air traffic management. Not of general interest to this group. Let the airlines decide what is needed for scheduled air route operations over highways in the stratosphere, and the issues of streaming traffic into and out of a separated and sequestered air terminal airport we can’t use anyway. A 10% problem, not our problem, why should we combine it with our common interests?
5. Knowledge of where significant fast moving objects are. Airplanes, skydivers, balloons, spacecraft, towers (they move relative to us), and their ilk. All collision avoidance. This one topic is the area that all technical conference participants felt was the most sorely neglected, and therefore most in the need of technical assistance. Attempts to date by the FAA to deal with this problem are inadequate, cumbersome, and generally unworkable. Some felt that the FAA's recent efforts are actually a significant step backwards.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

It was agreed upon that the best way to facilitate collision avoidance was to dedicate a single radio frequency to that one task, and that task alone.

It was agreed upon that a standard protocol definition was now required to determine the issues of:

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Each airplane would periodically broadcast its position, and may choose to listen to other broadcasts. Various schemes were proposed for managing these broadcasts in order to best utilize the spectrum.

1. Use a frequency manager. Proposed by Mark Kreger.

Said manager could be a series of ground stations, or a complex scheme of airborne managers. While this could be made to work, with spectrum efficiencies approaching 50%, all units would be required to receive as well as transmit, and the complexity involved is feared to raise the cost of each unit significantly. It is noteworthy that this was the initial state of the ADS-B proposal, wherein the frequency manager was the ATCRBS ground radar unit, addressing single messages to each aircraft individually, thereby clogging up the radar band for effective use as a radar. Someone would have to pay for the ground stations, too. It was agreed that one very important parameter be an absolute minimum initial cost of entry into this system by each user, specifically not requiring the ability to receive or process data in order to let the other participants know of his position relative them each. The Mode S transponder, at ~$10,000, to replace the Mode 3A/3C transponder, at ~$700, failed that requirement. If a TailLight transmitter is cheap and power thrifty enough, it might be feasible to mount one on sky-divers and balloons.

2. Some sort of time slot allocation. Proposed by Augie Beining in a Sport Aviation article prior to this meeting, and argued vociferously by Augie at the meeting.

Despite claims of 100% efficiency, Phil Hodge demonstrated that theoretical spectrum density could not exceed 25% due to the time delays required due to random distances between aircraft and perceived but unnecessary re-transmissions.

3. Random. Lowest level of complication.

Various studies have been performed in numerous applications that indicate spectrum utilization on the order of 20 to 25% is reasonable. Remembering the KISS principle (Keep It Simple, Stupid), it was agreed that a random protocol was the best.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Much discussion was held concerning what should be broadcast. The intent was to keep the data stream as short as possible, while maintaining maximum utility, and possible expansion in the future should the need and technology arise.

There was also considerable discussion prior to the meeting concerning repeating lateral position versus absolute over the entire surface of the planet. It was agreed to broadcast absolute position.

It was universally agreed upon, with little discussion, that the most important data was current position, second was current vector, and lastly various and sundry miscellaneous bits of information detailed below. The data stream evolved into:

Position:

Velocity:

Miscellaneous:

Message type: 1 bit

Overhead:

Total: 117 bits.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

ADS-B, as currently defined, requires every plane to broadcast every 1/2 second. We agreed that such a rate seemed excessive, and unnecessarily restricted system capacity. The tentative conclusion was for everyone to broadcast whenever one or more of the following was met. All of these parameters are to have +/-10% intentional time dither to reduce the chance of two aircraft getting into unintentional synchronism with each other.

There was some discussion of whether the balloon - jet scenario is adequately handled with this reporting frequency, with no resolution. A stationary balloon will report every 10 seconds, even to the jet approaching at 600 knots. As far as the jet is concerned, he's sitting in the middle of his bubble, with a balloon approaching at 600 knots! In the 10 seconds it takes for the balloon to report, they have closed 1.67 NM. Depending on how far away they are when the jet receives the first clear balloon report, is there enough time for the jet to recognize the hazard, react to it, and for the plane to maneuver?

This same scenario is currently not addressed by ATC, because the balloon does not have a transponder, and remains unseen. However, it was not acceptable to the participants to come up with a solution that was simply better than ATC.

Further simulation and parametric studies are needed to demonstrate the desired density would be safe. See article, above, for that more current simulation result efforts, and comparison to ADS-B capabilities.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

We all agreed that a system capable of handling 2000 aircraft within reception range should be adequate. Assuming a random protocol with no carrier detect, maximum 20% spectrum density, an average of 1 second between broadcasts (very conservative), and the previous 117 bits/message, that works out to about 1 MB. A more realistic mix of this much traffic might look like:

# of aircraft Speed (kts) Report period (seconds)

Total 2000 aircraft, average 690 transmissions/second = 8% spectrum density.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Operational Summary:

The questions asked near the beginning can now be answered:

Who talks to whom?

How often? Variable

Over what range?

Saying what?

With what assurance?

How many participants?

Demonstrator Architectural Overview

Bob Bruninga’s APRS (TM)

TailLight

Get a frequency assigned

Portable PC

Tekmos