‘Project Achilles’ – Final (Third) Report and Summary of Findings

A. K. Dewdney, 19th April 2003

During the early months of the year 2003, the author conducted three experiments to determine whether and how well cellphones could be operated from aircraft
. The first flight (Part One) was essentially a probe of the experimental situation, to acquire some primary data and to work out a simple, readily implemented protocol. The results of Part Two (Diamond Katana 4-seater) have already appeared in these pages. The results of Part Three (Cessna 172-R) appear immediately below.

Since this completes the suite of experiments, it is appropriate to summarize the findings and to draw some conclusions. The conclusions are based partly on the experiments and partly on two other sources. (See Appendix B at the end of the report.) Expert opinion and eyewitness testimony are acceptable not only in court, but in certain scientific inquiries where events are of short duration or experiments are too expensive or impossible to carry out. Of course, eyewitness accounts do not carry the same weight as expert opinions or actual experiments, but the eyewitness accounts quoted below seem to be both consistent and compelling.

Disclaimer: The companies hired to assist in this experiment, namely Empire Aviation and Cellular Solutions, both of London, Ontario, Canada, acted as disinterested commercial parties, with no stake in the outcome or even knowledge of the purpose of the tests.

Part Three – April 19th 2003

The previous experiment, called Part Two, established a distinct trend of decreasing cellphone functionality with altitude. It was conducted in a four-seater Diamond Katana over the city of London (pop. 300,000), Ontario in Canada, an area richly supplied with some 35 cellsites distributed over an area of about 25 square miles. The flight path was an upward spiral, punctuated every 2000 feet (abga) with a level circuit around the outskirts of the city. On each circuit a fixed number of cellphone calls were attempted by an expert operator employing a battery of well-charged phones broadly representative of those on the market both currently and in the year 2001.

(It should be remarked that not only is the cellphone technological base in Canada identical to its US counterpart, but Canadian communication technology is second to none, Canada being a world-leader in research and development.)

The purpose of Part Three was to test the effects of what might be called “Faraday attenuation” on the strength and success of calls. The presence of a metallic shell around some electronic devices can alter their behavior by its ability to attract and store electrons, especially electromagnetic waves. For this reason, the experimental craft was switched from the Katana, which is supposed to be relatively transparent to em radiation, to an aircraft with an aluminum skin, as below.

Equipment:

• Cessna 172-R (2002) four-seater (Empire Aviation)
• cellphones: C1, C2, C3, C4, C5 (See Appendix A for descriptions.)

Personnel:

• Corey Barrington (pilot – Empire Aviation)
• Darren Spicknell (operator – technician for Wireless Concepts, Inc)
• Kee Dewdney (director)
• Pat Dewdney (ground recorder)

Weather: unlimited ceiling, light scattered cloud at 5,000, solid/broken 24,000 feet, visibility 12 miles, wind 11 knots from SSW, air temperature +19 C.

For this experiment, we flew the same circular route as we did in Part Two, The circle centered on the downtown core and took us over most of the city suburbs. All locations below are referred to the city centre and are always about two miles distant from it.

Protocol:

At times specified by the director, the operator made a call to a specified number, stating the code number of the cellphone (1 to 5) and the altitude. The ground recorder noted whatever was heard and the time the call was received. At the first two altitudes of 2000, 4000 above ground altitude (abga) each cellphone was used once. At 6000 and 8000 feet abga, each cellphone was used twice only C2, C3, and C5 were tried, C1 and C4 being hors de combat.

Results with timeline:

time (pm)	call no.	C#	loc.	operator recorder
7:05 – started taxi to runway 7:12 – takeoff 7:15 – at 2000 feet (aboveground altitude)
7:17	Call #1	C1	N	success clear, slight breakup
7:18	Call #2	C2	W	success clear
7:20	Call #3	C3	SW	success clear
7:22	Call #4	C4	S	success (2 tries) clear
7:23	Call #5	C5	SE	success clear
7:27 – climbed to 4000 feet abga
7:28	Call #6	C1	NE	success clear
7:30	Call #7	C2	N	success clear
7:31	Call #8	C3	NW	“success” (frag) no complete word
7:32	Call #9	C4	W	failure no ring
7:34	Call #10	C5	SW	success clear
7:35 – climbed to 6000 feet abga
7:39	Call #11	C1	SE	success clear
7:41	Call #12	C2	E	success clear
7:42	Call #13	C3	E	success clear, slight breakup
7:44	Call #14	C4	NE	failure no ring
7:44	Call #15	C5	NE	failure no ring
7:45	Call #16	C1	N	failure no ring
7:46	Call #17	C2	N	success clear
7:47	Call #18	C3	NW	failure no ring
7:48	Call #19	C4	NW	failure no ring
7:49	Call # 20	C5	W	success clear
7:50	Call #21	C1	W	failure no ring
7:51	Call #22	C2	SW	failure no ring
7:52	Call #23	C3	SW	failure no ring
7:53	Call #24	C4	S	failure no ring
7:54	Call #25	C5	S	success clear
7:55 – begin climb to 8000 feet abga (cellphones C2, C3 and C5)
7:55	Call #26	C2	SE	failure no ring
7:57	Call #27	C3	E	failure no ring
7:59	Call #28	C5	E	success clear, slight breakup
8:00 – completed climb to 8000 feet abga
8:01	Call #29	C2	NE	failure no ring
8:02	Call #30	C3	NE	failure no ring
8:03	Call #31	C5	N	failure no ring
8:04	Call #32	C2	NW	success clear
8:05	Call #33	C3	NW	failure no ring
8:07	Call #34	C5	W	failure no ring
8:20 – landed at airport

The following table summarizes the results:

altitude (feet)	calls tried	calls successful	percent success
2000	5	5	100%
4000	5	3	60%
6000	15	6	40%
8000	15	2	13%

Note: calls “tried” includes retired cellphones C1 and C4 above the altitude of 4000 feet where, in the opinion of the cellphone expert, they would have failed to get through, in any case. Failure to include them in
the count would make the results at different altitudes non-comparable.

The results of this experiment may be compared to the results from Part Two where, instead of the Cessna, we used the Diamond Katana:

altitude (feet)	calls tried	calls successful	percent success
2000	4	3	75%
4000	4	1	25%
6000	12	2	17%
8000	20	1	5%

To make the results comparable, however, cellphone C5 was omitted from the calculations, since it was not used in the first experiment.

altitude (feet)	calls tried	calls successful	percent success
2000	4	3	75%
4000	4	1	25%
6000	12	2	17%
8000	12	1	8%

Analysis

Since the (1.5 mm) skin of the Cessna appears to have made little difference to the outcome of the experiment, the data of Parts Two and Three may be combined, as follows, to produce more reliable figures for the battery of test phones that were used in the experiment:

altitude (feet)	calls tried	calls successful	percent success
2000	9	8	89%
4000	9	4	44%
6000	27	8	30%
8000	35	3	9%

The data from the first three altitudes appear to fit an inverse-linear model of attenuation. In other words, the probability of a call getting through varies inversely as the altitude, according to the formula:

Probability of success = k/altitude, where k is a constant

It will be noted that the values of k implied by these data, at least up to 6000 feet abga are remarkably consistent. However, at 8000 feet the k-value falls precipitously, implying that a different regime may be in play.

altitude (feet)	k-value
2000	1780
4000	1760
6000	1800
8000	720

The expected model of attenuation with distance is of course inverse squared, a natural consequence of the three dimensions that any uniform radiation must travel through. Inverse squared attenuation follows a slightly different pattern or formula:

Probability of success = k/altitude²

To estimate k, it seems reasonable to use the data from 4000 feet and 8000 feet as benchmarks for the calculation of the constant k (not the same constant as was used in the foregoing analysis, of course.)

At 4000 feet abga the implied k-value if 7,040,000, while at 8000 feet, the implied k-value is 5,760,000. although here again the k-value appears to drop (indicating that the actual attenuation may be worse than inverse squared), we use an average of the two estimates, following our consistent practice of always giving the benefit of the doubt to the cellphones, so to speak.

Taking an average value of k = 6,400,000, we obtain the formula,

Probability of success = 6,400,000/altitude²

Using this formula, we can get a best-case estimate for the probability of cellphone success from a slow-moving light aircraft, as summarized in the following table.

altitude (feet)	probability of cellphone call getting through
4,000	0.400
8,000	0.100
12,000	0.040
16,000	0.025
20,000	0.016
24,000	0.011
28,000	0.008
32,000	0.006

Private pilots flying light aircraft are nowadays familiar with the fact that they may use their cellphones to make calls to the ground, at least if they are not higher than one or two thousand feet. Above that altitude, calls get rather iffy, sometimes working, sometimes not. The higher a pilot ascends, the less likely the call is to get through. At 8000 feet the pilot will not get through at all unless he or she happens to be using a cellphone with the same capabilities as C5 (See appendix A.) But even that cellphone begins to fail at 6000 feet.

Calls from 20,000 feet have barely a one-in-a-hundred chance of succeeding.

The results just arrived at apply only to light aircraft and are definitely optimal in the sense that cellphone calls from large, heavy-skinned, fast-moving jetliners are apt to be considerably worse.

Conclusions

It cannot be said that the Faraday attenuation experiment (Part Three) was complete, in the sense that the operator normally held the phone to his ear, seated in a normal position. This meant that the signals from the test phones were only partially attenuated because the operator was surrounded by windows that are themselves radio-transparent.

Although we cannot say yet to what degree the heavier aluminum skin on a Boeing 700-series aircraft would affect cellphone calls made from within the aircraft, they would not be without some effect as windows take up a much smaller solid angle at the cellphone antenna. Signals have a much smaller window area to escape through, in general.

As was shown above, the chance of a typical cellphone call from cruising altitude making it to ground and engaging a cellsite there is less than one in a hundred. To calculate the probability that two such calls will succeed involves elementary probability theory. The resultant probability is the product of the two probabilities, taken separately. In other words, the probability that two callers will succeed is less than one in ten thousand. In the case of a hundred such calls, even if a large majority fail, the chance of, say 13 calls getting through can only be described as infinitesimal. In operational terms, this means “impossible.”

At lower altitudes the probability of connection changes from impossible to varying degrees of “unlikely.” But here, a different phenomenon asserts itself, a phenomenon that cannot be tested in a propellor-driven light aircraft. At 500 miles per hour, a low-flying aircraft passes over each cell in a very short time. For example if a cell (area serviced by a given cellsite) were a mile in diameter, the aircraft would be in it for one to eight seconds. Before a cellphone call can go through, the device must complete an electronic “handshake” with the cellsite servicing the call. This handshake can hardly be completed in eight seconds. When the aircraft comes into the next cell, the call must be “handed off” to the new cellsite. This process also absorbs seconds of time. Together, the two requirements for a successful and continuous call would appear to absorb too much time for a speaking connection to be established. Sooner or later, the call is “dropped.”

This assessment is borne out by both earwitness testimony and by expert opinion, as found in Appendix B, below. Taking the consistency of theoretical prediction and expert opinion at face value, it seems fair to conclude that cellphone calls (at any altitude) from fast-flying aircraft are no more likely to get through than cellphone calls from high-flying slow aircraft.

A. K. Dewdney,
April 19th 2003

The author has not placed his university affiliations below his name, as the research described here was not conducted with any university facilities or supported by university-administered grants. He currently holds the titles of Professor Emeritus of Computer Science and Adjunct Professor of Biology at the University of Western Ontario, as well as Professor of Computer Science at the University of Waterloo.

---

APPENDIX A: Cellphone types

• C1 – Motorola i95cl – Telus Mike Network – 800 Mhz IDEN
• C2 – Motorola StarTac – Bell Mobility – 800 Mhz Analog
• C3 – Audiovox 8300 – Telus PCS Network – 1.9 Ghz CDMA / 800 MHz
• C4 – Nokia 6310i – Rogers AT&T – 1.9 Ghz GHz GSM. (Tri-Band – Has an 1.8 GHz and 900 Mhz GSM these are European frequencies)
• C5 – Motorola Timeport 8767 – Bell Mobility – 800 MHz Analog (CDMA Tri-Mode 1.9 GHz CDMA / 800 Mhz CDMA)

APPENDIX B: Letters

Professional opinions

==========================================================

Dear Sir

I have yet to read the entire [Ghost Riders] article but I do have a background in telecommunications. Using a cell phone on an air craft is next to impossible. The reasons are very detailed, but basically the air craft would run major interference, as well as the towers that carry the signal would have a difficult time sending and receiving due to the speed of the air craft. As well, calling an operator? Well that is basically impossible.

Having worked for both a major Canadian and American provider I had to instruct my staff that operator assistance is not an option. Have you ever tried to use a cell phone in some public buildings? Impossible. There are too many spots that service is voided. Just a tidbit of information to share.

Megan Conley <megan_conley@hotmail.com>

———————————————————————————–

Hi,

I am an RF design engineer, having built out Sprint, Verizon and another network in New Orleans. You are absolutely correct. We have trouble making these things work for cars going 55 mph on the ground. If you need another engineer’s testimony for any reason, let me know I will corroborate.

my engineering site: http://www.geocities.com/rf_man_cdma/

Brad Mayeux <cdmaman@engineer.com>

———————————————————————————-

Anecdotal evidence

==========================================================

Sir,

Yours is the first article I’ve read which focuses on those dubious ‘cell phone calls’. Last month my Wife and I flew to Melbourne, about 1000 miles south of here.

Cell phones are Verboten in Airliners here, but on the return journey I had a new NOKIA phone, purchased in Melbourne, and so small I almost forgot it was in my pocket. I furtively turned it on. No reception anywhere, not even over Towns or approaching Brisbane. Maybe it’s different in the US, but I doubt it.

There has to be an investigation into this crime. Justice for the thousands of dead and their families demands it.

Best

Bernie Busch <bbusch@iprimus.com.au>

———————————————————————————–

Hi Prof

I have repeatedly tried to get my cell phone to work in an airplane above 2-3000 feet and it doesn’t work. My experiments were done discreetely on [more than] 20 Southwest Airlines flights between Ontario, California and Phoenix, Arizona. My experiments match yours. Using sprint phones 3500 and 6000 models, no calls above 2500 ft [succeeded], a “no service” indicator at 5000 ft (guestimate).

There seem to be two reasons. 1. the cell sites don’t have enough power to reach much more than a mile, 2. The cell phone system is not able to handoff calls when the plane is going at more than 400 mph.

This is simply experimental data. If any of your contacts can verify it by finding the height of the Pennsylvania plane and it’s speed one can prove that the whole phone call story is forged.

Rafe <rafeh@rdlabs.com> (airline pilot)

—————————————————————————————–

Greetings,

I write in praise of your report, as I have felt from day one that the cell phone ‘evidence’ was perhaps the flimsiest part of the story, and am amazed that nobody has touched it until now.

I’d also like to bring up the point of airspeed, which is what made the cell calls a red-flag for me in the first place. I’m not sure what your top speed achieved in the small plane was, but, in a large airliner travelling at (one would think) no less than 450mph, most cell phones wouldn’t be able to transit cells fast enough to maintain a connection (at least, from what i understand of the technology) .. and we’re talking 2001 cell technology besides, which in that period, was known to drop calls made from cars travelling above 70mph on the freeway (again, due to cell coverage transits)

Anyway, thanks for shining the light, keep up the good work

Ben Adam

——————————————————————————————-

Dear Professor,

Responding to your article, I’m glad somebody with authority has taken the trouble to scientifically prove the nonsense of 9/11.

I was travelling between two major European cities, every weekend, when the events in the US occurred. I was specifically puzzled by the reports that numerous passengers on board the hijacked planes had long conversations with ground phone lines, using their mobile phones (and not on board satelite phones). Since I travelled every weekend, I ignored the on board safety regulations to switch off the mobile phone and out of pure curiosity left it on to see if I could make a call happen.

First of all, at take off, the connection disappears quite quickly (ascending speed, lateral reception of ground stations etc.), I would estimate from 500 meters [1500 feet approx.] and above, the connection breaks.

Secondly, when making the approach for landing, the descent is more gradual and the plane is travelling longer in the reach of cellphone stations, but also only below 500 meters. What I noticed was that, since the plane is travelling with high speed, the connection jumps from one cellphone station to another, never actually giving you a chance to make a phone call. (I have never experienced this behaviour over land, e.g. by car). Then, if a connection is established, it takes at least 10-30 seconds before the provider authorises a phone call in the first place. Within this time, the next cellstation is reached (travel speed still > 300KM/h) and the phone , always searching for the best connection, disconnects the current connection and tries to connect to a new station.

I have done this experiment for over 18 months, ruling out weather conditions, location or coincidence. In all this time the behaviour was the same: making a phone call in a plane is unrealistic and virtually impossible.

Based on this, I can support you in your findings that the official (perhaps fabricated) stories can be categorised as nonsense.

With kind regards.

Peter Kes <kpkes@yahoo.com>

---

It must be clearly understood that Prof. Dewdney’s tests were conducted in
slow-moving (<150kts) light aircraft at relatively low altitudes (<9000ft
AGL). The aircraft from which the alleged calls were made on 9/11 were
flying at over 30,000 ft at speeds of over 500 MPH.

During a recent round-trip flight from Orange County, CA to Miami, FL (via
Phoenix, AZ), I, personally conducted an unofficial “test” using a brand new
Nokia 6101 cellular phone [NB: 2005 technology]. En route, I attempted
(discretely, of course) a total of 37 calls from varying altitudes/speeds. I
flew aboard three types of aircraft: Boeing 757, 737, and Airbus 320. Our
cruising altitudes ranged from 31-33,000ft, and our cruising speeds, from
509-521 MPH (verified post-flight by the captains). My tests began
immediately following take
-off. Since there was obviously no point in taking
along the wrist altimeter I use for ultralight flying for reference in a
pressurized cabin, I could only estimate (from experience) the various
altitudes at which I made my attempts.

Of the 37 calls attempted, I managed to make only 4 connections – and every
one of these was made on final approach, less than 2 minutes before flare,
I.e., at less than 2,000ft AGL.

Approach speeds varied from 130-160 kts (Vref, outer marker), with flap and
gear extension at around 2,000ft (again, all speeds verified by flightdeck
crews). Further, I personally spoke briefly with the captains of all four
flights: I discovered that in their entire flying careers, NOT ONE of these
men had EVER been successful in making a cell phone call from cruising
altitude/speed in a variety of aircraft types. [NB: Rest assured the
ubiquitous warnings to “turn off all electronics during flight” are
completely unfounded. All modern aircraft systems are fully shielded from
all forms of RF/EMF interference (save EMP, of course). This requirement was
mandated by the FAA many years ago purely as a precautionary measure while
emerging advanced avionics systems were being flight tested. There is not a
single recorded incident of interference adversely affecting the performance
of airborne avionics systems.]

Obviously, my casual, seat-of-the-pants attempt at verifying a commonly
known fact can hardly be passed off as a “scientific” test. Ergo, I shall
offer Prof. Dewdney¹s conclusion, excerpted from his meticulously detailed
and documented paper re slow-flying light aircraft at low altitudes.

Nila Sagadevan

---

Prof. Dewdney:

I do not pretend to be any sort of expert of cellular communications, but I am an electronics engineer and hold both amateur and commercial FCC licenses, so I do have some understanding of the relevant principles of radio communication systems.

I read with interest your analysis of terrestrial contact probabilities via cellphones from aircraft. I believe your conclusions are sound, but would like to comment on an element which you pondered regarding the sort of apparent discontinuity in what seems otherwise to be an inverse-square relation beyond a certain altitude.

Cellphones operate by Frequency Modulation, and as such the (apparent) signal strength is not discernible to the listener because the intelligence is contained only in the frequency and phase information of the signal before demodulation. Hence, the system works pretty well until it is so weak that it is abruptly lost. That is, the system can no longer “capture” the signal. It does not get louder and softer with signal strength -until the signal is below the detection level of the receiver, at which point it is essentially disappears. The cellphone also adjusts the transmit power according to the signal level received at the tower end of the link. Once it is at maximum output, if the signal diminishes beyond some minimum threshold depending on the receiver design, it is lost altogether and not simply degraded in quality. Analogous behavior is experienced with FM broadcast stations; as you travel away from the transmitter the station is received with good fidelity until at some distance it rather suddenly cannot even be received any longer at all.

Additionally, cellphone towers are certainly not optimally designed for skyward radiation patterns. Since almost all subscribers are terrestrial that is where the energy is directed, at low angles.

In summary, if your observed discontinuous behavior is real, and I believe there is technical reasoning for such, the probability of making calls beyond some threshold altitude is not simply predictably less, but truly impossible with conventional cellphones under any condition of aircraft etc. because of the theoretical limits of noise floor in the receiving systems. I think the plausibility of completing the calls from 30,000+ ft. is even much lower than might be expected from extrapolations of behavior at lower altitudes which you investigated.

Thank you for your thoughtful work in this area.

Sincerely,

Kevin L. Barton

---

Dear Dr Dewdney

I too can verify that on a private charter airline, Champion Air, which was a 737-300, I believe that is correct or it might have been a 727-300. But regardless of that, we took off from Dallas/Ft. Worth International Airport at 0735 in July of 2003. As we were taxiing to the run way the pilot told us to please turn off all electronic equipment, i.e. Cell Phones, Laptops, etc. I did so, but shortly after take off and before the pilot said we could use our “electronic equipment” I thought I would call my mom and let her know we were in the air. We had not been off the ground for more than 2 minutes. I would guess between 2000 and 5000 ft. I was using at the time one of Motorola’s top of the line phones, a V60t. My cell phone carrier is Cingular which is quite a widespread carrier as you probably know, I had absolutely no signal at all. Since we were flying to Cozumel, Mexico I kept trying and watching for a signal until we got out past the coast line of Texas, when then I knew for sure I wouldn’t get a signal again until we landed in Cozumel. Again in June 2004 we flew out of DFW, same airline, same type of plane, and the same thing occurred. This time I left my phone on from take off and up until it lost the signal. Again we couldn’t have been more than 2000 to 3000 ft. off the ground. I lost the signal and never again got a signal until the plane landed in Cozumel. I find it highly unlikely that anyone could have used a cell phone on 9/11/01 at above 2000 feet.

Sincerely,
Brad Taylor

---

I’ve been using Nokia phones with automatic nationwide roaming, and Cingular before it was Cingular and long before 9/11. I confess to having turned my cell phone on while flying commercial airlines several times prior to 9/11, just to see if signals were available. At 2,000 feet the phone went totally flat. No calls out or in were ever possible. Of course all these stories are anecdotal, but according to cell phone engineers who have cared to comment have stated that commercial aircraft fly far too fast and far too high to expect that folks on flight 93 ever managed to get a call out on their own phones. They’ve said that the towers can’t transition or hand over private cell phones fast enough. I hope we can hear from other ATPs on this subject.

George Nelson (Col. USAF ret.)