The topic of the Lamarr-Antheil roll-driven torpedo controller crops up
from time to time, and its essential curiousness tends to swamp the
facts.  The Wikipedia article is a little light on the subject, and it
has never been properly described in MMD, so I thought readers might
like to read the following.  I scratched it together for the Player
Piano Group bulletin from what I could find at the time (the end of
1999, when Hedy Lamarr was still alive).

            The Lamarr-Antheil Patent

In recent years there has been a fair amount of publicity about a
"secret communications system" invented in 1941 by Hedy Lamarr, the
famous film star, and George Antheil, the infamous composer of the
pianola piece "Ballet Mechanique" and by that time a Hollywood com-
poser.

This used a pair of synchronized piano rolls to switch the frequency at
which a radio transmitter and receiver operated, the unpredictable
switches in frequency making it impossible to eavesdrop on or block the
signals being sent.  The patent was embargoed under wartime restric-
tions, and was not published until the mid-1980s.  The reason it has
come to prominence recently is that the frequency-hopping principle
which it described is identical to that used for modern day digital
phones, and if the patent had been published and protected, Lamarr and
Antheil would have become very wealthy indeed in recent years.

These may seem an unlikely pair to have developed such a system, but
Lamarr was experienced in military technology, having married armaments
manufacturer Fritz Mandl in her native Austria in 1933.  She had
listened to dinner-table conversations over the years, and been with
him when he reviewed field tests of torpedoes.  These frequently missed
their target, because the ship moved unexpectedly or because sea
currents caused the torpedo to drift.  It was easy to block control
signals sent to a torpedo on a single radio frequency.

Fleeing via London to America in 1936, largely out of dislike for her
husband, Lamarr was acutely aware of the rising Nazi threat.  She met
Antheil in 1940 at a dinner party hosted by Janet Gaynor, and outlined
her ideas for an anti-jamming control system for radio-controlled tor-
pedoes.  It seems reasonable to assume that Antheil was the one who
suggested the use of synchronized piano rolls, given his use of the
medium over the preceding twenty years; he was probably one of a hand-
ful of people in the world who could have developed the idea in that
manner.  He was also one of the few people in Hollywood intelligent
enough to interest her; she didn't suffer fools gladly and was famous
for her often-repeated quote that "Any girl can be glamorous.  All she
has to do is stand still and look stupid."

In 1940, the National Inventors' Council had been launched and widely
publicized.  Its intention was to attract ideas from members of the
public and assist their development.  Lamarr and Antheil sent their
ideas to the council in December 1940, and the patent was developed
with the assistance of the Massachusetts Institute of Technology.  The
patent was granted in August 1942.

At the time, the invention was announced in the New York Times thus:

    "Hedy Lamarr, screen actress, was revealed today in a new role,
    that of an inventor. So vital is her discovery to military defense
    that government officials will not allow publication of its de-
    tails.  Colonel L.B. Lent, chief engineer of the National Inven-
    tors' Council, classed Miss Lamarr's invention as in the 'red hot'
    category. The only inkling of what it might be was the announcement
    that it was related to remote control of apparatus employed in
    warfare."

The patent was never implemented by the War Department, and frequency
hopping was first used in 1962 in a system developed by Sylvania and
installed on ships blockading Cuba.  This was three years after the
Lamarr-Antheil patent expired and therefore fell into the public do-
main.  The Sylvania patent quotes Lamarr as the generic patent in this
field.  With the advent of digital transmission for cellular phones in
recent years, the idea has finally become commonplace.  The technique,
now called "spread-spectrum," permits effective and secure use of low-
powered transmissions.  Some fifty years after the idea, technology has
finally caught up.  Lamarr was awarded the Electronic Frontier Founda-
tion's Pioneer Award in 1997 as a belated acknowledgement for an inven-
tion she had never received any credit or payment for.

The latest episode in the saga that I know of is a report from Septem-
ber 1998, describing how a Canadian company called Wi-LAN has purchased
the original patent rights.  In return for a 49% claim to the patent,
Wi-LAN is promoting Lamarr's role as a wireless pioneer and has given
her some shares in the company ... her first payment, fifty-six years
late.  Wi-LAN's founder was a boyhood fan of Lamarr and was amazed when
working in wireless technology to find his idol had virtually invented
the technology.  He therefore decided to put together an agreement with
her.  If nothing else, it's a good advertising gimmick in the anonymous
but burgeoning world of hi-tech companies!

        The Patent Submission

(See the original images of U.S. Patent 2292387 at the USPTO or Google
Patents websites for the diagrams.)

The patent is claimed in terms of the specific application of control-
ling a torpedo.  This requires only two types of control to be sent:
turn left or turn right.  A ratchet device moves the rudder a step in
the required direction each time a control pulse is sent.

The patent Fig. 1 clearly shows the piano roll device, and how it
drives a series of seven switches, which place different capacitors in
an electrical circuit, which in turn cause the frequency of a radio
transmitter to change.  The control signal is then superimposed on this
carrier wave.  Fig. 2 shows the receiving device in the torpedo which
contains the same roll, and so changes the tuning of the receiver at
the same time as the transmitter.  The left-right rudder control is
achieved by the ratchet device.

The sending device as described contains seven channels, but the
receiver contains only four.  The other three channels are dummies,
which the operator can send signals on, simply to confuse any potential
eavesdropper.  Additional control channels operate lights on the con-
trol panel to tell the operator when the device is in dummy mode (or
about to enter it), so he doesn't send real controls at that time.  A
short "dummy" pulse is also coded at the start and end of actual
control frequencies to ensure that nothing is lost during the switch-
over period.  These can be seen as the top row of holes in Fig. 4.

The piano rolls in the two devices are to kept in synchronism by use of
calibrated clockwork drives, and started at the same time by means of a
locking pin which is released electrically at the moment the torpedo is
fired (Fig. 3).  The patent notes that it would be possible to send
synchronization pulses to periodically align the rolls (which clearly
harks back to the Ballet Mechanique of twenty years before).

Figs. 5 and 6 show the roll reader and switch-operating system, as well
as the locking pin device.  It can be seen that applied suction holds
the electrical contact device in the "off" position, compressing a
spring as it does so.  When air is admitted through the roll, it breaks
the suction and allows the spring to make the electrical contact.

The specific reason for using piano rolls in this device is stated as
being the entirely arbitrary nature of the pattern which may be coded
into it and the potential length of the roll, which is ample for the
time between firing and striking of a torpedo.  It is noted that the
number of frequencies may be expanded to use the full eighty-eight
available on a standard roll.

It is interesting to see this patent, with its clear ancestry back to
the early days of player devices, and compare it with its contempo-
raries--devices such as the German "Enigma" machine and its more
complex cousins--and the incredibly sophisticated means which were
adopted to crack the codes these produced, which led directly to the
first computers.

The essential arbitrariness of the roll, with its lack of an identi-
fiable pattern, would have actually made the Lamarr device much harder
to crack than the systematic mechanical coding devices, even if anyone
had been able to identify that the transmissions on different frequen-
cies were actually part of the same message.  The inclusion of spurious
components into the message was a nice touch.  A piano-roll controlled
torpedo perhaps wasn't the highest-tech encoding device of its time,
but the frequency-hopping idea it embodied has proved to be remarkably
powerful, a solution to a problem that took fifty years to arise.  It
is a fascinating story, and Lamarr's son has written a film script on
it.  So it may all come full circle yet.

Julian Dyer
Editor, Player Piano Group Bulletin
www.playerpianogroup.org.uk