 |
Claims  |
|
|
What is claimed is:
1. A method for the enciphered transmission of messages by splitting up the
clear signals to be transmitted into elements of equal length T.sub.o,
which are transposed at the transmitting end by being delayed by at least
partially unequal times and are re-transposed at the receiving end by
being further delayed by at least partially unequal times, said method
comprising the steps of transposing a pair of elements at a time, which
elements have a specific mutual spacing, at the transmitting end, and
re-exchange of the same elements in pairs at the receiving end, the pairs
of elements which are transposed at the transmitting end and re-transposed
at the receiving end being determined by providing irregular trains of
control pulses which are identical at the transmitting and receiving ends,
and delaying those elements which do not belong to the pairs of elements
by a fixed time T at the transmitting and receiving ends, and delaying the
element of each pair arriving first at the transmitting end and at the
receiving end by double the time 2T and passing the second element of the
pair without delay.
2. A method as claimed in claim 1, wherein the delay time T is selected to
coincide with the element length T.sub.o (FIG. 11).
3. A method as claimed in claim 1, wherein the delay time T is selected to
coincide with an integral multiple of the element length T.sub.o (FIG.
13).
4. A method as claimed in claim 1, characterized by repeated carrying out
of the exchange of pairs of elements at the transmitting end and of the
re-exchange at the receiving end, wherein the first step of performing the
exchange of elements at the transmitting end is carried out in accordance
with control pulses developed thereat and the step of performing the last
exchange of elements at the receiving end is determined by transmitting
said control pulses to the receiving end for controlling said last
exchange.
5. A method as claimed in claim 4, wherein equal time delay lengths are
employed at the transmitter and receiver ends in the repeated exchanges of
elements in pairs so that the element of a pair which has not been delayed
at the transmitter end is subjected to a delay at the receiver end which
is equal to the delay length imposed upon a delayed element at the
transmitter end.
6. A method as claimed in claim 4, wherein the exchanges of element pairs
at the transmitter end comprises the employment of unequal time delay
lengths during each such repetition to further increase the mixing of
elements of the message.
7. A method as claimed in claim 4, wherein repetition of the exchanges at
the transmitter end is performed with a varied delay time T (FIG. 16)
employed during each repetition.
8. A method as claimed in claim 7 wherein a ratio of 1:3 for the delay
times of the exchanges is employed.
9. A method as claimed in claim 1, further comprising the step of combining
the exchanges of elements in pairs with an additional permutation of the
elements in accordance with a predetermined program (FIG. 23).
10. A method as claimed in claim 1, further comprising the step of
combining the exchange of elements in pairs with an additional known time
coding with a varying program for the element transposition.
11. A method as claimed in claim 1, further comprising the step of
recording the enciphered signals on an information carrier at the
transmitting end and playing back the information carrier at a later time
at the receiving end.
12. A method as claimed in claim 1, further comprising the step of
developing the message elements by converting the message signals into
pulse form.
13. A method as claimed in claim 12, wherein the message elements are
formed from a pulse train which is quantized in two stages (FIGS. 2, 3).
14. A method as claimed in claim 12, wherein the message elements are
formed from a pulse train which is quantized in multiple stages.
15. A method as claimed in claim 12, wherein the elements are formed from
analogue pulses without fixed amplitude graduation by the step of sampling
said analogue pulses and converting the sampled pulses to digital form
prior to undergoing paired exchange.
16. A method as claimed in claim 1, wherein the elements are formed by the
step of dividing a variable analogue signal into equal element lengths
(FIG. 5).
17. A method as claimed in claim 1, wherein the elements are formed by the
step of periodically scanning an analogue signal (FIGS. 4, 5).
18. A method as claimed in claim 1, wherein the elements are formed by the
step of converting the message signals in elements whereby each element
consists of an individual pulse (FIGS. 2, 4).
19. A method as claimed in claim 1 wherein the step of forming elements
comprises forming elements each comprised of a plurality of individual
pulses (FIGS. 3, 5).
20. A system for encoding messages through the transposition of selected
message elements, comprising at least one element exchanger means provided
at the transmitting and at the receiving end for exchanging pairs of
elements and at least one control addition means (FIGS. 11, 13);
said element exchanger means having an input for receiving message elements
and an output for delivering exchanged elements for transmission and
containing signal retarder means having a constant delay tine T for
delaying elements selectively applied thereto, and first and second
electronic changeover switch means respectively positioned at the input
and output of the retarder means;
said first changeover switch means having first and second operating
positions for respectively connecting the input of the element exchanger
means directly to the output of the element exchanger means and for
connecting the input of the retarder means to the input of the exchanger
means;
said second changeover switch means having first and second operating
positions for respectively directly connecting the output and input of the
retarder means and for connecting the output of the retarder means to the
output of the element exchanger means;
said first and second electronic changeover switches being normally
maintained in their first positions;
means for generating a quasi-statistical pulse train;
said control addition means containing an electronic interrupter means for
selectively cancelling individual pulses of said statistical pulse train
(cipher signal) said interrupter means having a control input for
cancelling a pulse at its output upon receipt of a control inhibit pulse;
said interrupter means being actuated through pulse retarder means having a
delay time T for delaying pulses at an output of said interrupter means
and replacing the delayed pulses to the control input of said interrupter
means so that no pulses which have a mutual spacing T appear at the output
of the interrupter;
the output pulses of the interrupter being coupled to said first and second
electronic changeover switch means of the element exchanger means to move
the said first and second electronic changeover switch means to their
second positions.
21. A device as claimed in claim 20, wherein said retarder means has a
delay time T which coincides with the length of a message element.
22. A device as claimed in claim 20, wherein said signal retarder means has
a delay time T which coincides with an integral multiple of the length of
a message element.
23. A device as claimed in claim 20, wherein the pulses supplied to the
pulse retarder means in the control addition means, comprises the input
signal of the electronic interrupter means (cipher signal) (FIG. 9).
24. A device as claimed in claim 20, wherein the pulses supplied to the
pulse retarder means in the control addition means comprises the output
signal of the electronic interrupter means (FIG. 11).
25. A device as claimed in claim 20, further comprising a second element
exchanger means similar to said first exchanger means, said first and
second element exchanger means being connected in cascade fashion with the
input of the second exchanger means being coupled to the output of the
first exchanger means (FIG. 16).
26. A device as claimed in claim 25, wherein the signal retarder means of
the two element exchanger means have different delay times.
27. A device as claimed in claim 26, characterized in that the delay time
of one of said signal retarder means is one-third (1/3) the delay time of
the other signal retarder means.
28. A device as claimed in claim 25, wherein said element exchanger means
includes means adapted to exchange elements having unequal element
lengths.
29. A device as claimed in claim 20, further comprising signal scanner
means preceding said element exchanger means to convert variable input
signals into discrete analogue pulses for application to the input of said
exchanger means.
30. A device as claimed in claim 29, wherein the scanning frequency of said
scanner means is selected to be an integral multiple of the element
repetition frequency.
31. A device as claimed in claim 20, wherein the homologous element
exchanger means with the associated control addition means are connected
in cascade in a first sequence at the transmitting end and in a second
reverse sequence at the receiving end which second sequence is the reverse
of that employed at the transmitting end (FIG. 17).
32. A device as claimed in claim 20, further comprising analogue-digital
converter means for converting the analogue input into digital signals for
application to the input of said exchanger means, said signal retarder
comprising digital retarder means for delaying binary pulse trains derived
from said converter means.
33. A device as claimed in claim 20, further comprising digital-analogue
converter means at the output side of the receiver end exchanger means to
obtain analogue output signals fom the digital elements transposed to
return to their original order.
34. A device as claimed in claim 20, further comprising Delta modulation
converter means at the input side of said exchanger means for generating a
pulse train from the input signals by a Delta modulation method.
35. A device as claimed in claim 20, further comprising electric filter
means for dividing the whole frequency band of the message into at least
two component frequency bands with separate time coding of the individual
component frequency bands by element exchange in pairs in said exchanger
means in accordance with different transposition programs and separate
decoding thereof.
36. Means for the permutation of message elements in accordance with a
predetermined program comprising:
an element exchanger having an input for receiving the message elements to
be permutated and an output for delivering permutated elements;
first and second retarder means each having a delay time T.sub.1 for
delaying elements applied thereto;
first and second sets of changeover switches respectively associated with
said first and second retarder means and having a first normal position
for connecting the inputs of said first and second retarder means to said
exchanger means input and for connecting the outputs of said first and
second retarder means to said exchanger means output, and a second
operative position for connecting the output of each retarder means to its
input;
an auxiliary message element path;
a third set of switch means having a first normal position for connecting
said auxiliary path between said exchanger means input and output and
having a second operative position for connecting the output of the
auxiliary path to its input;
means for operating said first, second and third sets of switch means so
that only one of said sets of switch means is in the operative position
during the interval of any message element.
37. A device as claimed in claim 36, wherein the delay time T.sub.1 of the
retarder means is adapted to coincide with an integral multiple of the
element length.
38. A device for encoding or decoding messages by transposing selected
pairs of message elements comprising:
delay means having an input and an output;
a direct signal path for passing signals between its input and the output
of the device with no delay;
a feedback signal path for passing signals between the output and the input
of said delay means;
first switch means for receiving said message in serial fashion and for
selectively coupling said message either to said delay means input when in
a first condition or to said direct signal path input when in a second
condition to cause the message to appear at said device output with no
delay;
second switch means for selectively coupling the delay means output to said
device output when in a first position or to said feedback signal path
when in a second condition;
control means for operating both said first and second switch means;
said control means comprises means for generating a random pulse pattern;
means for sampling successive pulses in said random pulse pattern to
inhibit selected ones of the pulses, so that a pulse interval is always
followed by a no pulse interval, and applying the resulting pulses as
control signals to said first and second switch means.
39. The device of claim 38 wherein said generating means comprises means
for generating quasi-statistic signal pulses said sampling means
comprising logical gating means having a first input coupled to said pulse
generating means and a second input and an output, delay means coupled
between said gating means second input and output, said gating means
inhibiting the generation of a control means output pulse for operating
said switch means to their second conditions whenever pulses are
simultaneously present at said gating means first and second inputs.
40. A device for altering a message by transposing selected ones of the
message elements said device comprising;
an input for receiving the message to a utilization device after undergoing
element transposition;
first and second and third signal paths;
at least two of said signal paths having means for imparting a delay to
message elements applied thereto, while the remaining signal path imposes
no delay to message elements applied thereto;
first, second and third feedback paths;
first, second and third switch means respectively associated with one of
said signal and feedback paths whereby each switch means, either couples a
feedback path across its signal path when in a first switch condition or
couples the signal path between said input and said output when in a
second switch condition;
control means for operating said first, second and third switch means so
that no more than one of said switch means is in said first condition at
any given time.
41. A device as claimed in claim 36, wherein the delay time T.sub.1 of the
retarder means is adapted to coincide with the element length. |
|
 |
|
 |
Claims |
|
|
|
Description |
|
|
BACKGROUND OF THE INVENTION
The invention relates to a method and apparatus for the coded transmission
of messages by splitting up the clear (i.e. uncoded) signals to be
transmitted into elements of equal length, which are transposed at the
transmitting end by being delayed by at least partially different times
and are transposed back at the receiving end by being further delayed by
at least partially different times. The consecutively numbered elements
e.sub.1, e.sub.2, e.sub.3, . . . of the clear signal x have coinciding
lengths T.sub.o (see FIG. 1), and their transposition in time leads, for
example, to the coded signal y, of which the first element e.sub.4 appears
undelayed at the moment t = 3T.sub.o, while the other elements appear with
varying delay. After transmission of the signal y at a receiver, the
elements are restored to their original position by retransposition in
order to recover the original clear signal.
The elements e.sub.1, e.sub.2, . . . may, as shown in FIG. 2, be pulses of
the duration T.sub.o, which are keyed between -1 and +1 or between 0 and 1
in accordance with a telegraphic message. Each element may, however, also
comprise a plurality of individual pulses of a data signal s, as shown in
FIG. 3. The pulses may also be quantized in a plurality of stages. The
formation of elements, the amplitude of which corresponds to the scanned
values, formed at intervals T.sub.o, of a continuously variable clear
signal s (t), is shown in FIG. 4. Instead, however, sections of the clear
signal s (t) of constant length T.sub.o may be formed as elements e.sub.1,
e.sub.2, . . . as shown in FIG. 5. FIG. 5 also indicates that, instead of
these continuously variable signal sections, a train of short individual
pulses c (t) is suitable for forming the elements (see element e.sub. 3).
Now, as a result of the encodingprocess, the sequence of such elements in
time is altered, while the nature of the individual elements can remain
unaltered.
Methods and devices for time coding, that is to say for the transposition
in time of message elements, have become known for example through the
Swiss Pat. No. 212,742 and 232,786, which describe how omissions and also
repetitions of individual elements are avoided by periodically actuated
switches. A periodic repetition of the transposition program effected at
short intervals is undesirable, however, for cryptologic reasons.
Accordingly, in the Swiss Pat. No. 518,658, a method is described which
renders possible the control of the transposition process by random
signals, as a result of which, periodic repetitions of the transposition
program during a transmission are avoided. This control is achieved by
means of a separate position register which, however, considerably
increases the total expenditure necessary. The total expenditure on known
devices is also comparatively heavy because the storage devices used are
generally only partially filled with message elements wherein at least 50%
of the stored locations remain unoccupied at any moment.
BRIEF DESCRIPTION OF THE INVENTION AND OBJECTS
According to the invention, these disadvantages are avoided by
transposition in pairs of two elements at a time, which have a specific
mutual spacing, at the transmitting end and retransposition of the same
elements in pairs at the receiving end, the pairs of elements being
transposed or retransposed at the transmitting end and at the receiving
end being determined by irregular trains of control pulses which coincide
at the two ends, and the elements which do not belong to the pairs of
elements being delayed at the transmitting and receiving ends by a fixed
time T, while the element of each pair which arrives first is delayed by
double the time 2T at the transmitting and receiving ends and the second
element is not delayed.
It is therefore one object of the invention to provide method and apparatus
for encoding and/or decoding messages by transposing selected pairs of
message elements while leaving remaining message elements untransposed.
Another object of the invention is to provide method and apparatus for
encoding and/or decoding messages by transposing selected pairs of message
elements so that one element of the pair undergoes a delay 2T and the
remaining element of the pair undergoes no delay.
Still another object of the invention is to provide method and apparatus
for encoding and/or decoding messages by transposing selected pairs of
message elements so that one element of the pair undergoes a delay 2T and
the remaining element of the pair undergoes no delay and wherein message
elements not treated in pairs undergo a delay T, so that T = n T.sub.o
where T.sub.o = message element length, and n = 1,2,3, . . . , n.
BRIEF DESCRIPTION OF THE FIGURES
The above, as well as other objects of the invention, will become apparent
from the following description and drawings, in which:
FIG. 1 shows one manner in which message elements of a message may be
transposed.
FIGS. 2 - 5 show waveforms of various message formats which may undergo
encoding (and decoding) by the techniques and apparatus of the present
invention.
FIG. 6 shows a circuit for carrying out the exchange of message elements in
pairs,
FIGS. 7 and 8 are diagrammatical illustrations of the exchange of adjacent
elements,
FIGS. 9 and 11 show circuits for obtaining control signals for the
actuation of the transposition switch from cipher signals,
FIGS. 10 and 12 show examples of cipher signals and control signals
obtained therefrom,
FIG. 13 shows a circuit for the transposition in pairs of non-adjacent
elements with associated circuitry for obtaining the control signals,
FIGS. 14 and 15 are diagrammatic illustrations of the exchange in pairs of
non-adjacent elements,
FIGS. 16 and 17 show a circuit for the repeated exchange in pairs with
cipher-signal preparation and a circuit for the repeated re-exchange with
cipher-signal preparation,
FIG. 18 is a diagrammatic illustration of the repeated exchange in pairs,
FIG. 19 is an illustration of the delay times which occur with repeated
exchange in pairs,
FIG. 20 shows a circuit for permutation in accordance with a constant
program,
FIGS. 21 and 22 are diagrammatic illustrations of permutations in
accordance with a constant program,
FIG. 23 shows a block circuit diagram of devices for the time coding by
element exchanges in pairs in conjunction with permutations in accordance
with a fixed program and for the decoding by element exchanges in pairs in
conjunction with permutations.
DETAILED DESCRIPTION OF THE INVENTION
An explanation of the invention will now be given with reference to FIG. 6,
which shows a simple circuit for carrying out the exchange of elements in
pairs. The circuit contains a retarder R with the transit time T.sub.o,
which corresponds to the length of one message element. This retarder can
be connected, through the switches H.sub.1, H.sub.2 (in position "O", as
shown), to the input line and the output line of the circuit so that one
element at a time of the clear signal x is supplied to the retarder, while
at the same time a stored or delayed element is extracted therefrom as
output signal y. By means of a pulse of the control signal a with the
duration T.sub.o, on the other hand, the switches are brought into the
position designated by "I", so that one element of the input signal x at a
time again appears directly as an element of the output signal y, while
the preceding input element continues to be stored by being fed back from
the output to the input of the retarder. The position of the beginning of
the element in the retarder is indicated by the variable length d.
In the absence of a pulse of the control signal a, therefore, an element
e.sub.1 of the input signal x will reappear as element e.sub.1 of the
output signal y after the time T.sub.o, as shown in FIG. 7. In the course
of the duration of the element e.sub.6, on the other hand, for example, a
pulse of the control signal a appears so that this element reaches the
output without delay, through the switch H.sub.1, (indicated in broken
lines in FIG. 7), while the preceding element e.sub.5 is fed back to the
input of the retarder through the switch H.sub.2 and therefore only
reaches the output of the circuit after an additional delay time T.sub.o
or with a total delay 2T.sub.o. The passage through twice can be
recognized by the position d of the initial edge of the element, which can
be seen from FIG. 7. Whereas the element e.sub.1 is merely delayed by the
time T.sub.o, therefore, a delay reduced to 0 has occurred with the
element e.sub.6 and a delay increased to 2T.sub.o with the element
e.sub.5, so that these last two elements appear transposed in time in the
output signal y. In a similar way, the pair of elements e.sub.2, e.sub.3
is also transposed in time as shown in FIG. 7, while the element e.sub.4
for example is transmitted with delay but without transposition with any
other element. The same transpositions are indicated again
diagrammatically in FIG. 8. It should be noted that the time zero has been
advanced (i.e. shifted one "frame" to the left) by one time interval
T.sub.o in the signal y in order to achieve a clearer illustration.
It should be noted that during the transposition in pairs as described, the
switches H.sub.1, H.sub.2 should never be actuated for longer than the
duration T.sub.o of one element, in order that no element may be stored
longer than 2T.sub.o. Accordingly, immediate repetitions (for example
00110) of the switching pulses are not permitted on the control signal a.
In order to extract the control signals a.sub.o from a cipher-signal
w.sub.o following a quasi-random course, a cipher-signal addition circuit
SZ.sub.o as shown in FIG. 9 is therefore suitable. As a result of delaying
each individual pulse of the cipher signal w.sub.o by the element length
T.sub.o in the retarder V.sub.o, a blocking signal v.sub.o results which
suppresses a possible following pulse of the cipher signal in the
interrupter U.sub.o. The effect of this suppression is shown by way of
example in FIG. 10. The suppressed pulses are designated by underlining. A
disadvantage in this case, however, is that with an uninterrupted train of
three or more pulses, all the pulses except the first are cancelled. This
disadvantage is avoided with the cipher-signal addition circuit SZ.sub.1
shown in FIG. 11, in which the interrupter U.sub.1 is actuated by the
pulses of the control signal a.sub.1 delayed in v.sub.1. With an
uninterrupted train of a plurality of pulses of the cipher signal w.sub.1,
only every other pulse is suppressed in this case so that the control
signal a.sub.1 indicated in FIG. 12 results for example, and meets the
requirements for an exchange of elements in pairs. In FIG. 11, apart from
the device PT.sub.1 already explained for the exchange of elements in
pairs, a cipher-signal generator SG is indicated, the construction and
mode of operation of which may correspond to known constructions. Devices
for generating cipher signals with digital circuits are described for
example in the Swiss Pat. No. 361,839.
Depending on the nature of the clear signals x, digital or analogue stores
of known construction should be used as retarders R for exchanging the
elements in the pair exchanger PT. In this case, it may be a question of
delay lines or balancing networks, electro-mechanical retarders (for
example acoustic systems) or electromagentic stores (for example magnetic
sound recording with moving medium). Electrical shift registers are
particularly suitable, with which signals keyed digitally (for example as
shown in FIGS. 2 and 3) can easily be stored if operated at an appropriate
clock frequency. With analogue signals (for example as shown in FIGS. 4
and 5), periodic scanning and storage of the scanned values (c(t) in FIG.
5) is necessary. These scanned values can also be converted, by binary
coding, into corresponding pulse groups, the storage of which is then
effected with digital stores having an appropriately larger number of
stages. In this case, with the pair exchanger PT.sub.1 shown in FIG. 11,
it is necessary to connect an analogue-digital converted at the input side
to extract digital input signals from the clear signal x and to connect a
digital-analogue converter at the output side to extract output signals y
in analogue form. Delta modulation is also possible, however, instead of
the binary coding. The changeover switches H.sub.1, H.sub.2 may
appropriately be realized by suitably controlled semiconductor switching
elements, which is also true for the interrupter U.sub.1 in the
cipher-signal addition circuit SZ.sub.1.
The effectiveness of the time coding is increased by transposition in
pairs, of elements which are not immediately adjacent. In FIG. 13 a device
PT.sub.3 is shown for the transposition in pairs of two elements at a
time, the beginnings of which have a mutual spacing of three element
lengths T.sub.o, and corresponding element trains are illustrated in FIGS.
14 and 15 to explain the operation by way of example. When the switches
H.sub.3, H.sub.4 are in the normal position shown, the elements of the
output signal y appear delayed by 3T.sub.o in comparison with the input
signal x, if the delay of the retarder R.sub.3 likewise amounts to
3T.sub.o. This is the case, for example, with the element e.sub.2 (see
FIG. 14), because said switches are in the normal position shown both
during the supply and also during the extraction of this element. Although
the element e.sub.3 is likewise supplied to the retarder through the
switch H.sub.3, nevertheless after a first passage through this retarder,
it is again fed back to the input of the retarder through the switch
H.sub.4, because at this time, this switch is brought into the operative
position (not shown) by a pulse of the control signal a.sub.3. At the same
time, an element e.sub.6 of the input signal x is conveyed, without delay
to the output through the switch H.sub.3 which is likewise actuated
(indicated in broken lines in FIG. 14). Only three element lengths later
does the stored element e.sub.3 finally appear through the switch H.sub.4
restored to the normal position, in the output signal y. In a similar
manner, the elements e.sub.1, e.sub.4 and e.sub.7, e.sub.10 for example
are also transposed, while e.sub.5 and e.sub.8 are passed on with simple
delay without being transposed. This process is illustrated again, with
the associated control signals, in FIG. 15. The advancing of the time zero
(i.e., the shifting left of the time frame) should again be noted in this
simplified illustration. As a result of operation with control pulses
having the uniform length T.sub.o , the effect is achieved that a
plurality of elements of corresponding length always travel through the
retarder.
In order to avoid a further feedback of all elements which have already
been delayed twice, care must be taken to ensure that no further pulse
follows a pulse of the control signal a.sub.3 with the spacing 3T.sub.o.
For this reason there is provided in the cipher-signal addition circuit
SZ.sub.3, a blocking switch U.sub.3 which is actuated by the pulses of the
control signal a.sub.3 delayed by three element lengths T.sub.o in
V.sub.3, so that any following inadmissible control pulses are eliminated.
Here, too, the cipher signals w.sub.3, from which the control signals
a.sub.3 are obtained by suppression of inadmissible pulses, are taken from
a cipher-signal generator SG.
In order to further increase the effectiveness of a time coding, the
interconnection of a plurality of pair-exchange process circuits is
advisable so that an increase in the possible displacements of each
element comes about. In FIG. 16, a device ZT can be seen in which a first
transposition in pairs is effected of elements of the clear signal x
through the retarder R.sub.3 and the switches H3, H4, as a result of which
a signal y results, the elements of which may have additional
displacements by 3T.sub.o or 6T.sub.o as in FIGS. 13 and 15. A second
transposition in pairs is then effected through the retarder R.sub.1 and
the switches H.sub.1 and H.sub.2 with smaller displacements similar to
FIGS. 6 and 8. The cipher-signal addition circuit SZ is also equipped with
retarders V.sub.3 and V.sub.1 respectively, corresponding to FIGS. 13 and
11 respectively, in accordance with the unequal displacement times. This
cascade connection of two transposition processes in pairs produces, from
a clear signal x, the element numbers of which are designated by n(x) in
FIG. 18, first the intermediate signal y, of which the element numbers
n(y) are likewise given in FIG. 18, and finally, as a result of further
element exchange in pairs, the output signal z with the element numbers
n(z). Whereas displacements of 0 and +3 element lengths occur in the
intermediate signal, the second exchange produces displacements of 0,
+T.sub.o, +2T.sub.o, +3T.sub.o, +4T.sub.o can appear in the output signal
z in comparison with a mid position of the elements. In view of the fact
that even this mid position has a displacement of 4T.sub.o, because
negative displacements in time are impossible, the output elements of the
time coding device ZT therefore appear with delays of O, T.sub.o, 2T.sub.o
3T.sub.o, . . . to 8T.sub.o in comparison with the input elements. The
delays occurring in the example shown are given in FIG. 19 as integral
multiples r(n) of the element length T.sub.o over the element numbers n of
the input signal x. It can be seen that a very effective mixing of all the
elements of the message comes about already as a result of pair exchanging
twice. This process could be extended by one or more further pair
exchanges. In this case, it is advisable to avoid the same storage times
for the various exchange processes. The number of possible displacements
becomes particularly high if the storage times are graduated in accordance
with a ternary system, in that retarders are used having transit times of
T.sub.o, 3T.sub.o, 9T.sub.o . . . =3.sup.i T.sub.o (i = a whole number),
because thus all total delays mT.sub.o between 0 and (3.sup.k.sup.+1 -
1)T.sub. o are possible (m = a whole number, k = total number of the pair
transposition devices).
A device which as shown in FIG. 17, corresponds largely to the
transposition device at the transmitting end, serves for the re-exchange
of the message elements at the receiving end. From the coded signal z*
received, which coincides with z, as a result of a first re-exchange with
the retarder R*.sub.1 and the switches H*.sub.1, H*.sub.2, an intermediate
signal y* is again formed which coincides with y and (apart from the delay
of the transmission channel) is delayed by 2T.sub.o in comparison with y,
because the untransposed elements are subjected to a delay of T.sub.o at
the transmitting end and at the receiving end. With the transposition of
the elements e.sub.5 and e.sub.6 shown in FIG. 7, re-exchange of these
elements comes about when a following analogue transposition device
receives a control pulse a at the moment the element e.sub.5 is received,
so that this element is not further delayed, while the preceding element
e.sub.6 is delayed by 2T.sub.o and so comes back into the original
position in relation to e.sub.5. Accordingly, the control pulses a*.sub.1
of the first re-exchange with the switches H*.sub.1, H*.sub.2 must be
displaced by T.sub.o in comparison with the control pulses a.sub.1 of the
exchange shown in FIG. 16 with the switches H.sub.1, H.sub.2, in the
device also shown in FIG. 17. This displacement is achieved by an
additional delay T.sub.o of the cipher signal w*.sub.1 at the receiving
end (FIG. 17). In this case, it is assumed that the cipher-signal
generator SG* at the receiving end is synchronized with the cipher-signal
generator SG at the transmitting end by auxiliary signals u and u*
transmitted separately, for example by the method described in the Swiss
Pat. No. 361,839. In the case of element exchange in pairs with
displacement by three element lengths as shown in FIGS. 13 and 14, it
should be noted that an element e.sub.3 which is displaced by six element
lengths in the exchange process at the transmitting end (see FIG. 14),
must not be further delayed during the re-exchange at the receiving end,
while the element e.sub.6 which is not delayed at the transmitting end has
to be delayed by six element lengths at the receiving end. The control
pulse for the re-exchange at the receiving end must therefore coincide
with the element e.sub.3 received; that is to say the control of the
re-exchange must be delayed by 3T.sub.o in comparison with the control at
the transmitting end, if no additional delays have to be taken into
consideration. In the transmission system as shown in FIGS. 16, 17,
however, as already explained, there is a difference in time of 2T.sub.o
between the signals y and y*, so that the control signal a*.sub.3 for the
re-exchange in pairs in the retarder R*.sub.3, the transit time of which
amounts to 3T.sub.o, must be delayed altogether by 3T.sub.O + 2T.sub.O =
5T.sub.o in comparison with the control signal R.sub.3 for the exchange in
pairs in R*.sub.3. The retarder W*.sub.3 is provided in the cipher-signal
addition circuit SZ* at the receiving end to ensure this delay time (FIG.
17).
The effectiveness of an enciphering by exchanging elements in pairs is also
increased by additional permutation of the elements in accordance with a
fixed program. A device ZT.sub.o, which is suitable for this, may contain
two retarders R.sub.1, R.sub.2 with an identical transit time, as shown in
FIG. 20. Individual elements of the input signal y.sub.1 can be supplied
to these retarders through the switches A.sub.1 and B.sub.1 respectively,
while the extraction of elements to form the output signal y.sub.2 is
possible through the switches A.sub.2 and B.sub.2 respectively. When the
switches are not actuated, however, the retarder output is connected back
to its input in each case. Finally direct passing-on of elements of the
input signal y.sub.1 to the output of the device is possible through the
further switches C.sub.1, C.sub.2. The switches A.sub.1, A.sub.2 are
always actuated simultaneously, likewise the switches B.sub.1, B.sub.2 and
C.sub.1, C.sub.2, for example in accordance with the periodic program S
given at the top in FIG. 21 (the switches not recited in a time interval
being in the normal position in each case). The elements of the input
signal y.sub.1 are numbered consecutively with the numbers given below the
switch program S in FIG. 21. The switching through by the switch C is
indicated diagrammatically underneath (DC). The element No. 3 is passed on
directly through the switch C to the output so that this element appears
without delay in the output signal y.sub.2 (FIG. 21 bottom). The element
No. 5 on the other hand, passes through the simultaneously actuated switch
A.sub.1 to the retarder R.sub.1 (the delay in R.sub.1 is illustrated
symbolically in the next line "VR.sub.1 "), and immediately after being
delayed only once, it is conveyed to the output through A.sub.2. The input
element No. 4, which reaches the retarder R.sub.2 through the switch
B.sub.1 (see next line "VR.sub.2 "), on the other hand, is fed back from
the output of the retarder to the input thereof through the switches
B.sub.1, B.sub.2 which alternate in the normal position after this input;
it is only extracted therefrom again after passing through three times and
added to the output signal y.sub.2, as soon as the switches B are actuated
again. On the assumption that the transit time of a retarder R coincides
with the element length T.sub.o, such storage and switching-over finally
leads to an output signal y.sub.2 with elements transposed in time, as can
be seen from the resulting numbering shown at the bottom of FIG. 21.
Mutual displacements of the elements by greater times are possible with an
increased transit time of the registers R. With a delay time 3T.sub.o of
the registers R.sub.1 and R.sub.2, the displacements which can be seen
from FIG. 22 result, as the switch control is effected in accordance with
program S given across the top of FIG. 22. The element No. 2 for example
is transmitted directly through switches C.sub.1, C.sub.2 while the
element No. 3 is delayed by three element lengths in the retarder R.sub.1.
The element No. 5, on the other hand, after being fed back twice, is
subjected to a delay of 9T.sub.o in the retarder R.sub.2. The element No.
4 is subjected to a delay of 6T.sub.o in the same retarder and the element
No. 1 is actually delayed by 12T.sub.o in R.sub.1. Because of the periodic
repetition of the switching-over program, the elements No. 1, 6, 11 . . .
are delayed by the same amounts, likewise the elements 2, 7, 12 . . . and
the elements 3, 8, 13 . . . and so on. Further possibilities for carrying
out the periodically repeated transposition are provided, for example, by
increasing the delay times of R.sub.1 and R.sub.2 to 4T.sub.o or even
greater amounts, or by using three or more retarders which are connected
to the inputs and outputs of the device in a similar manner by switches
actuated in pairs.
An interconnection of the device ZT.sub.o, which has been explained, for
the periodically repeatd permutation of message elements, with devices
PT.sub.1 and PT.sub.2 for the exchange of such elements in pairs, is shown
in FIG. 23. The control-signal additions for obtaining the control signals
a.sub.1 and a.sub.2 from the cipher signals w.sub.1 and w.sub.2 are
designated by circuits SZ.sub.1 and SZ.sub.2. A further control-signal
addition circuit SZ.sub.o serves to produce the periodically repeated
control signals a.sub.o for the actuation of the switches A, B, C of the
permutation device ZT.sub.o. The corresponding devices at the receiving
end for reversing the transpositions and the signals appearing in the
course of this are shown in FIG. 23 using the same symbols. An additional
asterisk (for example y*.sub.2) serves to make a distinction from the
devices and signals at the transmitting end. The transit times of the
retarders contained in PT.sub.1 and PT.sub.2 are preferably selected
unequal in order to obtain, once again, as great a multiplicity as
possible of the element displacements which can be achieved.
The interconnection described, between devices for exchanging elements in
pairs and a device for permutating elements in accordance with a fixed
program, leads to resulting transpositions of the message elements which
are still very difficult to take in at a glance even with knowledge of the
fixed permutations. In particular, the fact should be noted that the
number of possible displacements of elements is considerably greater than
with simple exchange of elements in pairs and that the total expenditure
necessary remains comparatively low because even with the permutations,
operation involves optimum utilization of all signal stores.
Supplementing the exchange of elements in pairs by an additional time
coding of known type is, of course, also possible. In this case, too, the
individual transposition operations at the receiving end must be provided
in reverse sequence compared with the transmitting end. There is also the
possibility, however, of an effective amplification of the exchange of
elements in pairs according to the invention by enciphering processes of
another kind, such as additional splitting up of the elements into
individual frequency bands which are transmitted in a transposed frequency
position. In particular, there is also the possibility of a division into
two or more frequency bands, which are each subjected, independently of
one another and in accordance with a different program, to a time coding
by exchange of elements in pairs. Thus apart from at least two devices for
the exchange of elements in pairs, separate filters for dividing the
message into at least two sub-bands are necessary for carrying out such
enciphering.
The effectiveness of the exchange of elements in pairs can also be
increased by interconnecting two or more devices for the exchange of
elements in pairs, working with different lengths of element. The element
lengths are preferably in an integral ratio to one anothe | | |
