Note: Descriptions are shown in the official language in which they were submitted.
CA 02241041 1998-06-19
"An optical device and a method of processing a digital optical signal in
parallel and in free space"
**********
The present invention relates to an optical device and a method of
processing a digital optical signal in parallel and in free space.
Further to the widespread development of optical communication
systems, the need is felt for devices capable of performing, at high
speed, various operations on those bits which constitute the digital
information of an optical signal and of creating advanced "super
computers" and network nodes suitable for modern methods of
infGrmation transfer (e.g. ATM).
The equipment currently used for processing optical signals is
inadequate to manage the ever increasing transmission speeds possible
in transmission systems of optical fibre type. In fact, such equipment
consists of digital electronic devices with a limited band in comparison
with the optical band available in optical fibre transmission systems
based on serial information processing.
Furthermore, the equipment which enables the signal to be maintained
in optical form, e.g. intensity or phase modulators, processes signals
under the control of an electrical signal and suffers from the inherent
limitations of electronic devices.
Consequently, there is still a great need to obtain the full benefit of the
optical band available in optical fibre transmission systems, processing
the signal by means of purely optical control signals and thus overcoming
the inherent limitations of electronic devices.
US patent 5 589 967 describes a method and a device for transmitting
and switching packets in an optical network which carries out multiplexing
by a synchronous time division system, in which the packets are
transmitted at a given speed and the transmission rate is determined by
line occupancy time.
CA 02241041 1998-06-19
Patent EP 0 742 660 A1 describes a signal processor for processing
digital signals in the physical domain (e.g. optical). A symbol flow is
conducted towards different delay branches. The number of delay
branches is such that at any given time, in at least one of the branches, a
"1" symbol and a "0" symbol are available. By controlling the opening and
closing of the on/off switches, the symbol values can be changed.
Furthermore, extra light sources are not necessary and the processor is
transparent, that is the processor output symbols possess exactly the
same physical characteristics as the processor input symbols.
Processing of the optical signals takes place in a serial way and in guided
propagation.
Boffi P. et al. ["Optical time-to-space converter", Optics
Communications,123,473-476 (1996)] describe an all-optical time-to-
space converter of free propagation type which translates time-coded
binary words of an optical communication signal of 1550 nm to equivalent
space-coded words. Conversion is carried out by means of four optical
gates, one for each of the four polarized optical signals. Each optical gate
comprises a first and second indium-doped cadmium tellurium crystal
(CdTe:ln). There are two optical control beams, one for the first four and
one for the second four crystals. In the right-hand column of page 476,
lines 10-12, the authors wish that it might be possible for this time-to-
space converter to constitute the input stage of an all-optical optic signal
processor in free space, but do not indicate what means could be used to
construct such a processor.
The present invention is designed for creation of such means and
such a processor.
A first object of the present invention is therefore an optical device for
processing a digital optical signal in parallel and in free space, the said
device comprising:
CA 02241041 1998-06-19
a) means of input in guided propagation of a digital optical signal
comprising at least one temporal series of n bits;
b) means capable of converting the said digital optical signal
to a beam of n signals in guided propagation, each of the same
comprising at least one temporal series of n bits;
c) means capable of converting the said beam of n digital optical signals
in guided propagation, each of the same comprising at least one
temporal series of n bits, to a beam of n digital optical signals in free
space;
d) means capable of selecting, in parallel and in free space, a bit pre-
selected from the said at least one temporal series of n bits of each of
the said n digital optical signals, so as to transform the said at least
one temporal series of n bits into a spatial figure of the said n bits
carrying the same information as that previously contained in the said
at least one temporal series,
characterised in that the said device also comprises
e) optical means capable of modifying, in parallel and in free space, at
least one bit of the said spatial figure of the said n bits, the said means
being selected by the group consisting of means capable of
eliminating at least one bit, means capable of inserting at least one bit
and means capable of modifying the form of at least one bit, and
f) means of output of the said at least one bit of the said spatial figure of
n bits.
Throughout the present description and of the following claims, the
expression "propagation in free space" is used to indicate all the modes
of propagation of an optical signal which, in a device according to the
present invention, are not guided by an optical fibre.
Typically, the said means capable of converting the said digital optical
signal to a beam of n digital optical signals in guided propagation, each of
the same comprising at least one temporal series of n bits, comprise:
CA 02241041 1998-06-19
a) means of cloning the said digital optical signal in the said n digital
optical signals in guided propagation;
b) first n lines capable of interval-timing the said digital optical signals in
guided propagation according to predetermined time intervals;
c) means of controlling and, if necessary, changing the state of
polarization of the said n digital optical signals in guided propagation.
Preferably, the said means capable of converting the said beam of n
digital optical signals in guided propagation to a beam of n digital optical
signals in free space comprise means of collimation capable of guiding
the said n digital optical signals in free space in a predetermined direction
and maintaining them within predetermined transverse dimensions.
Typically, the said means capable of selecting, in parallel and in free
space, a bit pre-selected from the said at least one temporal series of n
bits of each of the said n digital optical signals, so as to transform the
said at least one temporal series of n bits into a spatial figure of the said
n bits carrying the same information as that previously contained in the
said at least one temporal series comprises:
a) a first optical switching module;
b) a second optical switching module arranged in series in relation to the
said first optical switching module;
c) means of supplying to the said first and, respectively, to the said
second optical switching modules a pair of a first and a second beam
of optical control pulses having a predetermined time interval between
each other.
d) a dichroic mirror capable of guiding the said at least one pair of a first
and a second beam of optical control pulses in collinear form in
relation to the propagation direction of the said n digital optical signals.
e) means of collimation to guide the said at least one pair of a first and
second beam of optical control pulses in free space in such a way that
they are incident upon the said dichroic mirror at a pre-fixed angle.
CA 02241041 1998-06-19
Preferably the said first and second optical switching modules
comprise, respectively, a first and second element, capable of causing
the plane of polarization of the said n digital optical signals in free space
to rotate by a predetermined angle under the action of the said at least
one pair of a first and a second beam of optical control pulses, and they
also comprise, respectively, a first and a second polarization analyzer
capable of filtering, along a predetermined plane of polarization, the said
n digital optical signals output from the said first and, respectively,
second elements.
The said first and second elements preferably consist of a first and a
second indium-doped cadmium tellurium monocrystal (CdTe:ln).
Typically, the said first and second polarization analyzers are oriented
essentially orthogonal to each other.
Typically, the said means of output comprise optical focusing means
capable of guiding the said spatial figure of n bits in free space in means
of guided propagation.
Preferably the said means of output also comprise second n lines
capable of interval-timing the said n bits of the said spatial figure
according to predetermined time intervals.
More preferably, the said means of output also comprise means
capable of conveying the said interval-timed n bits and in guided
propagation to a processed digital optical signal comprising at least one
temporal series of n bits.
Typically, the means of eliminating the said at least one bit of the said
spatial figure of n bits comprise a third switching module and means of
supplying at least one optical elimination signal.
Preferably the said third switching module comprises an element
capable of causing the plane of polarization of at least one of the said n
digital optical signals, output from the said first n interval-timing lines and
the said means of collimation, to rotate by a predetermined angle, under
CA 02241041 1998-06-19
the action of the said at least one optical elimination signal, and also
comprises a polarization analyzer capable of filtering, along a
predetermined plane of polarization, the said n digital optical signals
output from the said element.
More preferably, the said element consists of an indium-doped
cadmium tellurium monocrystal (CdTe:ln).
According to a preferred solution, the said at least one optical
elimination signal is co-linear to and overlaps a single one of the said n
digital optical signals output from the said first n interval-timing lines.
In a preferred design form, the electrical field is applied perpendicular
to face 1100 of the said crystal whilst the signal and the control beam are
applied perpendicular to face 1900.
Preferably the said means of elimination of the said at least one bit of
the said spatial figure of n bits also comprise, downstream of the said
third switching module, means transparent to the wavelength of the said
n digital optical signals and capable of reflecting the wavelength of the
said at least one optical elimination signal.
Typically, the means capable of inserting at least one bit into the said
spatial figure of n bits comprise means of supplying at least one optical
insertion signal to the said first and second optical switching modules.
Preferably, the said at least one optical insertion signal possesses the
same wavelength and power as the optical signals constituting the said
spatial figure of n bits.
Furthermore, the said means of supplying the said at least one optical
insertion signal also comprise
- means transparent to half the power of the said n digital optical signals
and of the said at least one optical insertion signal and
- means capable of reflecting the other half of the power.
Typical examples of the said transparent means comprise a 50/50
optical beam splitter.
CA 02241041 1998-06-19
Preferably the said 50/50 optical beam splitter is essentially inclined
45~ in relation to the direction of the said n digital optical signals and the
said at least one optical insertion signal, the direction of the said digital
optical signals being essentially orthogonal to the direction of the said at
least one optical insertion signal.
More preferably, the said means of supplying at least one optical
insertion signal also comprise means capable of collimating in free space
the said at least one optical insertion signal so that it is incident upon the
said 50/50 beam splitter at an angle of approximately 45~ and so that it
overlaps one of the said n digital optical signals.
Typically the said means of supplying the said at least one optical
insertion signal also comprise means of controlling and, if necessary, of
varying the state of polarization thereof.
Preferably the means of modifying the form of the said at least one bit
of the said spatial figure of n bits comprise means capable of varying the
said time interval between the said first and second beam of optical
control pulses of the said at least one pair of a first and a second beam
of optical control pulses.
According to an embodiment, the said second n lines interval-time at
least one bit of the said spatial figure of n bits by a time such that in
output the said at least one bit is delayed, in relation to the other bits of
the said spatial figure of n bits, by a different time quantity from that of
the said bit in output from the said input means.
According to a variant, the said optical means capable of modifying, in
parallel and in free space, at least one bit of the said spatial figure of n
bits also comprise means capable of executing algebraic operations on
the said spatial figure of n bits.
According to another variant, the said optical means capable of
modifying, in parallel and in free space, at least one bit of the said spatial
CA 02241041 1998-06-19
-
figure of n bits also comprises means capable of executing symmetry
operations on the said spatial figure of n bits.
Preferably the said means capable of executing the said algebraic
operations comprise at least one element capable of causing the plane of
polarization of the said n bit(s) of the said spatial figure, output from the
said first and second optical switching modules, to rotate by a
predetermined angle, under the action of an optical beam enabling the
said algebraic operations.
More preferably, the said mean capable of executing the said
algebraic operations also comprise at least one means transparent to a
first predetermined plane of polarization of the said n bits of the said
spatial figure, the said transparent means also being capable of deviating
a second plane of polarization essentially orthogonal to the said first
plane.
Again more preferably, the said means capable of executing the said
algebraic operations also comprise at least one mirror capable of
reflecting the said n bits of the said spatial figure.
Typically the said means capable of executing the said symmetry
operations comprise at least one element capable of causing the plane of
polarization of the said n bits of the said spatial figure, output from the
said first and second optical switching modules, to rotate by a
predetermined angle under the action of an optical beam enabling the
said symmetry operations.
Furthermore, the said means capable of executing the said symmetry
operations also comprise at least one means transparent to one first
predetermined plane of polarization of the said n bits of the said spatial
figure, the said transparent means also being capable of deviating a
second plane of polarization essentially orthogonal to the said first plane.
CA 02241041 1998-06-19
- 10-
Moreover, the said means capable of executing the said symmetry
operations also comprise means capable of reflecting the said n bits of
the said spatial figure.
Preferably the said means capable of executing the said symmetry
operations also comprise means capable of varying the state of
polarization of the said n bits of the said spatial figure.
A method of processing a digital optical signal in parallel and in free
space constitutes a second object of the present invention, the said
method comprising:
a) feeding of a digital optical signal comprising at least one temporal
series of n bits;
b) conversion of the said digital optical signal to a beam of n digital
optical signals in guided propagation, each of the said signals
comprising at least one temporal series of n bits;
c) conversion of the said beam of n digital optical signals in guided
propagation, each of the said signals comprising at least one temporal
series of n bits, to a beam of n digital optical signals in free space;
d) selection, in parallel and in free space, of a bit pre-selected from the
said at least one temporal series of n bits of each of the said n digital
optical signals, so as to transform the said at least one temporal series
of n bits into a spatial figure of the said n bits carrying the same
information as that previously contained in the said at least one
temporal series,
characterised in that the said method also comprises modification, in
parallel and in free space, of at least one bit of the said spatial figure of n
bits.
A third object of the present invention consists of a device for
modifying the temporal duration of at least one bit of a temporal series of
n bits transformed into a corresponding spatial figure of n bits,
characterised in that
CA 02241041 1998-06-19
- the said n bits of the said spatial figure are guided in a beam of n
optical waveguides at the end of which the said n bits are
subsequently guided in a single optical waveguide in the form of
temporal series of n bits, and in that
- the said beam of n optical waveguides comprises at least a section of
optical waveguide having a length which is selected such that in input
to the said single optical waveguide one bit which has passed through
the said section is separated from at least one of the other bits of the
said series by a different time interval from that by which it was
separated in the initial temporal series.
The special characteristic of the optical device according to the
invention consists of the fact that the device enables processing of a
digital optical signal, maintaining the said signal in optical form and under
the control of optical control signals.
As already stated above, this enables the band limitations of
conventional electronic devices and the slow response times of
electronically controlled optical devices(e.g. Iiquid crystal and thermo-
optic modulators) to be overcome.
Further advantages of the device according to the present invention
consist in the fact that the device enables processing of signals in free
space, thus benefiting from the "space" resource offered by optics, as
well as processing in parallel, thus enabling a higher speed to be
obtained in comparison with the serial type processing typical of
conventional electronic processors and optical fibre guiding systems and
furthermore in waveguide.
Furthermore, the optical device according to the present invention
enables processing of signals having the wavelength typical of optical
fibre type transmission systems, e.g. approximately 1300 and 1550 nm.
CA 02241041 1998-06-19
Characteristics and advantages of the invention are now illustrated
with reference to embodiments represented by way of non limiting
example, in the appended drawings, in which:
Figure 1 shows a first embodiment of an optical device according to
the invention;
Figure 2 shows means of collimation and focusing of the device
illustrated in Figure 1;
Figure 3 shows the operation of an optical switching module of the
device illustrated in Figure 1;
F,gure 4 shows the mode of operation of two optical switching
modules of the device illustrated in Figure 1;
Figure 5 shows the mode of operation of a dichroic mirror interposed
between two elementary modules of the device illustrated in Figure 1;
Figure 6 shows means of the optical device illustrated in Figure 1
capable of converting a temporal sequence of n bits into a
corresponding spatial figure;
Figure 7 shows the logic operation diagram of the means illustrated in
Figure 6;
Figure 8 shows an embodiment of means which are insertable in an
optical device in accordance with the invention for execution of
algebraic operations.
Figure 9 shows an embodiment of means which are insertable in an
optical device in accordance with the invention for execution of
symmetry operations;
Figure 10 shows an embodiment of means of collimation for eight
digital optical signals of an optical device according to the invention.
Figures 11 a and 11 b show a graphic representation of a 4 bit signal in
input (Figure 11 a) and in output (Fig. 11 b) following an operation of
elimination of one bit in the device as illustrated in Figure 1;
CA 0224l04l l998-06-l9
- 13-
- Figures 1 2a and 1 2b show a graphic representation of a 4 bit signal in
input (Fig. 12a) and in output (Fig. 12b) following an operation of
insertion of one bit in the device illustrated in Figure 1;
- Figures 13a and 13b show a graphic representation of a 4 bit signal in
input (Fig. 13a) and in output (Fig. 13b) after an operation of
modification of the form of four bits in the device illustrated in Figure 1;
- Figures 14a and 14b show a graphic representation of a 4 bit signal in
input (Figure 14a) and in output (Figure 14b) following an operation of
modification of the bit rate of four bits in the device illustrated in Figure
1.
The embodiment of Figure 1 represents an optical device for
processing in free space and in parallel four bits of a digital optical signal
6 comprising at least one temporal series 7a of four bits.
Means of input, for generating the said optical digital signal 6,
comprise a laser 60 with DFB (pigtailed) semiconductor emitting at
approximately 1550 nm and modulated by a word generator 600 at a
frequency of approximately 140 MbiVs.
The power of the said digital optical signal 6 output from the said laser
60 is approximately 1 mW. An optical amplifier (not shown in the Figure),
e.g. a fibre doped with erbium and having variable gain, enables a power
of approximately 15 mW to be obtained in output from the said means of
input 60 and 600, so as to compensate any losses from the subsequent
stages.
An optical fibre 1 x 4 coupler 16 clones the said digital optical signal 6
thus amplified so as to obtain four identical copies 1, 2, 3 and 4 of the
four bit temporal series 7a.
Four lines (line sections) made of optical fibre 110, 120, 130 and 140
in different lengths delay, in relation to each other, the said four digital
optical signals 1, 2, 3 and 4 by multiple quantities of the bit time which, at
approximately 149 Mbit/s, is approximately 7.12 ns. Because the speed
CA 02241041 1998-06-19
- 14-
of the light in the glass is approximately 2 x 108 m/s, in order to work a
delay of 7.12 ns it is necessary to lengthen the path by 1.428 m. The
optical fibre line section 110 from which the first bit is extracted is
lengthened by 4.284 m, section 120 corresponding to the second is
Iengthened by 2.856 m, section 130 corresponding to the third is
lengthened by 1.428 m, while the fibre line section 140 relating to the
fourth bit is not lengthened. The said optical fibre sections 110,120,130
and 140 are wound onto a reel having a sufficiently large diameter so as
not to cause losses due to excessive curvature.
Having thus delayed the first digital optical signal 1 for a time
equivalent to three times the bit time, the second digital optical signal 2
for a time equivalent to twice the bit time, the third digital optical signal 3
for a time equivalent to once the bit time and not having delayed the
fourth digital optical signal 4, output from the said four optical fibre line
sections 110,120,130 and 140, are simultaneously found the first, the
second, the third and the fourth bit of the said temporal series 7a to be
converted, corresponding, respectively, to the said four optical fibre line
sections 110,120,130 and 140 (Figure 7).
Figure 7 illustrates the conversion of a four bit temporal series 7a to a
four bit spatial figure 7d which carries the same information as 7a. The
procedure necessitates:
- cloning of the temporal series 7a in four identical temporal series,
indicated all together by 7b;
- delaying, by means of delay lines 110,120,130 and 140, of temporal
series 7b, in relation to each other, by a multiple of the bit time, thus
creating time-space figure 7c;
- appropriately selecting, from the time-space figure 7c, the figure 7d by
means of suitable switching modules 2000, 4000.
As shown in Figure 6, suitable means 101,102,103 and 104 control
and, if necessary, vary the state of polarization of the said four digital
CA 02241041 1998-06-19
- 15-
optical signals 1, 2, 3 and 4 output from the said four optical fibre line
sections 1 10, 120, 130 and 140.
In a preferred embodiment, the said means 101, 102, 103 and 104 of
controlling and, as if necessary, varying the state of polarization of the
said four signals 1, 2, 3 and 4 consist of four polarization controllers
which in turn consist, for example, of four pairs of fibre optic polarization
rotators. Preferably each pair of polarization rotators consists of two
discs, made, for example, of metal and/or plastic, of suitable diameter,
onto which turns of optical fibre coils are wound. The said turns induce a
birefringence in the normal plane of the fibre in the direction of
propagation of the said signals 1, 2, 3 and 4. By suitably selecting the
diameter size of the said coils it is possible to create a ~/4 plate by one
turn of an optical fibre line and a ~2 plate by 2 turns. Given that it is
possible to obtain any state of polarization whatsoever through rotation of
a ~2 plate and a ~J4 plate, using the said means 101, 102, 103 and 104
it is possible to regulate the state of polarization of each of the four
signals 1, 2, 3 and 4 with a high degree of accuracy.
Means of collimation 5 located downstream of the said means 101,
102, 103 and 104 capable of controlling and, as the case may be, varying
the polarization state of the said signals 1, 2, 3 and 4 act as an interface
between an optical fibre line section in guided propagation and a section
in free space (approximately 20 cm) in which a first 2000 and a second
4000 switching modules are located.
The said means 5 are capable of converting the said four digital
optical signals 1, 2, 3 and 4 in guided propagation to four digital optical
signals 111, 112, 113 and 114 in free space. Furthermore, the said
means 5 are capable of collimating these latter digital optical signals 111,
1 12, 1 13 and 1 14 for the entire section in free space and maintaining
them parallel and within the transversal dimensions of the said first and
second optical switching modules 2000 and 4000.
CA 0224l04l l998-06-l9
- 16-
As shown in Figure 2, fou r optical fibres 11,12,13 and 14 in output
from the said means 101,102,103 and 104 capable of controlling and, if
necessary, varying the state of polarization are arranged in a linear
geometry having essentially the minimum dimensions. Such linear
geometry can be obtained by removing the plastic cladding from the said
four optical fibres 11,12,13 and 14 and gluing the fibres thus uncovered
close to each other on a glass mounting.
Because the external diameter of the cladding of the monomode
fibres, at 1550 nm, is approximately 125 !lm, the overall dimension of an
array of fibres 1000 thus created is approximately 50011m. The gap
between the extreme outside signals 1 and 4 at output from the said set
1000 is consequently approximately 375 llm.
A grin-type lens 1001 having a pitch equivalent to, for example, 0.25
collimates the four digital optical signals 111,112,113 and 114 output
from the said array 1000. The functioning of the grin lens 1001 is based
on a radial variation of the index of refraction rather than on the curvature
of the lateral surfaces as in conventional lenses. The said grin lens 1001
is preferable to a conventional lens because it can be placed immediately
after the said fibres 11,12,13 and 14 of the said array 1000, thus
enabling gathering of all the said signals 111,112,113 and 114 output
from the said array 1000 before the said signals diverge excessively.
In the embodiment shown in Figure 2, a convex lens 1010 having a
focus of approximately 80 mm is placed approximately 8 cm away from
the said grin lens 1001 to correct divergence of the said four signals 111,
112,113 and 114 output from the said grin lens 1001. The said four
signals 111,112,113 and 114 input into the said grin lens 1001 are, in
fact, offset in relation to the axis of the said grin lens 1001 and
consequently when output from the same possess optimum collimation
but considerable divergence.
CA 02241041 1998-06-19
According to a variant, the means of collimation 5 as previously
described can be created using a set of microlenses, one for each of the
said four digital optical signals 1, 2, 3 and 4, as described below referring
to Figure 10.
At output from the said means of collimation 5, the said four signals
1 1 1, 1 12, 1 13 and 1 14 are essentially well collimated and parallel over
the entire section in free space. In particular, the two extreme outside
signals 1 1 1 and 1 14 are separated by approximately 3.8 mm.
As shown in Figure 3, the said first switching module 2000 preferably
consists of a first indium-doped cadmium tellurium (CdTe:ln) monocrystal
200 and a first polarization analyzer 20.
The said first monocrystal 200, having dimensions of approximately 5
x 5 x 15 mm, is placed in a plexiglass mounting provided with openings
corresponding to the input and output faces of the said signals 1 1 1, 1 12,
113 and 114 and equipped internally with electrodes for application of
voltage.
In this embodiment the signal is propagated perpendicular to the
electrical field applied.
The said first polarization analyzer 20 preferably consists of a
polarization divider cube.
The state of polarization, represented in Figure 3 by the arrows 10, of
the said optical signals 111, 1 12, 1 13 and 114 input into the said first
module 2000 is adjusted, as stated above, by the said means 101, 102,
103 and 104 capable of controlling and if necessa~ changing the state of
polarization. This adjustment is carried out in such a way that the said
optical signals 1 1 1, 1 12, 1 13 and 1 14 are polarized in linear form at 45~
in relation to the axes of birifringence induced in the said first monocrystal
200. The said birifringence is induced by applying a voltage to the said
first monocrystal 200 capable of rotating, by electro-optical effect, the
plane of polarization 10 of the said signals 1 1 1, 1 12, 1 13 and 1 14
CA 0224l04l l998-06-l9
through an angle of 90~ so that the plane of polarization orthogonal to the
input plar~e 10 is obtained as represented by the arrows 100.
In Figure 3a, the said first polariz~tion analyzer 20 is oriented so as to
block the said signals 1 1 1, 112, 113 and 114 output from the said first
monocrystal 200. In this condition (OFF condition) the said first module
2000 is closed and does not allow transmission of the said signals 111,
112, 113and 114.
By illuminating the first monocrystal 200 with a first optical control
beam 320 at 1064 nm (the wavelength at which the CdTe:ln shows a
photoconductivity peak), the counterfield generated by the
photogenerated carriers inhibits the electro-optical effect. Consequently,
in the said first monocrystal 200 the plane of polarization 10 of the said
signals 11 1, 112, 1 13 and 114 is no longer rotated and the said analyzer
20 allows the said signals to pass through as shown in Figure 3b (ON
condition).
The said first monocrystal 200 is characterised by two response times:
- a time ton relating to the process of photogeneration of the carriers and
creation of the counterfield, and
- a time toff relating, on the other hand, to the process of re-combination
of such charges and restoration of the initial conditions.
Experiments have demonstrated that the time ton is very fast (typically
a few ns) and tends to follow exactly the rise time of the said first optical
control pulse 320 when these possess a sufficiently high power density
(approximately 105 W/cm2 higher). Time to~f on the other hand is
considerably slower (typically a few lus) and is strictly correlated to the
spatial distribution of the said first optical control beam 320.
Consequently, only with the said first monocrystal 200 is it not possible to
create a device having a response speed of the order of nanoseconds.
An important characteristic of the said first monocrystal 200 is the fact
that it is essentially transparent (possesses an absorption coefficient
. . .. ..
CA 0224l04l l998-06-l9
- 19-
approximately lower than 0.2 cm~1) to wavelengths above approximately
1250 nm. The optical device according to the invention can therefore be
used in optical communication systems in second and third window.
The description and the comments made concerning this first optical
switching module 2000 also apply for the said second switching module
4000, consisting of a second monocrystal 400 and a second polarization
analyzer 40.
As shown in Figure 4, the said first and second switching modules
2000 and 4000 are arranged in series, with the respective polarization
analyzers 20 and 40 cross-oriented, and they are controlled by a first and
a second beam of optical control pulses 320 and 340 delayed in relation
to each other by a predetermined time.
Initially the plane of polarization of the said signals 111,112,113 and
114 input to the said first monocrystal 200 is rotated by electro-optical
effect and, when output from the latter, the said signals 111,112,113
and 114 are blocked by the said first polarization analyzer 20 (1 st module
in OFF condition). On arrival of the said first optical control pulse 320 the
electro-optical effect of the said first monocrystal 200 is inhibited (1st
module in ON condition). The said signals 111,112,113 and 114 can
therefore pass through the said first polarization analyzer 20 and enter
the said second monocrystal 400, where their plane of polarization
undergoes a 90~ rotation due to an electro-optical effect. Because the
said second polarization analyzer 40 is oriented perpendicular in relation
to polarization analyzer 20 of the said first module 2000, the OFF
condition of the second module 4000 enables the said signals 111,112,
113 and 114 to be transmitted with a response time equivalent to ton and
to leave the said modules 2000 and 4000. The said optical switching
modules would stay in these conditions, i.e. open, for the entire
extinguished time toff of the said first monocrystal 200.
CA 02241041 1998-06-19
- 20 -
To obtain a total opening time of the said first and second modules
2000 and 4000 in series of iess than toff~ suitable guiding means 30 lead
the said second control pulse 310 to the said second monocrystal 400
after a predetermined time interval of tw less than toff. The second module
4000 is then activated (2nd module in ON condition), the said signals
111,112,113 and 114 no longer undergo rotation of the said plane of
polarization and are thus cut off by the said second analyzer 40, with a
~esponse time of ton.
The total opening time of the said first and second modules 2000 and
4000 in series can be adjusted by selecting the delay tw between the said
first and second optical control pulse 320 and 310. This delay tw must not
be greater than the extinction time toff typical of the said monocrystals
200 and 400. It must furthermore not be less than ton to enable the said
first monocrystal 200 to respond to the said first control pulse 320.
In Figures 3 and 4, the case in which illumination of the said
monocrystals 200 and 400 by the said optical control beams 320 and 310
is crosswise to the direction of propagation of the said signals 111, 112,
113 and 114 is illustrated, whilst Figure 5 represents the case in which
the said first and second optical control beams 320 and 310 arrive in the
said monocrystals 200 and 400 in collinear fashion in relation to the said
signals 111,112,113 and 114.
As shown in Figure 5, a dichroic mirror 50 is placed between the said
first and second modules 2000 and 4000; the said mirror is transparent
to the wavelength of the said signals 111,112, 113 and 11 ~ and reflects
the wavelength of the said first and second optical control beam 320 and
310. Typically, the wavelength of the said signals 111, 112, 113 and 114
is approximately 1550 nm and that of the said first and second optical
control beams is approximately 1064 nm.
In the case, not illustrated, in which the wavelength of the said digital
optical signals 111, 112, 113 and 114 is essentially the same as that of
CA 02241041 1998-06-19
the said first and second optical control beams 320 and 310, the said
dichroic mirror 5 can be substituted, e.g. by a beam splitter capable of
transmitting approximately half the power of the said digital optical
signals 111,112,113 and 114 and of the said first and second optical
control beams 320 and 310 and reflecting approximately the other half of
the same.
The said dichroic mirror 50 is essentially inclined at an angle of 45~ in
relation to the propagation direction of both the said 111, 112,113 and
114 and the said first and second optical control beams 320 and 310
(Figure 5). The said first control beam 320, coming from a direction
essentially perpendicular to that of the said signals 111,112, 113 and
114, after reflection on the said dichroic mirror 50 is incident upon the
said first monocrystal 200 in a direction collinear to, but opposite to, that
of the said signals 111,112, 113 and 114 (counterpropagating
illumination).
In turn, the said second control beam 310, also coming from a
direction essentially perpendicular to that of the said signals,1, 2, 3 and
4, after reflection on the said dichroic mirror 50, is incident upon the said
second monocrystal 400 in a direction which is collinear to, and is the
same direction as, the said signals 111,112,113 and 114 (co-
propagating illumination).
Both illumination modes, counterpropagating and co-propagating, are
equally effective for the purposes of achieving switching.
As shown in Figure 6, the said first and second optical pulse control
beams 320 and 310 having wavelength of approximately 1064 nm are
preferably generated by Nd:YAG Q-switched laser 300. The FWHM
duration of the said pulses is approximately 5 ns. The optical control
beam 3000 output from the said laser is divided into two equal parts by
means of a beam splitter 33 and coupled to a first 32 and a second 31
multimode optical power fibre of which the core diameter is
CA 02241041 1998-06-19
- 22 -
approximately 600,um. The lengths of the said first and second optical
fibres 32 and 31 are different from each other in such a way that the said
first and second optical pulses 320 and 310 are incident upon the said
first 200 and, respectively the said second 400 monocrystals at different
time instants. The delay tw between the said first and second optical
control pulses 320 and 310 determines, as described above, the total
opening time of the said first and second optical series switching
modules 2000 and 4000.
In this case, the said total opening time will be equal to the bit time,
whilst the interval between two successive selections will be equal to the
duration of the bit series which is being converted, this, in the example
given in Figure 6, being four times the bit time.
Two flat-to-convex lenses 15 and 25 with 30 mm focus collimate the
said first and second optical control beams 320 and 310 output from the
said first and second optical fibres 32 and 31 so that the diameter of the
beams (spots) incident upon the said monocrystals 200 and 400 is
approximately 7 mm. In this way the said first and second optical control
beams 320 and 310 completely illuminate the input surface of the said
monocrystals 200 and 400 which is approximately 5 x 5 mm. In addition,
the said flat-to-convex lenses 15 and 25 guide the said first and second
optical control beams 320 and 310 in free space in such a way that they
are incident upon the said dichroic mirror 50 at an angle of approximately
45o
In the embodiment illustrated, the energy of the said first and second
optical control pulses 320 and 310 incident upon the said monocrystals
200 and 400 is approximately 1 mJ. With energy of this order, the time ton
of the said monocrystals 200 and 400 is the same as the rise time
(corresponding to the time taken by the pulse to increase from
approximately 10% to approximately 90% of the maximum intensity
value) of the said first and second control pulses (approximately 3 ns).
CA 02241041 1998-06-19
In the case of CdTe:ln monocrystals, it has been observed in
experiments that in order to cause the said first and second switching
modules 2000 and 4000 to switch in series, the energy of the said first
and second optical control pulses 320 and 310 only needs to be
equivalent to or greater than 350 ,uJ.
The four bits output from the first and second optical switching
modules 2000 and 4000 which form the said spatial figure 7d are thus
ready to undergo subsequent processing operations in parallel and in
free space.
The said subsequent processing operations are carried out using
means 7000, 71, 72, 73 and 74 to eliminate one bit of the said spatial
figure 7d of four bits and means 81, 82, 83, 84, 80 and 8 to insert one bit
into the said spatial figure 7d.
In particular, Figure 1 illustrates an alternative in which the third bit is
eliminated and the second bit is inserted.
The said means of eliminating the third bit of the said spatial figure 7d
comprise a third optical switching module 7000, placed downstream of
the said means of collimation 5 and upstream of the said first and second
switching modules 2000 and 4000, and means (not illustrated) of
supplying the third, 73, of four optical elimination signals 71, 72, 73 and
74.
The descriptions and comments made concerning the said first optical
switching module 2000 also apply to the said third switching module
7000, which consists of a third indium-doped cadmium tellurium
(CdTe:ln) monocrystal 700 and a third polarization analyzer 70.
In the embodiment as shown in Figure 1, the said third polarization
analyzer 70 is oriented so as to allow the said optical signals 111, 112,
113 and 114 output from the said third monocrystal 700 to pass through
when the said elimination signals 71, 72, 73 and 74 are absent (3rd
module: OFF condition), and to block them when the said optical
CA 02241041 1998-06-19
- 24 -
elimination signals 71, 72, 73 and 74 illuminate the said third monocrystal
700 (3rd module: ON condition).
The said third digital optical signal 113, relative to the said third bit to
be eliminated, is thus blocked, illuminating the region of the said third
monocrystal 700 where the said third digital optical signal 113 is
propagated by means of the said third optical elimination signal 73.
In the case illustrated of a CdTe:ln monocrystal and a power density of
the said elimination signals 71,72, 73 and 74 of less than approximately
0.5 mW/mm2, it has been found that when the said elimination signals
71, 72, and 74 are situated more than approximately 0.5 mm away from
the said third digital optical signal 113, elimination of the said third bit of
the said spatial figure 7d does not interfere with propagation of the other
digital optical signals 111,112 and 114 corresponding to the other bits of
the said spatial figure 7d.
The dimensions of the said third monocrystal 700, i.e. 5 x 5 x 15 mm,
are such that they can fully contain the said optical signals 111,112,113
and 114, with a gap of at least 0.5 mm between each other.
The said third optical elimination signal 73 has a wavelength of
approximately 1064 nm and is generated by a laser (not illustrated) made
of neodymium-doped fibre and pumped by an approximately 810 nm
laser diode which emits a continuous signal.
In a preferred embodiment, four on/off switches (not shown) select the
said four optical elimination signals 71, 72, 73 and 74 dependent on the
bit(s) to be eliminated in the said spatial figure 7d. The said on/off
switches can be controlled either electrically, for example, by using a
common integrated electrical-optical modulator, or optically, for example,
by using an optical switching module similar to those described above.
The said elimination signals 71, 72, 73 and 74 and the said signals
111,112,113 and 114 relative to the said four bits of the said spatial
figure 7d are coupled in fibre, before the said means 5 of collimation
CA 02241041 1998-06-19
- 25 -
previously described, by means of four 2 x 2 fusion couplers to combine
a first signal at about 1550 nm with a second signal at about 1064 nm.
Because the said elimination signals 71, 72, 73 and 74 are output,
respectively, from the same fibres 11,12,13 and 14 as the said digital
optical signals 111, 112,113 and 114 and are subjected to the same
collimation process, they are perfectly aligned with and overlap the said
digital optical signals 111,112,113 and 114.
In the embodiment shown in Figure 1, in which the said elimination
signals 71, 72, 73 and 74 and the said digital optical signals 111,112,
113 and 114 are coupled in guided propagation before the said means of
collimation 5, the said third switching module 7000 is arranged, as
previously mentioned, downstream of the said means of collimation 5
and upstream of the said first and second switching modules 2000 and
4000.
In another embodiment, in which the coupling between the said
elimination signals 71, 72, 73 and 74 and the said digital optical signals
111, 112, 113 and 114 is carried out in a different way, for example in
free space, the said third switching module 7000 can also be placed
downstream of the said first and second switching modules 2000 and
4000, it being possible at this point to insert the said elimination signals
71, 72, 73 and 74 at any given point of the section in free space. The two
solutions are completely equivalent.
Downstream of the said third module 700 and upstream of the said
first module 2000, an interference filter (not shown) is located; the said
filter is transparent to the wavelength of approximately 1550 nm of the
said digital optical signals 111,112, 113 and 114 and reflects the
wavelength of approximately 1064 nm of the said elimination signals 71,
72, 73 and 74. The said interference filter thus prevents the said
elimination signals 71, 72, 73 and 74 relative to the said third switching
module 7000 and the said counterpropagating first optical control beam
CA 02241041 1998-06-19
- 26 -
320, relative to the said first switching module 2000, from reaching the
modules which do not concern them.
Using the optical device illustrated in Figure 1, an experimental
operation was carried out to eliminate the second bit of a temporal series
"1111" of bits, in RZ format and at approximately 140 MbiVs, input into
the device. In Figure 11, the recordings of the temporal series "1111"
input (Figure 11a) into and the modified "1011 " series (Figure 11b) output
from the device, obtained by means of suitable photodetectors and an
oscilloscope with passband of approximately 1 GHz, are reported.
The said means of inserting the second bit in the said spatial figure 7d
comprise means (not shown) of supplying the second, 82, of four
insertion optical signals 81, 82, 83 and 84, means 80 of collimation of the
said insertion signals 81, 82, 83 and 84 and a beam splitter 8.
The said means of supplying the second, 82, of four optical insertion
signals 81, 82, 83 and 84 comprise, preferably, a laser diode DFB (not
shown) which supplies a continuous signal having a wavelength of
approximately 1550 nm and an output power into fibre of approximately 1
mW.
For example, the said means 80 of collimation of the said insertion
signals 81, 82, 83 and 84 comprise a grin lens to collimate the said
signals 81, 82, 83 and 84 in free space and guide them in such a way
that they are incident upon the said beam splitter 8 at an angle of
approximately 45~ and are then propagated in the spatial position relative
to the respective said digital optical signals 111,112, 113 and 114, in
other words they overlap one of the said digital optical signals 111,112,
113and 114.
The said beam splitter 8 is capable of transmitting approximately half
the power of the said digital optical signals 111,112,113 and 114 and of
the said insertion signals 81, 82, 83 and 84 and reflecting approximately
the other half of the said signals.
CA 02241041 1998-06-19
The said beam splitter 8, placed downstream of the said third
switching module 7000 and upstream of the said second switching
module 2000, is essentially inclined 45~ in relation to the direction of the
said digital optical signals 111,112,113 and 114 and to that of the said
insertion signals 81, 82, 83 and 84, the direction of the said digital optical
signals 111,112,113 and 114 being essentially orthogonal to the
direction of the said insertion signals 81, 82, 83 and 84.
Means (not shown) such as, for example, a pair of polarization rotators
as described earlier, placed upstream of the said means 8 of collimation,
control and, if necessary, change the state of polarization of the said
second insertion signal 82 in such a way that it is polarized in linear form
at 45~ in relation to the axis of induced birifringence, in the said first and
second switching modules 2000 and 4000, by the said voltage applied to
the same.
In the embodiment according to Figure 1, asynchronous insertion of
the said second bit into the said spatial figure 7d takes place at the
moment when the said first and second switching modules 2000 and
4000 in series select, in parallel and in free space, a predetermined bit
from each of the said four digital optical signals 111,112,113 and 114
and consequently at the moment when they convert the said temporal
series 7a of four bits to the said spatial figure 7d. Consequently, the
synchronism of the said insertion is guaranteed by the synchronism of
the said first and second switching modules 2000 and 4000 in series in
relation to the said four digital optical signals 111,112,113 and 114.
According to another embodiment, insertion of the said second bit can
be carried out by inserting the said bit into the correct spatial position in
synchronism with all the said digital optical signals 111,112,113 and 114
constituting the said spatial figure 7d of four bits. In comparison with this
latter embodiment, however, the embodiment illustrated in Figure 1 does
not require any timing between the said digital optical signals 111,112,
CA 02241041 1998-06-19
- 28 -
113, 114 and the said bit to be inserted. Furthermore, the embodiment
shown in Figure 1 is far simpler to make, since, for insertion of the said
bit, it is possible to use the same first and second switching modules
2000 and 4000 in series which convert the said four bit temporal series
7a into the said spatial figure 7d of four bits.
The temporal duration of the said second inserted bit will be equal to
the total opening time of the said first and second switching modules
2000 and 4000. The said duration will, preferably, be equal to or less
than the bit time in order to prevent partial overlapping of the said four
bits (intersymbolic interference) in a subsequent conversion of the said
spatial figure 7d to a temporal series 7e of four bits of a processed digital
optical signal 900 output from the device.
Preferably the said means of insertion of the second bit also comprise
a system of four on/off switches (not shown) capable of blocking or
transmitting the said optical insertion signals 81, 82, 83 and 84
depending on the bit(s) to be inserted into the said spatial figure 7d. The
said on-off switches can be controlled either electrically, for example by
using a common integrated electrical-optical modulator, or optically by
using an optical switching device similar to those previously illustrated.
By means of the optical device illustrated in Figure 1, an experimental
operation has been carried out for asynchronous insertion of the third bit
in a temporal series "1101" of bits, in RZ format and at approximately 140
Mbit/s, input into the device. In Figure 12, the recordings of the temporal
series "1101" input into (Figure 12a) and the modified temporal series
"1111" (Figure 12b) output from the device, obtained by means of
suitable photodetectors and an oscilloscope with passband of
approximately 1 GHz, are reported.
In the case under consideration, the third inserted bit was found to be
lower and wider than the other bits. Its narrower amplitude is due to the
fact that insertion of the third bit was carried out by a continuous laser
CA 02241041 1998-06-19
- 29 -
signal having, after the beam splitter 8, lower power (approximately 1
mW) than the power of the digital optical signals output from the optical
amplifier, which had, after the beam splitter 8, a power of approximately
1.5 mW each, and also to inadequate alignment of the third optical
insertion signal with the spatial position relative to the third bit. The
greater width of the third inserted bit in comparison with the others is, on
the other hand, due to the fact that, the said optical control beams 320
and 310 having a rise time of approximately 3 ns, the total opening time
of the said first and second optical switching modules 2000 and 4000,
was greater than the physical duration of the bits in RZ format at
approximately 140 Mbit/s.
These flaws can nevertheless obviously be corrected by using a
continuous 1.5 mW laser signal, correctly aligning the optical insertion
signal with the spatial position relative to the third bit and using control
beams with a rise time such that the total opening time of the said first
and second optical switching modules 2000 and 4000 is equal to the
physical duration of the bit.
The embodiment illustrated in Figure 1 can also comprise means of
changing the form of the said four bits of the said spatial figure 7d.
The said means convert the said "Non Return to Zero" (NRZ) coded
bits of the said spatial figure 7d to "Return to Zero" (RZ) coded bits.
According to a preferred embodiment, the said means halve the ratio
(duty cycle) between the physical duration of the said digital optical
signals 1 1 1, 1 12, 1 13 and 1 14, corresponding to 1 bit, and the bit time. Inthis way, since in the NRZ format the said ratio is approximately 1, the
said means enable a ratio of approximately 0.5 to be obtained.
The said halving is carried out by adjusting the total opening time of
the said first and second optical switching modules 2000 and 4000 in
series so that they stay open for a time equal to approximately half the bit
time.
CA 02241041 1998-06-19
- 30 -
By adjusting the said total opening time in this way, the said first and
second optical switching modules 2000 and 4000 are capable of carrying
out both the said transformation of the said four bit temporal series 7a
into the said spatial figure 7d and modification of the duty cycle of the
said bits of the said spatial figure 7d.
Determination of total opening time has to be carried out very carefully
because this determines the form the selected bit shall take. This being
so, it is necessary to consider the total response time during opening and
closing of the said first and second switching modules 2000 and 4000 in
series, i.e. ton, which, as stated above, is the same as the rise time of the
said first and second optical control pulses 320 and 310 (approximately 3
ns in the embodiment illustrated).
The said first and second switching modules 2000 and 4000 of the
embodiment illustrated in Figure 1 cannot, therefore, stay open for a total
time of less than approximately 6 ns.
The said conversion is therefore feasible, for example, in the case of a
temporal series 7a with a bit rate of 70 MbiVs (corresponding to a bit time
of approximately 14.28 ns) which, therefore, needs a total opening time
of the said first and second switching modules 2000 and 4000 in series
of approximately 7.14 ns.
The said total opening time is adjusted by changing the lengths of the
said first and second optical fibres 32 and 31 to take into account the
total response time ton during opening and closing of the said first and
second switching modules 2000 and 4000 in series.
For example, in the case described above of a bit rate of 70 MbiVs, in
order to obtain a total opening time of the said first and second switching
modules 2000 and 4000 of approximately 7.14 ns, the lengths of the said
first and second optical fibres 32 and 31 are selected so that the said first
and second optical control pulses 320 and 321 are incident upon the said
first and respectively the said second modules 2000 and 4000 with a
CA 02241041 1998-06-19
- 31 -
delay of approximately 4 ns. This delay is obtained by lengthening the
said second optical fibre 31 by approximately 80 cm in relation to the said
first optical fibre 32.
In such a case, the operation of conversion of a temporal series of four
bits "1111" input into the device, at approximately 70 Mbit/s and NRZ
coded, into four corresponding RZ coded bits with a duty cycle of
approximately 0.5 was carried out in an experiment. In Figure 13 the
recordings are reported of the input temporal series "1111" (Figure 13a),
NRZ coded, and the corresponding temporal series (Figure 13b) output
from the device, RZ coded, obtained by means of suitable
photodetectors and an oscilloscope with passband of approximately 1
GHz.
The device illustrated in Figure 1 enables the same optical switching
modules 2000 and 4000 to be used both for transforming the said
temporal series of bits 7a into the said spatial figure 7d, and for inserting
the said second bit and also for modifying the form of the said bits of the
said spatial figure 7d. This therefore enables creation of a device with
extremely simple and compact architecture benefiting from all the
capacities of the parallel optical structures in free space.
As illustrated in Figure 1, means of output transform the said four bit
spatial figure 7d, processed if necessary, into a temporal series 7e of a
processed digital optical output signal 900.
The said means of output comprise optical means 9 of focusing the
said four bits, in free space, in four optical fibre lines (sections) 910, 920,
930 and 940 and a coupler 90 to convey the said bits output from the
said lines 910, 920, 930 and 940 to the said processed digital optical
output signal 900.
As illustrated in Figure 2, the said optical means 9 of focusing the said
four bits of the said spatial figure have, preferably, a symmetrical
structure in relation to the said means of collimation 5. The said means 9
CA 02241041 1998-06-19
comprise a first lens 9010, with focus of approximately 80 mm, for
focusing the said digital optical signals 111,112,113 and 114 in free
space, corresponding to the said four bits, output from the said first and
second optical switching modules 2000 and 40900, on the input surface
of a grin lens 9001, as previously described, located on the focal plane of
the said lens 9010.
Emergent from the said grin lens 9001, the said digital optical signals
111,112,113 and 114 are highly focused and are thus easily guided in
guided propagation by means of an array 9000 of the said four optical
fibre lines 910, 920, 930 and 940 arranged in a linear geometry similar to
that previously described for the said means of collimation 5.
This linear geometry can easily be reproduced with an extremely high
degree of accuracy. The precision with which the said arrays 1000 and
9000 of fibres have to be created depends on the diameter of the core of
the monomode fibres, which at approximately 1500 nm is approximately
9,um. Consequently, if one of the fibres 910, 920, 930 and 940 of the
output array 9000 is offset in relation to the corresponding fibre 11,12,
13 or 14 in the input array 1000 by a quantity greater than the said
dimension, the bit which is propagated in the said fibre is badly coupled
and consequently the information relative to one of the bits can be lost.
The precision of the linear geometry as shown in Figure 2 is solely due
to the external diameter of the cladding of the 1550 nm monomode
fibres, which is approximately 125 ,um with tolerances of less than
approximately 1 ,um. This guarantees optimum coupling in fibre of the
said digital optical signals 111,112,113 and 114.
According to a variant, the said four optical fibres 11,12,13 and 14
input into and fibres 910, 920, 920 and 940 output from the device can
be arranged in a two-dimensional geometry similar to that illustrated in
Figure 10 in the case of eight digital optical signals.
CA 02241041 1998-06-19
The said four optical fibre lines 910, 920, 930 and 940 in different
lengths delay, in relation to each other, the said four bits of the said
spatial figure 7d by a multiple time quantity of the bit time which, at
approximately 140 Mbit/s, is approximately 7.12 ns. That is, the lengths
of the said lines 910, 920, 930 and 940 differfrom each others by
multiple quantities of approximately 1.428 m, the necessary length to
operate a delay of approximately 7.12 ns.
To construct the said four bit output temporal series, if necessary
processed in relation to the said temporal series 7a input into the device
according to the invention, the said four bits are inversely delayed in
relation to the input bits. For example, the first bit, which is delayed in
input by three times the bit time, is not delayed in output, whilst the fourth
bit, which is not delayed in input, is delayed in output by three times the
bit time.
Consequently, the fibre section 910 from which the first bit is extracted
is not lengthened, fibre section 920 corresponding to the second bit is
lengthened by 1.428 m, fibre section 930 corresponding to the third bit is
lengthened by 2.856 m and section corresponding to the fourth bit is
lengthened by 4.284 m.
The said lines 910, 920, 930 and 940 are also wound onto a reel
having sufficiently large diameter to avoid causing losses due to
excessive curvature.
The said 4 x 1 coupler 90 conveys in a single optical fibre 990 the said
four bits emergent from the said lines 910, 920, 930 and 940 so as to
obtain the said digital optical output signal 900 comprising the said
temporal series of bits 7e, possibly processed in relation to the said input
temporal series 7a.
Preferably, upstream of the said grin lens 9001, a differential filter (not
shown) is arranged which is transparent to the wavelength of
approximately 1 550 nm of the said digital optical signals 111, 11 2, 11 3
CA 02241041 1998-06-19
- 34 -
and 114, corresponding to the said four bits, and which reflects the
wavelength of approximately 1064 nm of the said second co-propagating
optical control beam 310. In this way the said filter prevents the said
second optical control beam 310 from being coupled in fibre and
overlapping the said four bits of the said spatial figure.
According to another embodiment, the said optical fibre lines 910, 920,
930 and 940 can also delay, in relation to each other, the said four digital
optical signals 111,112,113 and 114 by multiple quantities of a time
greater or lesser than the bit time so as to change the duration of the said
four bits of the said spatial figure 7d and, consequently, the bit rate.
More particularly, the dimensions of the said lines 910, 920, 930 and
940 can be adjusted according to the new bit time reguired.
For example, to increase from a bit time of approximately 7.14 ns to a
time of approximately 9.64 ns, pieces of optical fibre are added in
multiples of 50 cm to the said lines 910, 920, 930 and 940, the
dimensions of which are suitable for working at approximately 140 Mbit/s.
Accordingly, line 940, corresponding to the fourth bit of the said temporal
series 7e to be reconstructed, is lengthened by approximately 150 cm
(50 cm x 3) and is therefore increased from approximately 4.248 m to
approximately 5.748 m, the line corresponding to the third bit is
lengthened by approximately 100 cm and increased from approximately
2.856 m to approximately 3.856 m, the line corresponding to the second
bit is lengthened by approximately 50 cm, increasing it from
approximately 1.428 m to approximately 1.928 m, whilst the line
corresponding to the first bit does not need any added delays.
The said temporal series 7e output from the device according to the
invention has, in this case, a bit rate of approximately 103.7 Mbit/s (the
inverse of the new bit time of approximately 9.64 ns), which is less than
its bit rate at the time of input.
CA 02241041 1998-06-19
- 35 -
In such a case, an experimental operation was carried out to change
the duration of the bit time from approximately 7.14 ns to approximately
9.64 ns in the case of a four bit temporal series "1111", in RZ format and
with a duty cycle of approximately 0.5, during input into the device.
Figure 14 shows the recordings for the temporal series input into the
device (Figure 1 4a) at approximately 140 Mbit/s, and for the
corresponding tFigure 1 4b) temporal series output from the device, at
approximately 103.7 Mbit/s, the said recordings having been obtained by
means of suitable photodetectors and an oscilloscope with passband of
approximately 1 GHz. Although the duration of the bit time output from
the device was changed, the physical duration of the bit remained
unchanged and therefore the duty cycle of the bits output from the device
was reduced (approximately 0.37).
Other experiments have demonstrated the possibility of varying the
length of the said lines 910, 920, 930 and 940 continuously so as to be
able to continuously change the bit rate of the said bits of the said spatial
figure 7d.
This is feasible due to the elasticity of optical fibres, which can be
lengthened, in elastic conditions, by at least approximately 4%.
By means of a device capable of lengthening the fibres, consisting, for
example, of two rows of pulleys onto which 10 m of fibre are rolled and a
stepper electric motor, fibre lines 910, 920, 930 and 940 can be
lengthened by as much as approximately 40 cm, thus causing a
maximum extra delay of approximately 2 ns between the said bits of the
said spatial figure 7d. By using three devices of this type arranged in
series in relation to the said lines 920, 930 and 940 (line 910 which
corresponds to the first bit needs no extra delays), it is possible to
lengthen the bit time of a four bit word by a quantity less than or equal to
approximately 2/3 ns (in this way the maximum delay corresponding to
the fourth bit will be approximately 2 ns). This quantity is not very
CA 02241041 1998-06-19
- 36 -
significant at low bit rates but becomes considerably so at high bit rates.
In fact, in the case of a bit time of approximately 1 ns (equal to a bit rate
of approximately 1 GbiVs) it is possible with the said devices to
continuously lengthen the bit time to approximately 1.67 ms (equal to a
bit rate of approximately 600 MbiVs).
In a preferred embodiment, the device according to the invention also
comprises means of execution of algebraic operations (Figure 8).
In this case, for example, a fourth CdTe:ln monocrystal 1200 is
arranged in output from the said means (first and second switching
modules 2000 and 4000) capable of transforming the said temporal
series 7a of four bits into the said spatial figure 7d, the said CdTe:ln
monocrystal 1200 enabling actuation of algebraic operations by optically
controlling the state of polarization, represented in Figure 8 by an arrow,
of the said four digital optical signals 111,112,113 and 114
corresponding to the said four bits of the said spatial figure 7d.
The description and comments made above concerning the aforesaid
first monocrystal 200 also apply to the said fourth monocrystal 1200.
The said fourth monocrystal 1200, in fact, under the action of a
continuous voltage Vp and when a first optical beam enabling the said
algebraic operations is absent, rotates the plane of polarization of the
said four digital optical signals 111,112,113 and 114 by approximately
90~ or leaves it unchanged when it is illuminated by the said first enabling
beam.
A first polarization separator 1210 is arranged downstream of the said
fourth monocrystal 1200. The said first polarization separator 1210 is
oriented in such a way that it is transparent to the said four digital optical
signals 111,112,113 and 114 when the said first optical beam enabling
the said algebraic operations is absent and so that it deflects them in a
direction essentially orthogonal to the direction of incidence when the
said first enabling beam illuminates the said fourth monocrystal 1200.
CA 02241041 1998-06-19
- 37 -
In this way, as illustrated in Figure 8, the said four digital optical signals
111,112,113 and 114 output from the said fourth monocrystal 1200
continue in their direction of propagation when the said optical enabling
beam is absent, i.e. when no operation has to be executed. The said
optical signals 111,112,113 and 114 are, on the other hand, deflected in
a direction essentially orthogonal to the direction of propagation when the
said first enabling beam illuminates the said fourth monocrystal 1200, i.e.
when an algebraic operation has to be carried out.
A fifth CdTe:ln monocrystal 1400 conveys the said optical signals 111,
112,113 and 114 deflected by the said first polarization separator 1210
according to the operation to be carried out.
This fifth monocrystal 1400 too, in fact, underthe action of a
continuous voltage Vp, rotates the plane of polarization of the said four
digital optical signals 111,112,113 and 114 by approximately 90~ when
a second optical beam enabling the said algebraic operations is absent
and leaves it unchanged when it is illuminated by the said second
enabling beam.
Downstream of the said fifth monocrystal 1400, a second polarization
separator 1230 is arranged, oriented in such a way that it is transparent
to the said four digital optical signals 111,112,113 and 114 when the
said second enabling beam is absent, and such that it deflects them in a
direction essentially orthogonal to the direction of incidence, when the
said second enabling beam illuminates the said fifth monocrystal 1400.
In the embodiment according to Figure 8, therefore, when the said
second enabling beam is absent, a first and second mirror 1410 and
1420 reflect the said digital optical signals 111,112,113 and 114 output
from the said second polarization separator 1230 and guide them onto a
50/50 beam splitter 1150, so as to translate the said spatial figure 7d of
four bits to the left and thus execute the operation of multiplication by
two, as represented by 77 in Figure 8. When, on the other hand, the said
CA 02241041 1998-06-19
second enabling beam illuminates the said fifth monocrystal 1400, a third
mirror 1430 reflects the said digital optical signals 111,112,113 and 114
output from the said second polarization separator 1230 in order to guide
them onto a third polarization separator 1220. The said third polarization
separator 1220, in turn, guides the said digital optical signals 111,112,
113 and 114 onto the said 50/50 beam splitter 1550 so as to translate the
said four bit spatial figure 7d to the right and thus execute the operation
of division by 2, as represented by 777 in Figure 8.
After processing, the said spatial figure 7d is then re-converted in
output into a temporal series 7e of bits of a processed digital optical
signal 900, as previously described.
In another embodiment, the device according to the invention
comprises means of execution of symmetry operations (Figure 9) on a
circular, eight bit spatial figure obtained, for example, by means 5,
illustrated in Figure 10, of collimation of eight digital optical signals outputfrom eight optical fibres 11,12,13,14,15,16,17 and 18.
Of course, according to a variant not illustrated, the said symmetry
operations can also be executed on a spatial figure of bits having a linear
geometry similar to that, for example, illustrated in Figure 2.
The said means of collimation 5 comprise a circular set 47 (Figure
1 Ob) of eight collimators 48 (Figure 1 Oa) in which the said circular set 47
has, for example, an external diameter of approximately 3.6 mm and
each of the said collimators 48 has a diameter of approximately 1 mm.
At one end of each of the said collimators 48, a spherical collimation
lens 46 is present, the diameter of which is approximately 600,um.
Coupling between each of the said eight optical fibres and each of the
said spherical lenses 46 is preferably carried out by means of a glass
microcapillary 49 with external diameter of 1 mm, internal hole 126,um (1
,um larger than the diameter of the 1550 nm monomode fibre cladding)
and two conical widened sections at the ends. The said two conical
CA 02241041 1998-06-19
- 39 -
widened sections allow, at one end, assistance with insertion of one of
the said eight fibres and, at the other end, housing and gluing-in of the
said spherical lens 46 (Figure 1 Oa).
The focus of the said spherical lens 46 is preferably positioned on the
surface of the said sphere in such a way that the said eight optical fibres
can be inserted and glued in contact with the said spherical lens 46 with
no need for special measurements to determine the best point for
collimation.
The said eight collimators 48 thus created are then inserted in a circle
between two metal tubes each having a predetermined diameter, one
inside and the other outside the circular set 47, so as to obtain an
essentially regular and symmetrical arrangement.
In this case, execution of the said symmetry operations is enabled by
a sixth CdTe:ln monocrystal 1300, placed downstream of suitable means
(not illustrated) capable of transforming a temporal series of eight bits
into a spatial figure.
As illustrated in Figure 9, the said sixth monocrystal 1300, under the
action of a continuous voltage Vp, rotates by approximately 90~ the plane
of polarization (represented in Figure 9 by an arrow) of the said eight
digital optical signals (corresponding to the said eight bits), when an
optical beam 1500 enabling the said symmetry operations is absent, and
leaves it unchanged when it is illuminated by the said enabling beam
1 500.
Downstream of the said sixth monocrystal 1300, a fourth polarization
separator 1310 is located, oriented in such a way that it is transparent to
the said eight digital optical signals when their plane of polarization is
rotated approximately 90~ by the said sixth monocrystal 1300, i.e. when
the said beam 1500 enabling the said symmetry operations is absent.
When, on the other hand, their plane of polarization is not rotated, i.e.
when the said enabling beam 1500 illuminates the said sixth monocrystal
CA 02241041 1998-06-19
- 40 -
1300, the said fourth polarization separator 1310 deflects them in a
direction essentially orthogonal to the direction of incidence.
In the absence, therefore, of the said beam 1500 enabling the said
symmetry operations, the said eight signals continue unchanged. When,
on the other hand, the said enabling beam 1500 illuminates the said sixth
monocrystal 1300, a ~2 plate 1390, a prism 1350, a ~4 plate 1370, and
a mirror 1330 process the said eight bit circular figure so as to execute
required symmetry operations such as, for example, symmetry in relation
to an axis of the bit configuration (Figure 9).
