NZ245766A

NZ245766A - Detecting chromatically dispersed fm optical signals

Info

Publication number: NZ245766A
Application number: NZ245766A
Authority: NZ
Inventors: Berthold Wedding
Original assignee: Alcatel Australia
Priority date: 1992-02-01
Filing date: 1993-01-27
Publication date: 1995-11-27
Also published as: DE4202863A1

Description

<div class="application article clearfix" id="description"> 2 24 5 7 6 6 Priority DatG(s): — Com plate Specification Filed: .23.'A.'..c0.s. Class: (f?)..I3L,\.<a.Q?r.Ois» 2 7 NOV 1995 ' Publication Date:.. P.O. Journal No: .130.^ NEW ZEALAND PATENTS ACT 1953 COMPLETE SPECIFICATION "AN OPTICAL SIGNAL DECODING ARRANGEMENT" WE, ALCATEL AUSTRALIA LIMITED, A Company of the State of New South Wales, of 280 Botany Road, Alexandria, New South Wales, 2015, Australia, hereby declare the invention for which we pray that a patent may be granted to us, and the method by which it is to be performed, to be particularly described in and by the following statement: 1 245766 I i This invention relates to optical transmission. Generally optical transmission systems have some dispersion at the operating wavelength. The invention will be described in the context of a frequency modulated digital system. A frequency modulated digital optical transmission system is known from: A.R. Chraplyvy et al: "8 Gbit/s FSK Modulation of DFB Lasers with Optical Demodulation", Electronics Letters, 2 March 1989, Vol. 25, No. 5, Pages 319 to 321. With the transmission of digital signals with high bit rates (in the Gigabit range), the chromatic dispersion (also known as material dispersion) displayed by the optical waveguide of the optical transmission system becomes a problem, insofar as it limits the path length over which a digital signal with a high bit rate can be transmitted. Nowadays, it is desirable on the one hand to have an operating wavelength in the vicinity of 1500 nm for the optical transmission because suitable fibre-optic amplifiers are available for such wavelengths, and on the other hand to use standard single-mode optical waveguides, since these have been installed extensively. Therefore, the problem of the chromatic dispersion of the optical waveguide must be solved by means other than through the choice of the operating wavelength or the choice of the optical waveguide type. The known system provides the following solution: the semiconductor laser at the transmitting end does not, as is otherwise usual, have its amplitude modulated by the digital signal being transmitted, but has the frequency of its 245766 output signal modulated. This modulation is called FSK modulation (FSK = Frequency Shift Keying), while the previously usual amplitude modulation is called ASK modulation (ASK = Amplitude Shift Keying). FSK modulation is achieved when the injection current of the semiconductor laser is modulated significantly less, that is with a significantly smaller modulation amplitude, than is required for the usual ASK modulation. With frequency modulation, the transmitted optical signal occupies a smaller spectrum range than would be the case with amplitude modulation, so that the chromatic dispersion is no longer such a disadvantage. At the receiving end, the known system has an optical interferometer which converts the frequency modulation of the received optical signal into amplitude modulation, and a direct-reception optical receiver which accepts the optical amplitude-modulated signal and recovers the transmitted digital signal from it. By a direct-reception optical receiver one usually understands an arrangement consisting of an optical detector, a pre-amplifier, an amplifier and a regenerator (this latter is sometimes also called a decision circuit), where the optical detector, together with the pre-amplifier and the amplifier, converts the time sequence of the amplitude of the received optical signal into a corresponding time sequence of an electrical signal, and the regenerator recovers the digital signal from the electrical signal. Such optical receivers are discussed, for example, in the book "Optical Fibers" by J. Geissler et al, Pergamon Press, Oxford, New York, Toronto, Sydney, Frankfurt, 1986, page 439 or in H. Hamano et al, Proc. ECOC '90, Amsterdam, pages 45 to 48. 3 245766 Because of its optical interferometer, the known system is more costly than earlier systems and therefore at a disadvantage from a cost point of view, even though it has greater capacity. Therefore, it is desirable to provide a simpler and thus more cost-effective solution for the problem of transmitting a digital signal with a high bit rate over an optical waveguide with dispersion at the operating wavelength. This specification describes a system for the optical transmission of a digital signal over an optical waveguide with dispersion at the operating wavelength, having an optical transmitter at the transmitting end of the system whose optical output signal is frequency modulated by the digital signal, and having an optical receiver at the receiving end which converts its optical input signal into an electrical signal corresponding to its amplitude pattern and recovers the digital signal from the electrical signal, characterised in that the frequency modulated optical signal transmitted by the optical transmitter , which is transmitted to the receiving end via the optical waveguide , is applied as the input signal to the optical receiver, which reacts to the amplitude pattern of its input signal. The specification also discloses a new receiver arrangement was provided for the reception of a frequency modulated optical signal which has travelled through an optical waveguide which has dispersion at its wavelength. The specification also describes an optical waveguide with dispersion used to convert an original frequency modulation of an optical signal into an amplitude modulation. 245766 *x ;During the transmission of digital signals, undesired distortion occurs to the digital signal, which causes the transmission quality to deteriorate. ;In practical applications, therefore, so-called eye diagrams are used for the evaluation of transmission quality. This method is, for example, known from the book "Datenubertragung Band I", by P. Bocker, Springer Verlag 1979, page 102ff. However, no provision is made for influencing the transmission quality. ;It is also proposed to provide a solution for continuously influencing the transmission quality. ;Apart from the undesired distortion during transmission, errors can occur during the reception of the digital signal. This can, for example, happen when long sequences of identical binary signals occur, e.g. a sequence of logical "1", and interference pulses occur in the receiver, for example because of noise in the receiver or the electrical preamplifier. Conventionally, the digital signal is scrambled before transmission (by a scrambler) in order to avoid long sequences of identical binary signals. However, there are communication systems in which the digital signal is not scrambled. With these, it possible for errors to occur in the received digital signal. One of the tasks of the invention is to provide a solution for which the frequency of errors is decreased. This is the subject of Claim 16. ;Before the invention is described in more detail with the aid of the diagrams, the idea which is basic to it will be explained. In accordance with the invention, the transmitting-end equipment remains unchanged compared with the known system mentioned earlier, so that a frequency modulated optical ;24 57 66 ;signal is transmitted over the optical waveguide to the receiving equipment. For the receiving equipment, however, an in itself basically known direct-reception optical receiver is used, which typically has the property that it converts the amplitude variations of its optical input signal into an electrical output signal, ;that is, it is sensitive to amplitude modulation and not to frequency modulation. ;Thus the question arises why the system nevertheless functions. The explanation lies in the following recognition of the invention: the optical signal coupled into the optical waveguide at the transmitting end consists of components with differing wavelengths which are coupled into the transmitting end one after the other. Because of the chromatic dispersion of the optical waveguide, the two signal components are delayed differently during their transmission over the optical waveguide. At the end of the optical waveguide, the two signal components with different wavelengths are shifted in time relative to one another. The interference between the two signal components caused by this leads to a variation of the amplitude of the optical signal emerging from the end of the optical waveguide. The time sequence of the amplitude variation has the property that the sequence of the digital signal being transmitted can be recovered from it. ;Thus, according to the invention, use is made of the recognition that the optical waveguide path itself (just because of the, in principle, undesirable property of chromatic dispersion) has the effect of converting an initial frequency modulation into an amplitude modulation such that it contains the sequence of the digital signal. ;24 5 7 6 ;In other words, the optical waveguide itself, which is affected by its dispersion, is used for the conversion of the initial frequency modulation into an amplitude modulation. The optical signal transmitted over it is used immediately as the input signal to the optical receiver, which is sensitive to amplitude. This does not mean that other optical arrangements cannot be used before the optical receiver, which are not intended for the conversion of the modulation but serve other purposes such as, for example, an isolator or optical amplifier. ;If the length of the optical waveguide which is necessary to cover the transmission path is not sufficient to provide the required modulation conversion, it can be increased by means of a supplementary length of optical waveguide. In this case, it is the total optical waveguide length whose optical output signal is used as the input signal for the optical receiver. Another aspect is that in this case the total length of the optical waveguide as well as its partial sections are used for the conversion of the modulation type. ;Concerning the concept used here of "frequency modulation" or "FSK modulation", the following should be pointed out: the concept is not restricted to pure frequency modulation. What is significant is that, as with the already-mentioned known system, a semiconductor laser modulation is involved which is carried out with a low modulation amplitude. With this, the amplitude of the optical output signal mostly does not remain constant, but varies together with the frequency. Nevertheless, such modulation is described as FSK or frequency modulation. What is important is that the optical output signal produced is not converted at the receiving end to ASK modulation by means of some special ;2 4 5 7 6 6 ;arrangement, but that it is processed directly in a receiver sensitive to amplitude modulation. ;The invention will now be further explained with the aid of the diagrams. Fig. 1 shows a block diagram representation of the system according to 5 the invention ;Fig. 2 shows the signal waveforms at different positions of the transmission system, in order to explain its operation ;Fig. 3 shows the typical characteristic of a Schmitt trigger circuit which can be used as the decision circuit (6) (Fig. 1) ;10 Fig. 4 shows a design example of the decision circuit depicted in Fig. 1 ;Fig. 5 shows the amplitude waveform of the input signal of the optical receiver when frequency modulation with simultaneous phase modulation occurs at the transmitting end. ;15 Fig. 6 shows two design examples for an opto-electric converter ;Fig. 7 shows equivalent circuit diagrams for explaining the low-pass filter operation of the opto-electric converter ;Fig. 8 shows a block diagram of a system with an arrangement for producing eye diagrams, and a computer 20 Fig. 9 shows a representation of the received voltage waveform, which corresponds to the optical signal, together with an interfering pulse ;Fig. 10 shows a representation of the received voltage waveform together with positive and negative voltage variations and an interfering pulse ;8 ;24 5 7 6 ;i) ;Fig. 11 shows a design example for a circuit arrangement for improving the protection against interference. ;The transmission system for digital signals according to the invention, shown in Fig. 1, has at the transmitting end (left side of the figure) an electro-optic converter 2 whose significant component is a semiconductor laser and which has the property, as in the prior art, of converting an electrical input signal in the form of a binary digital signal into a correspondingly modulated optical output signal, by means of frequency shift keying. The optical output signal thus has different frequencies for the different binary states of the digital signal being transmitted. In Fig. 1 a wavelength A0 corresponds to binary state 0 and a wavelength to the binary state 1. The amplitude normally remains constant with this modulation, as shown schematically by the waveforms with different frequencies but constant amplitudes. ;This transmitting-end equipment is connected to the receiving-end equipment of the system by an optical waveguide 3 which forms the transmission path. The new factor in the system according to Fig. 1 is that there is no equipment at the receiving end to convert the transmitted optical signal into amplitude modulation, but that instead the transmitted optical signal directly becomes the input signal for a basically known optical receiver 4. ;The optical receiver 4 has an opto-electric converter 5 which consists of an optical detector, a pre-amplifier and an amplifier and converts the time sequence of the amplitude of the optical signal into the time sequence of the amplitude of its electrical output signal. Furthermore, the optical receiver 4 ;24 5 7 6 6 ;-V. ;*) contains a decision circuit 6 (sometimes also called regenerator), which recovers the transmitted digital signal from the electrical output signal of the opto-electric converter 5. The latter signal is shown schematically at the output. The question immediately arises here why an optical receiver 4, which typically reacts to amplitude variations and not to the frequency variations of its optical input signal, should be in a position to convert a transmitted frequency modulated optical signal into the digital signal contained therein. The explanation is: it is the optical waveguide path 3 itself which produces an amplitude modulation from the initial frequency modulation of the optical signals coupled into it, and the resulting amplitude modulation has the property that the transmitted digital signal is contained within it. Above the optical receiver 4 in Fig. 1 it is shown schematically that the amplitude waveform of the received signal, which the converter 5 transforms into a corresponding waveform of its electrical output signal, has a positive pulse first and then a negative pulse, as viewed from left to right. As will be explained later with the aid of Fig. 2,the positive pulse means that the digital signal changes from state 0 to state 1, and the negative pulse means that it changes from state 1 to state 0. For the decision circuit 6, therefore, any circuit is suitable which changes its digital output signal from 0 to 1 when its input signal exceeds a predetermined upper threshold value, and which changes its digital output signal from 1 to 0 when its electrical input signal falls below a predetermined lower threshold value. Essentially, as with any known optical receiver, the decision circuit has the task of recovering, from the electrical 10 24 5 7 6 6 output signal of the opto-electric converter 5, the digital signal contained therein. As always, how it operates in detail depends on the waveform of its electrical input signal. For the case considered here, some examples will be given at a later stage. With the aid of Fig. 2, it will now be explained how the conversion of the initial frequency modulation into amplitude modulation is achieved during the transmission of the optical signal through the optical waveguide 3. First, the example shown in Fig. 1 is recalled, where, for the transmission of the bit sequence 0 10, wave trains with the wavelengths A0, Av and A0 are coupled successively into the optical waveguide. In Fig. 2 the wave trains (or signal components of the optical signal) which are coupled successively into optical waveguide 3 are shown as two rows, not one, the wave train with the wavelength A0 in Row a, and the wave train with wavelength A, in Row b. Both are referred to a common time axis running from left to right. The chromatic dispersion of the optical waveguide 3 causes the light signal with the higher wavelength to have a longer propagation time than the signal with the shorter wavelength. Observing the arrival of the signal components at the end of the optical waveguide, it is established that there is a time interval At during which the last part of the first wave train with wavelength A0 and, simultaneously, the first part of the wave train with wavelength A, are present. In other words: the wave train with wavelength A^ already arrives at the end of the optical waveguide before the wave train with wavelength AQl which propagates more slowly, is completed. The difference in 11 24 5 7 6 the propagation times of the two wave trains is At and can be calculated from: At = A/I . D . L where A/1 is the difference between A0 and A•, D is the chromatic dispersion of the optical waveguide L is the length of the optical waveguide 3 The magnitude of At must be no greater than the duration of one bit of the digital signal. A suitable value can be chosen through the selection of AA, i.e. in practice through the modulation amplitude, and possibly of the length L (D has a fixed predetermined value). Because of the differing propagation times, there is an effect at the end of the wave train with wavelength Au similar to that at its beginning, namely a state with the duration At, the difference in the propagation times, during which the wave train A, is already finished and the second wave train with wavelength /10 has not yet begun. Thus, when changing from A0 to A,, the propagation time difference for the different wavelengths results in a state with duration At during which the two wave trains are received simultaneously, and when changing from A0 to A,, a state with the same length during which neither of the two wave trains is received. The first state represents an interference between the two wave trains, with a resulting wave with a higher amplitude than either of the individual waves, and the second state results in a reduction of the received amplitude to 0. 12 245766 Row c of Fig. 2 shows schematically the resultant amplitude waveform of the optical signal emerging at the end of the optical waveguide 3, which is converted by the opto-electric converter into a corresponding waveform of its electrical signal. Therefore, depending on the sign of the dispersion and of AA , this has, for example, a positive pulse during the change from A0 to A, and a negative pulse during the change from A, to A0. From its waveform, the decision circuit 6 can recover the transmitted binary digital signal, by changing the binary state of its output signal from 0 to 1 when its input signal exceeds a first prescribed threshold value Vv and changing from this state to the binary state 0 when its input signal falls below a second prescribed threshold V0. Row d of Fig. 2 shows the waveform of the binary digital signal recovered in this way. The first threshold value V, is chosen in such a way that it lies above the signal value resulting from reception of only a single wave train and the second threshold value V0 in such a way that it lies below this "normal" signal value. The output signal of the optical receiver is therefore a bipolar signal, and the decision circuit 6 has the task of recovering the binary digital signal from it in the usual NRZ format (NRZ = Non Return to Zero). In the following, several designs will be explained for the decision circuit 6 which are suitable for this. A "Schmitt Trigger Circuit", with the characteristic of Fig. 3, is suitable as a decision circuit for changing from binary state 0 to binary state 1 when its input voltage, as shown in Fig. 2c, exceeds a first threshold value V1f and for changing from binary state 1 to binary state 0 when its input voltage falls below a second threshold value V0 which lies below Vr 13 24 5 7 Fig. 3 shows the characteristic of such a Schmitt Trigger circuit as the output voltage Va versus the input voltage Ve. If the input voltage Ve is raised, at a threshold value V, its output voltage changes from a lower value representing the binary value 0 to an upper value representing the binary value 1. If the input 5 voltage is lowered from a value above Vv the output voltage only flips back from the second state to the first state at the lower threshold value V0. Such Schmitt Trigger circuits are well known, for example from the book "Halbleiter-Schaltungstechnik" {Semiconductor Circuit Techniques) by U. Tietze and C. Schenk, 8th edition, Springer-Verlag Berlin, Heidelberg, New York, pages 10 168 and 169, as well as pages 180 and 181. A further design example is an integrator in the form of an RC low-pass filter. Its output signal rises with a positive pulse, and falls again with a negative pulse, so that from its rising and falling edge, the rising and falling edge of the transmitted digital signal can be recovered in a simple manner by a simple D flip-5 flop. In the case of the positive and negative pulses having differing energy contents, an integrator with differing time constants should be used. This is possible when an RC low-pass filter with a voltage- dependent capacitance, e.g. the capacitance of a varactor diode, is used. A further example for a suitable decision circuit is the circuit shown in Fig. 20 4, consisting of two D flip-flops and an RS flip-flop. The input signal is applied in parallel to the two D flip-flops 10 and 11. The RS flip-flop 12 has inputs R and S which are respectively connected to the Q output of the flip-flop 10 and the Q output of the flip-flop 11. The D flip-flop 10 flips into the state with Q = 1 14 24 5 7 fr tf when the input signal of the decision circuit exceeds the threshold value V1r and the flip-flop 11 flips into the state with Q = 1 when the input signal becomes less than the threshold value V0. The RS flip-flop 12 flips into the state with Q = 1 when its S input flips into this state, that is when the Q output of the D flip-flop 10 becomes equal to 1, and it flips into the state Q = 0 when its R input becomes equal to 1, that is when the Q output of the D flip-flop 11 becomes equal to 1. The Q output of the RS flip-flop 12 therefore produces the recovered binary digital signal. Instead of the two D flip-flops 10 and 11 shown in Fig. 4, a single D flip-flop with two mutually inverting outputs Q and Q can also be used, in which it is ensured, by internal asymmetry of the layout, that it has an upper threshold value, exceeding which puts it into the state Q = 1, and a lower threshold value V0, below which it flops back into the state Q = 1. It then has a switching hysteresis characteristic of the type shown in Fig. 3. A special version of the decision circuit 6 of Fig. 1 is possible when the optical signal generated at the transmitting end of the system has the property that the frequency modulation described is accompanied by amplitude modulation. This is the case when the semiconductor laser is modulated by the injection current. Such modulation is different from the normal amplitude modulation of a semiconductor, which is carried out with a much greater modulation amplitude. It is therefore primarily frequency modulation. The frequency modulation considered here is one which results in a higher level of the optical output signal for the wavelength A1 (i.e. for the higher 15 24 5 7 frequency) than for the wavelength /l0. In such a case, the signal appearing at the output of the opto-electric converter 5 has a waveform as shown in Fig. 5. It starts with a level P0 with which the wave train with wavelength A 0 is received, rises like the signal in Fig. 2c as a pulse, falls back to a level P, with which the wave train with wavelength A, is received, and falls from there again in the form of a negative pulse to the level P0. In this case it is sufficient to detect the exceeding of a single threshold Vm at the start of the positive pulse, and the reduction below it at the start of the negative pulse. Here also use is made of the existence of the positive and negative pulse which results from the frequency modulation. For the decision circuit, a conventional decision circuit can then be used, e.g. a D flip-flop with a threshold value Vm, such as is customarily used with the known optical receivers. Naturally, a decision circuit of the type shown in Fig. 4 can also be used, with two threshold values V, and V0. The following realisation of the integrator is also possible, different from the one described above in which the integrator is a component of the decision circuit and is located after the opto-electric converter. According to the invention, use is made of the fact that the photodetector has a capacitance CD and operates as an RC low-pass filter in conjunction with an additional resistor R or the input resistance Rj of a following amplifier. By adjusting the bias VD, the capacitance CD of the photo- detector can be varied and thus the integration effect of the opto-electric converter can be influenced. Fig. 6 shows design examples of an opto-electric converter 5. Here a bias 16 24 5 7 6 6 voltage VD is applied to the photo-detector 13. The capacitance CD of the photo-detector 13 functions as an RC low-pass filter in conjunction with the input resistance Rj of the amplifier 15 (Fig. 6a), or in conjunction with an additional resistor R (Fig. 6b). In Fig. 6a the input resistance Rj is drawn separately as a component of the amplifier 15. The explanation of the integration or low-pass operation takes place with the aid of Fig. 7, using the design example of Fig. 6b. In Fig. 7b a simplified equivalent circuit is used for the photodetector. The current source delivers a photo current whose magnitude depends on the coupled-in light power. The following AC equivalent circuit, Fig. 7c, shows the resultant parallel arrangement of the ideal current source 17, the resistor R, and the capacitance CD. The RC low-pass filter can be observed there. As already mentioned earlier, eye diagrams are produced in order to assess the transmission quality. The production of eye diagrams is generally a laboratory procedure for transmission quality assessment. Here it is used to influence the transmission quality and forms a component of the system. How the effect on the transmission quality is achieved in detail, is described in the following. The system with the additions according to the invention is shown in Fig. 8. It consists of a controllable amplifier 20, an arrangement 60 for generating the eye diagrams, a computer 105, and an arrangement 30 for recovering the clock signal from the digital signal. This arrangement for clock recovery is a component of every digital communication 17 245766 system, but was not shown in Fig. 1. It is connected to an output 165 of the opto-electric converter 5 and to an input 160 of the decision circuit 6, and has an output 150 for the clock signal. The amplifier 20 has an input 25 for the digital signal and an input 135 for a control voltage Vv and is connected to the electro-optic converter 2. The decision circuit has, in addition to the already- mentioned input 160, another input 140 for a control voltage Vs and an input 145 for a control voltage VT and a data output 155. The opto-electric converter 5 has an input 130 for a control voltage VD which is connected to the photodiode 13. The arrangement 60 for producing the eye diagrams is connected via an input 120 to the output 165 of the opto- electric converter 5, and via an input 125 to the output 150 of the clock recovery arrangement 30. It consists of a sample-and-hold circuit 35, an analogue/digital converter 40, a frequency divider 55, a phase shifter 50 and a pulse generator 45. The frequency divider 55, which is connected to the input 125, is followed by the phase shifter 50 and the pulse generator 45, which is connected to the sample-and-hold circuit 35. The pulse generator has an input 65 for a control voltage Vp and the phase shifter 50 has an input 70 for a control voltage Vf. The computer 105 is connected via an input 110 to an output 115 of the arrangement 60 for producing eye diagrams. It has the following outputs for the following control voltages: - output 75 for Vp 18 2 4 5 7 6 8 - output 80 for Vf - output 85 for Vv - output 90 for VD - output 95 for Vs - output 100 for VT The operation of the arrangement 60 for producing eye diagrams corresponds to that of a sampling oscilloscope, whose operation is described, for example, in Klein, P.E., "Das Oszilloskop", Franzis Verlag, Munich 1979. The arrangement 60 produces an eye diagram which is determined from the received digital signal. This eye diagram will be denoted the actual eye diagram. The computer 105 compares this actual eye diagram with a prescribed eye diagram and from the differences it deduces control voltages which are available, e.g. as analogue voltage values, at the outputs, and are transferred to the controllable system components. How the system components are controlled in detail will be described in the following. In order to control the amplifier 20 at the transmitting end, the control voltage Vv must be transmitted from the receiving end to the transmitting end. With a two-way communication system, this transmission can, for example, be achieved by converting the control voltage Vv into an optical signal by means of electro-optic converter, and transmitting this to the transmitting end by the wavelength multiplex method. The conversion takes place there from an optical signal to an electrical signal, which once again represents the control 19 24 voltage Vv. With this, the amplifier and, therefore, also the electro- optic converter can be controlled as follows: The control voltage Vv causes a change Al in the output current of the 5 amplifier which is proportional to the control voltage Vv, i.e. Al _ Vv. A current change of Al in the electro-optic converter 2 causes a frequency change At/ of the optical digital signal, which is proportional to the current change Al, i.e. Au = Al . The frequency change Au is equivalent to a wavelength change A^. By this means, the frequency shift of the opto-electric converter 2 can be 10 influenced. A further control voltage VD controls the bias voltage of the photo diode in the opto-electric converter 5 and thereby the capacitance of the photo diode, as already described. With this, the frequency response of the opto-electric converter can be controlled, which is an advantage when an integrator with different time constants is to be used. This has already been mentioned in 15 connection with the integrator. The threshold value of the decision circuit 6 can be controlled with the control voltage Vs. The clock recovery arrangement 30 provides a clock signal with constant frequency. The control of the decision circuit by means of the control voltage VT 20 is achieved by controlling the phase of the clock pulse, i.e. the decision instant is so controlled that the opening of the eye diagram is biggest. The pulse generator 45 and the phase shifter 50 of the arrangement 60 can also be controlled for the production of the eye diagrams. The control 20 2 4 5 7 6 6 voltage Vp controls the pulse generator, and the control voltage Vf controls the phase shifter. By this means, the sample-and-hold circuit is controlled, and thus the time of sampling. With the measures described here, the eye diagram which results from the received digital signal can be influenced and the transmission quality of the system can be optimised. Independently of the system described so far, the production of eye diagrams, their comparison with a prescribed eye diagram and the control of system components can be used for every optical communication system which is for the transmission of digital signals. With the transmission described here, via an optical waveguide which has dispersion, an amplitude modulated signal results in the optical receiver. A decision circuit with intermediate buffer recovers the data from this. This signal can in some cicumstances be distorted by noise puses, as already mentioned. In the following the formation of an error, and means for reducing the frequency of errors, is described. Fig. 9 a) shows, as an example, the distortion of the electrical signal in the receiver by means of a negative noise pulse. The case is considered there of a negative noise pulse appearing at time t0 during the formation of the electrical output signal in the opto-electric converter. The magnitude of the noise pulse is here so large that the signal falls below the threshold value V0 of the decision circuit and that the digital signal recovered by the circuit changes from state "1" 21 24 5 7 to state "0". This state transition at the time t0 represents an error in the recovered digital signal which will be propagated onwards. If required, in order to reduce the frequency of such errors, the optical signal being transmitted, which is frequency modulated, has its amplitude varied in the optical transmitter as a function of the digital signal. This procedure leads to the fact that the electrical signal which appears at the output of the opto-electric converter 5 (Fig. 1), has voltage changes additional to those pulses which are due to the digital signal. Details of this procedure are explained later. Fig. 10 a) shows the electrical signal (voltage waveform) with a positive and a negative voltage shift. If now during the positive voltage shift, at time t0, there appears a negative noise pulse whose height in this example is equal to that of the pulse in Fig 9 a), a reduction of the voltage below the lower threshold V0 is prevented because of the positive voltage shift. Therefore, no error occurs in the received digital signal. The same applies if, during a negative voltage shift, a positive noise pulse occurs. Because the magnitude of noise pulses can vary considerably, in the positive as well as in the negative direction, errors can nevertheless still occur. In fact, whenever the pulse height is large enough, for example, for the voltage to fall below the lower threshold. In that case, there is only an error for the duration of the noise pulse. Afterwards the output signal is set back to the correct value. There is thus no error propagation. 2 4 5 7 6 0 With the aid of Fig. 11, the following will explain how and when positive or negative voltage shifts occur, and a circuit arrangement is described which produces the voltage shifts. Fig, 11 shows the transmitting end of the system of Fig. 1, here supplemented with a circuit arrangement 200 and an optical amplitude modulator which serve to impress amplitude shifts onto the optical signal. An optical amplitude modulator is a component whose optical transmission is dependent on the applied voltage. An example of one is a Mach-Zehnder Interferometer. Further examples are described in the publication of Robert G. Waller "High-Speed lll-V Semiconductor Intensity Modulators", IEEE Journal of Quantum Electronics, Vol. 27, No. 3, March 1991, pages 654-667. The circuit arrangement 200 consists of an N-bit shift register 220, two AND gates 210, 215 and a differential amplifier 205. The shift register 220 has an input 240 for the digital signal and an input 245 for a clock pulse. An output 225 is connected to an input 25 of the electro-optic converter 2. Qj and Q| (i = 1,2,...,N) are outputs of the shift register which are connected to the AND gates 210 and 215: Q|, for example, to the AND gate 215 and Q| to the AND gate 210. The outputs of the AND gates are connected to the inputs of a differential amplifier 205 whose output is connected to an input 235 of optical amplitude modulator 230. The digital signal being transmitted is sequentially written into the shift register. The AND gates form a logic circuit which checks if all N outputs Q( or Qi of the shift register have the same logical binary state. If all outputs Qi = "1,,> 23 245766 a positive voltage from the output of the differential amplifier 205 is applied to the optical amplitude modulator, and if all the outputs Qj = "1", a negative voltage is applied. A positive voltage at the optical amplitude modulator results in a stronger transmission signal, a negative voltage in a smaller one. The amplitude shifts are produced by the changes in the transmission signal, and the consequent changes to the power of the optical signal. A positive or negative amplitude shift is only generated and transmitted when a predetermined number of equal binary signals occur in sequence. An amplitude shift lasts until the first output Q| becomes different to the remaining outputs. An amplitude shift can therefore last for several clock periods. Amplitude shifts of the optical signal are converted to voltage changes in the receiver. This method reduces the danger of noise pulses during long sequences of identical binary signals, and their resultant corruption of the binary signal. 24 </div>

Claims

<div class="application article clearfix printTableText" id="claims"> 245766 What we claim is:-

1. An arrangement for communicating information in digital form over an optical waveguide between transmitter means and an optical receiver, wherein the waveguide exhibits chromatic dispersion in the region of the frequency of the optical carrier wave, the transmitter means including: frequency modulating means to modulate the optical carrier wave with digital information signals, the modulated optical carrier wave being transmitted over the waveguide; the optical receiver including amplitude detection means to detect the amplitude of the modulated optical carrier wave at the optical receiver, the information being extracted from the output of the amplitude detection means.

2. An arrangement as claimed in claim 1, wherein the transmitter means includes amplitude modulation means to modulate the amplitude of the carrier wave.

3. An arrangement as claimed in claim 2, wherein the amplitude modulation means modulates the amplitude of the carrier wave so that the output of the amplitude detection means for a series of bits of a first binary value in the information in digital form, has a first steady state value greater than a median value, and wherein the output of the amplitude detection means, for a series of bits of a second binary value in the information in digital form, has a second steady state value less than the median value.

4. An arrangement as claimed in any one of claims 1 to 3, including: an eye diagram generator means to generate a received eye diagram from 245766 the output of the amplitude detector means; processor means to compare the received eye diagram with a predetermined eye diagram and to generate one or more control signal to control one or more components of the arrangement to optimize transmission quality of the arrangement.

5. An arrangement as claimed in claim 4, wherein a first control signal is fed back to the transmitter means to control adjustment of the frequency modulation of the carrier wave.

6. An arrangement as claimed in claim 5, wherein the first control signal controls the gain of a first amplifier which controls the frequency shift of the frequency modulation means.

7. An arrangement as claimed in any one of claims 4 to 6, wherein the processor means generates a second control signal to control the frequency response of the optical receiver.

8. An arrangement as claimed in claim 7, wherein the optical receiver includes an integrator to recover the digital information.

9. An arrangement as claimed in claim 7 or claim 8, wherein the amplitude detection means includes a photo-diode with a capacitance controlled by the second control signal, the photo-diode being associated with a series resistor, whereby the amplitude detection means functions as an integrator.

10. An arrangement as claimed in claim 8, wherein the integrator is an integrating low-pass filter fed by the amplitude detection means. 245 7 66

11. An arrangement as claimed in any one of claims 1 to 10, wherein the optical receiver includes a decision circuit responsive to the output of the amplitude detection means, wherein the output of the decision circuit switches to a first state when the output of the amplitude detection means exceeds a 5 first threshold value and switches to a second state when the output of the amplitude detector means falls below a second threshold value, the second threshold value being equal to or less than the first threshold value.

12. An arrangement as claimed in claim 11 as appended to claim 3, wherein the first steady state value is greater than the first threshold value and, wherein 10 the second steady state value is less than the second threshold value.

13. An arrangement as claimed in any one of claims 1 to 10, wherein the optical receiver includes a decision circuit responsive to the output of the amplitude detection means, wherein the output of the decision circuit switches from a first binary state to a second binary state when the output of the 15 amplitude detection means has a rising pulse edge which exceeds a first threshold value, and from the second state to the first state when the output of the amplitude detection means has a falling pulse edge which falls below a second threshold value.

14. An arrangement as claimed in claim 11 as appended to claim 4, wherein 20 the processor means generates at least a third control signal to control the threshold values of the decision circuit.

15. An arrangement as claimed in claim 4 or any one of claims 5 to 14 as _n.z. patent office 3-JUL 1995 27 received 245766 appended to claim 4, including a clock generating circuit to generate a clock signal from the modulated carrier wave to provide a timing base signal for the eye diagram generator means, and wherein the processor means generates a fourth control signal to control the phase of the clock signal.

16. An optical transmission system substantially as herein described with reference to the accompanying drawings.

17. A method of communicating information in digital form over an optical waveguide between transmitter means and receiver means, wherein the optical waveguide exhibits chromatic dispersion over the range of the frequency of the modulated optical carrier wave, the method including: frequency modulating an optical carrier wave with digital information signals at the transmitter means; transmitting the modulated optical carrier wave over the optical waveguide; detecting the amplitude of the optical carrier wave at the optical receiver; extracting the information from the detected amplitude of the carrier wave.

18. A method of demodulating transmission dispersed frequency modulated digital optical signals substantially as herein described with reference to the accompanying drawings.

19. An optical receiver connected to an optical waveguide in a transmission system in which digital binary information is transmitted over an optical waveguide as frequency modulation of an optical carrier wave, wherein the 245766 waveguide exhibits chromatic dispersion over the range of the modulated optical carrier wave frequency, wherein the receiver includes optical amplitude detection means to detect the amplitude of the optical carrier wave, binary information of a first type being transmitted at a first frequency, and binary information of a second type being transmitted at a second frequency, wherein one of the first or second frequencies travels slower in the waveguide than the other frequency, so that when a binary bit represented by the slower frequency is followed by a bit represented by the faster frequency there is a partial overlap of the slower and faster frequencies at the receiver for a period determined by the difference in the speed of the first and second frequencies and the length of the fibre, causing an increase in the amplitude of the signal received for the duration of the overlap, and wherein when the faster frequency is followed by the slower frequency, the amplitude of the signal received tends to fall towards zero for a period determined by the difference in the speed of the first and second frequencies and the length of the fibre, wherein the receiver detects changes from binary information of the first type to binary information of the second type when the amplitude of the received signal increases above a first threshold, and wherein the receiver detects changes from binary information of the second type to binary information of the first type when the amplitude of the received signal falls below a second threshold. ALCATEL AUSTRALIA LIMITED P.M. Conrick Authorized Agent P5/1/1703 </div>