DE2025920A1 - transmission system with phase-modulated pulses - Google Patents

Description

Patent attorney

7 Stuttgart -30 ■

Short street 8

AH. Reeves-M.M. Ramsay - A.E. Mounter S.P. Laurence 92-3-2-1

INTERNATIONAL STANDARD ELECTRIC CORPORATION, NEW YORK

Transmission system with phase-modulated pulses

The invention relates to a transmission system with phase modulated Impulses.

Such a transmission system is explained with reference to FIGS. La and lb. In Fig. La is a reference carrier shaft shown, which has the same wavelength as the signals transmitted with the pulse-modulated carrier of Fig. Ib. These modulated pulses represent unequal characters because they have a different phase relationship to the reference carrier wave exhibit. The first pulse of the carrier is in phase with the reference carrier wave while the second pulse is in phase is in phase opposition to the reference carrier wall. Fig. Ib shows a system in which the phase separation between unequal sign with Jf is given. This so-called Pulse phase modulation can easily be carried out with pendulum oscillators, since these have the property that they can be triggered with a part of their Ira pulse cycle by a control pulse. The carrier wave of the

25th time 1970 009 849/17

AH. Reeves 92-3-2-1

Control pulse has the same frequency as the oscillator. The oscillator is therefore synchronized with the phase of the control pulse ignited.

It is the object of the invention to provide a transmission system with phase-modulated pulses in which the simplest On the receiving side, the transmitted pulse trains not only contain the information but also the reference signals

"can be derived. The transmission system with phase-modulated Impulse is characterized according to the invention that the binary characters of the information through different phase position of the carrier wave of the impulses with respect to a reference wave of the same frequency, the reference signals being at regular time intervals are transmitted as synchronization pulses over the same channel. From the preferably transmitted with the same phase Synchronization pulses can receive a reference signal for determining the phase of the phase-modulated information pulses be derived. Details of the new transmission system can be found in the description of an exemplary embodiment

Wk > and the subclaims.

The invention is illustrated by means of one in the drawings optical pulse phase modulation transmission system explained in more detail. Show it:

2 shows a semiconductor laser as a pendulum oscillator,

Fig. 3 © in vector diagram showing the phases of the pulses in the case of a pulse repeater of the transmission system, in which the Pkaseparation between unequal characters is lf / 2,

009849/1709

AH. Reeves 92-5-2-1

Fig. 4 is a vector diagram for explaining the Pulse phase correction via several pulse repeaters connected in series,

Fig. 5 is a transmission scheme showing phase recovery shows in a transmission system in which the phase separation between unequal Character Jf is

and

Fig. 6 is a schematic diagram of an optical pulse repeater for a transmission system in which the Phase separation between dissimilar characters is 3f.

Since the exemplary embodiment uses a pendulum oscillator, the mode of operation of such a circuit will first be explained.

The mathematical analysis of how a pendulum oscillator works Is known, and it is clear, that these principles are also used for the construction of an oscillator working with optical frequencies. A pendulum oscillator is an oscillator whose damping constant varies over time. So that this oscillator itself cancels, the damping factor must be generated within the oscillator and a delayed function of its output signal be. This implies that the equation for the operation of the oscillator must be a non-linear differential equation of the second order.

A laser, which is an optical oscillator, can in principle be designed as a pendulum oscillator if in the driver stage of the laser creates a feedback loop

0098 4 9/1709 ./.

AH. Rewes 92-3-2-1

is introduced. This requires the use of photoelectric receivers and delay devices in the feedback loop in order to obtain the desired pendulum frequency oscillation. It is difficult and expensive to obtain a pulse repetition frequency greater than 1 QEz with this type of pendulum oscillator. The pulse repetition frequency is essentially determined by the rise and fall times of the laser driver wave is constantly pending and an optical feedback is used. The fact that eisie can visually © feedback 'generation eiiiQs "Pendelfregaeaz-Sßtejii ^ Mig verw © ai ©, is zvm part darlia fesgrtiadet, tHe the dtttefak'eöi? Q and the running time of the Leeqpq ®ine ¥ eFSsög © rimg and that the ratio of the Pppislatiöaswikelii ⁹ and the speed of the light build-up © la der Hesonanskammmer are both functions that are dependent on the optical field strength in the Issonanskafiiiier. This ia.t to era @ hen "tHe © in laser permanently pending üireiberwell® äareh insertion of a nonlinear optical attenuation @ inrichtung to a pendulum" oscillator-made canoe »The nonlinear Däaipfungs can" device from known Substsnssen formed, whose damping decreases below the Einflui of strong laughs. A half-parent injection diode ~ is used for the laser. The non-linear damping device can then be made from the same material. The semiconductor diode used as nonlinear DMmpfungseinrichtung may treat electrically energized in this case, "vm and reduce the optical field%, which for AbsohwMchung required νύ.Έ? _Ά α Diode MUB in each Pail far below the 'lÄ3 © s ^ S © liw © lle be operated so that it in the absence of external optical Jeios ³ - not z "i @ liwiiig" n GeBr characterizing lighting eht be the; Illumination of the diode with light _Λ that is roughly the wavelength

009849/1700 ./.

AH. Reeves 92-3-2-1

has, leads to the excitation of transitions between the two energy states that determine the laser wavelength. Since a larger population of the If the switching state is lower, network energy is absorbed from the incident lighting. if the lighting is strong enough to suppress the self-created transitions then the condition becomes achieved in which the two populations are equal. The probability that an incident photon has a j

Stimulates transition from low to higher energy state, becomes equal to the probability that a reverse transition will take place. Under these conditions will made the diode statistically transparent.

From the above it can be deduced that an optical pendulum oscillator has an optical coupling can be constructed from two semiconductor injection laser diodes. In operation, both diodes are constantly pending Driver shaft exposed. The control of a diode is strong enough that the diode emits laser-corrugated waves.

The other diode is deliberately controlled in such a way that it does not emit any laser-correlated% waves without additional external influences. It is experimental reasons that led to the view that two separate laser resonance chambers are absolutely necessary. At least under certain conditions it is sufficient to provide a single laser, the current density of which can be made smaller via the PN junction in a part of the length of the resonance chamber than in the remaining part. To achieve this, an electrode sheet of the device can be divided into two parts, like Big. 2 shows. The two parts are switched on separately, so that the current density across the PN junction

009849/1709

AH. Reeves 92-3-2-1

can be set differently. This can, as indicated in FIG. 2, be fed. FIG. 2 shows the semiconductor piece 1 with the PM junction 2. A metal electrode is denoted by 3 _s while the other electrode is divided into the two parts 4 and 5. The division is carried out by a channel β which runs parallel to the reflective surfaces 7 of the laser. The two parts of the laser are connected to a common voltage source 8, namely via the llIdQFStiacle 9 u ^ d io, which are chosen so that the current density tifeei? the part of the laser under the electrode 4 is greater than the oil current density in the part under the electrode 5 »

The output transmitter of a UberferagiHigssyBfeeia ^ fessteöt from a simple fashion follow-up line which is followed by an electro-optical phase plate. This laser ⁵ tmt irorzugsweise constructed to d & S oils Mifcnafeao bsw. "Uer MitlaMf Within the laser reaches viirdp ü © tHe no Moönlatloasfrequenz = band signals are needed produces" The 'fashion litlaiaflaseF phase-coherent pulses in regelraäSigen Zeifcabstäuden. Before these pulses emitted they pass an electro-optical crystal, so that the impulses can be delayed individually by half a wavelength. "This delay is caused by a corresponding electrical potential on the crystal.

The restoration procedure will now be correct Phase position of the impulses explained when they are an impulse » iterate repeatedly. The one on an impulse repeater incoming impulses have a. determined © amplitude and Phase fluctuation on. The cancellation of the iUnplifoad fluctuation is relatively easy. Is becomes a threshold limiter used, e.g. as a pendulum oscillator isfe. Dl © cancellation the phase fluctuations is w @ ifc more difficult ws & d becomes 3, 4 and 5 explained with reference to FIGS

AH. Reeves 92-3-2-1

In Pig. 3 the correct phase position of the input pulses for the characters "1" and ^H O ^{W is shown} by the vectors OA and OB. In reality, however, these impulses are disturbed due to noises which cause amplitude and phase fluctuations. The amplitude fluctuations are first compensated for by a threshold circuit so that the vector representation of the limited pulses "1" and "O ^{" lies in the ranges OA 1} to OA ¹¹ and OB ¹ to OB ". These pulses are then mixed with a locally generated signal , the phase and amplitude of which is given by the vector CO. The phase of this vector is chosen so that it lies in the middle between the phases of the vectors OA and OB. The amplitude of this vector is chosen twice the amplitude of the limited input signals Small errors in phase are therefore given by the vectors OA ', OA ", OB * and OB" and are completely converted into amplitude fluctuations represented by the vectors CA ¹ , CA ", CB * and CB * and by a subsequent one Threshold circuit can be compensated »If the effective signal-to-noise ratio at the input is, for example, 20 dB, then the effective phase fluctuation, which is given by this noise, is about 4, 3 ° and can be reduced to around 10 * with this system. It must now be noted that as a side effect of this cancellation of the phase fluctuationaa there occurs a reversal of the phase difference between the characters ^M 0 "and" 1 *. If a character ⁿ l ^{tt leads} a character "0 ^{n" by} JT / 2 before the cancellation of the phase fluctuations, then it will lag behind by JJ / 2 after cancellation and vice versa.

If the phase error is relatively large, but still smaller than Tf / 4, then a pulse repeater old with this phase angle correction is not sufficient to completely compensate for the phase error, as shown in FIG. 4 shows. When the phase error at the beginning

009849/1709,

AH. Reeves 92-3-2-1

is given by the angle XjOA = ck , then the phase error at the output of the pulse swinger is a value which is given by the angle X ₁ CA - & . However, if this signal passes through a series of cascaded pulse repeaters, then the correction is improved at each stage. The phase error O1 therefore appears at the next pulse repeater as angle XgOB. For the third, fourth and fifth pulse repeater, the phase errors are given by the angles Χ, ΟΑ, X ₁ ^ OB and XgOA. The phase correction can therefore be applied several times. However, the prerequisite is that the phase error is never greater than 3f / 4.

An improvement of this type can also be achieved in a transmission system in which the characters ^w l ⁿ and ^η 0 ^Μ are transmitted with a phase separation of JT. The system provides more phase correction according to FIG. 4 if the error is not greater than JT / 2. The most essential of this system is that the phase separation of dissimilar characters is first converted from Jf to Jf / 2. The signals are then processed further in the catfish, as shown in Flg. 2 was explained. Finally, the signals are converted back to their original phase separation of 3 [. In Fig. 5 the transmission scheme for this system is shown. The incoming signals, which have a normal phase separation of T between dissimilar characters, are represented either by vector 40 or by vector 41 due to the amplitude and phase fluctuations and depending on their normal phase. These incoming signals are first mixed by a threshold value circuit 42 in order to compensate for the amplitude fluctuations, and are then mixed as a locally generated signal, the phase and amplitude of which is represented by the vector 43

009849/1709 ,

AH. Reeves 92-3-2-1

are. This results in a signal that, due to the Amplitude and phase fluctuations and depending on the original phase at the input either by the Vector 44 or represented by vector 45. the The amplitude fluctuation is detected via the threshold value circuit 46 balanced and then mixed the signal with a second locally generated signal, whose amplitude and phase through the vector 47 are given. This is due to the fluctuation of the araplitude resulting signal obtained has a much smaller phase fluctuation. This signal, which is in Depending on the original phase at the input is given either by vector 48 or by vector 49, passes through a threshold value circuit 50 to cancel the amplitude fluctuation and is then localized with a third The signal generated is mixed, the amplitude and phase of which are given by the vector 51. That obtained after this mixture The signal is essentially free of amplitude and phase fluctuations and, depending on the phase, is through at the input the vector 52 or the vector 53 given. This is to see that by adding the signals to the first locally generated signal, the phase separation between dissimilar Signals from 3f to 3JT / 2 is converted while the phase errors can be reduced via the second addition. The third addition finally becomes the original Phase separation of 3f restored. If the phase error is relatively large, then it can be about a Pulse repeaters cannot be fully compensated for. The However, almost complete repeal.

009849/1708

AH. Reeves 92-5-2-1

Because the phase of a signal only in relation to a specific one The form of a reference point, e.g. a reference shaft, can be measured, requires the transmission of information Phase modulation also means the transmission of a reference signal from which the reference wave can be derived * It is on the simplest way to transmit this reference signal over the same channel over which the modulated signals are transmitted so that all slow and random changes in the path length in the transmission system cause a phase error, since they affect both the modulated signals and the reference signals in the same way.

One method of deriving this reference signal uses a transmission system in which the transmitted pulses contain both the information and the reference signal at equal intervals. Some pulses in the pulse train are used to transmit the base signal. This Synchronisierimpalse _β involving the Besagssigpaal should preferably occur in regeJümäSigen time periods. Therefore, for example, every single pulse is used as a synchronization pulse.

The pulse repeater is designed so that successive synchronization and reference pulses in a transmission system are separated by η modulated signal pulses. The synchronization pulses occupy position po in the pulse train. The first modulated signal pulse ISj, which occurs after the synchronization pulse, occupies position P ₁ , the second _{occupies position p g} and the n-fce occupies position pn. This position pn is immediately before the next synchronization pulse.

009848/1709

AH. Reeves 92-3-2-1

If the pulse repeater is also to restore the number of pulses received in the manner described, then it must be able to distinguish between the synchronization pulses and the modulated signal pulses. A local reference wave which is in phase with the transmitted reference signal is then derived from the synchronization pulses. For this purpose the transmission system provides a code character which comprises a form of periodic character transitions and which results in the condition for the probability that a pulse in position ρ (1 ί χ ί "η) has the same phase as the previous pulse in position _ρ χ has. This probability is only about half. the probability is the one opposite that all Synchronisierimpulae in the position p _Q have the same phase.

Under these conditions, the synchronization pulses can be separated from the modulated signal pulses. A Fabry-Perot interferometer is used for this purpose, the basic resonance of which corresponds to the pulse train frequency of the synchronization pulses. If such an interferometer switched into the transmission path of the pulses, then a pulse can position ρ only old subsequent pulses interfere ρ of the position. There is no mutual influence between this pulse p_ and pulses that occupy other positions ρ (yfx). First, the synagogue impulses are considered. All synchronizing pulses have the same phase, the chamber is in resonance and the energy is taken over by the incident light, the majority of which is transmitted through the interferometer. The impulses that take the positions ρ _χ (1 £ χ ^ η) can be divided into two phase groups. The individual elements of each

009849/17 09

AH. Reeves 92-5-2-1

Group would interfere with each other in the same way like the synchronization pulses, because the numbers in the groups roughly the same and the phases are opposite. The interferometer is therefore almost the entire incident Reflect energy contained in these impulses. It follows that the interferometer continues to transmit the synovial impulses with little attenuation during the Main part of those contained in the modulated signals Energy is reflected. The figure of merit of this interferometer is made as large as possible for two reasons. Firstly, it improves the differentiation between the synchronization pulses and the other pulses second is the phase of the transmitted pulses Average value of the largest possible number of synchronization pulses. In this way the phase noise becomes of the individual output pulses emitted by the interferometer compared to the phase noise of the individual incoming synchronization pulses reduced.

Another method for separating the synchronizing pulses from the other signals uses part of the incoming signal to control a pendulum oscillator, the pulse rate of which is slightly less than the pulse rate of the synchronizing pulses. The size of this control signal is set in such a way that it is just insufficient to pull the pendulum 'oscillator along. Under these conditions, the ignition of this oscillator will coincide approximately with a pulse position, eg p_>, for a number of cycles, before the coincidence ar * passes over _{the next pulse position ρ χ 4-1.} If the pendulum oscillator fires while a trigger pulse is pending, it fires with the same phase as the trigger pulse, even if the pendulum oscillator is not driven with this pulse. That

009849/1709 · / * ■

AH. Reeves 92-3-2-1

Output signal of the pendulum oscillator can be used as a feedback signal can be used for entrainment by the synchronization pulses. For this purpose, the output signal of the pendulum oscillator fed to a Fabry-Perot interferometer, which is in resonance with the light frequency. The basic resonance this interferometer can be set to the pulse repetition rate of the sync pulses or to a low harmonic this frequency. The output signal of this interferometer is fed back to an auxiliary control signal for amplifying the control signal derived from the incoming signal. When the pendulum oscillator is free oscillates and roughly in synchronism with the impulses in the positions ρ (1 = χ ^ "η) ignites, then the phases

Jt

of the emitted pulses tend to change because the character reversal is periodic. The interferometer will therefore reflect the major part of the energy of the pulses and the auxiliary control signal actually assumes an amplitude of zero. As a result, the pendulum oscillator will drive via approximate synchronism over the various pulse positions until it comes into approximate synchronism with the pulse position ρ _{after passing through the position p n.} Due to the fact that all the pulses have the same phase, the amplitude of the auxiliary control signal begins to increase. The control signal derived from the incoming signals is increased and the pendulum oscillator is pulled along by these pulses. If for any reason this pendulum oscillator is disturbed, the oscillator ignites synchronously with a modulated signal pulse and the next control pulse with the opposite phase is no longer amplified via the auxiliary control pulse, but rather reduced. If this is not sufficient to immediately cancel the entrainment of the oscillator, then this takes place immediately after the auxiliary control signal has decayed, since the pulses with different phases which are fed to the interferometer are caused by this.

009849/1709

AH. Reeves 92-3-2-1

It is not necessary to derive a local, continuous reference wave for each pulse repeater, since the locally generated signals to restore the pulse phases are only required during the pulses. For this Basically, it is sufficient to feed the synchronization pulses to a second Fabry-Perot interferometer, its fundamental frequency is equal to the pulse repetition rate of the entire system. This interferometer then also generates output pulses the desired pulse repetition rate. This results in inevitably a certain drop in amplitude of the pulses between two successive synchronization pulses. If this drop is significant, then it can be canceled simply by sending the signal to the controller another pendulum oscillator is used.

One difficulty in setting up an optical pulse repeater is the elimination of spurious feedback paths, the appearance of which is attributable to the fact that the majority of optical devices are bidirectional are. Spurious feedback loops of this type can be eliminated by using Faraday directional lines. Deviating from this, damping lines can also be used in order to save costs and design difficulties to avoid. The application of these damping lines is based on the fact that a pendulum oscillator is a Gain of about 60 dB and only in a relatively short period of time over an input pulse can be controlled. The separation between pendulum oscillations, that are connected in cascade can be achieved with damping lines that are connected between the oscillators will. The prerequisite for this, however, is that the path lengths are chosen between the successive devices so that one transmitted in the opposite direction Signal at the wrong time .. arrives to the upstream Control facility. It is understandable that, though

009849/1709

AH. Reeves 92-3-2-1

this light does not control the facility «a considerable part of the light is reflected and so in the reverse Direction flows back. An improvement in this separation could therefore be achieved by reducing the reflectivity of the fender oscillator. A procedure to produce a variable reflectivity of this type, a coating appears at the end of the semiconductor laser permeable dielectric in front of «one strength of one half wavelength and whose refractive index is matched to the real portion of the laser refractive index. A multi-layer, highly reflective stack is then deposited on the half-wavelength thick layer. If the laser reaches the threshold, then occurs in this arrangement at the interface between the laser material and the layer with a diode of half Wavelength no reflection, since the two refractive indices are matched. When the laser is below the threshold is operated «then the laser material is strongly attenuated for light with the laser wavelength. The refractive index is therefore complex and has a large imaginary part. Will If the laser is operated below the threshold, then the refractive index is at the interface between these layers mismatched, resulting in considerable reflective ability. -

Adding these layers to a laser has none Influence on the performance of the laser, if this is operated above the threshold. The resonance chamber is only lengthened by half a wavelength. on on the other hand, a laser operated below the threshold acts as an interference filter with a dielectric Spacer with half a wavelength. Despite the high reflectivity stack, a large proportion of the

009849/1709 /

Century 2Q2.5920

AH. Reeves 92-3-2-1

energy hitting the layers is allowed through and absorbed in the laser material. The layer with a thickness of half a wavelength is made of aluminum, in which gallium arsenide is enriched.

With reference to Fig. 6, a complete pulse repeater for a transmission system is described, which a phase separation of J | between unequal characters. With 61 the input of the Xmpulswlederholers is designated. The incident light is fed via a beam splitter 62 to a pendulum oscillator 63 which works as an input pulse amplifier and as a threshold value circuit. Part of the output light of the pendulum oscillator 63 is transmitted via the beam splitter 62 and deflected by the beam splitter 64 into the Fabry-Perot interferometer 65, the basic resonance of which is equal to the pulse repetition frequency of the synchronization pulses. As already mentioned, this interferometer works as a filter that allows the synchronization pulses to pass and blocks the other pulses. These synchronization pulses are then fed to a further Fabry-Perot interferometer 66 , the basic resonance of which is equal to the pulse repetition frequency over the entire transmission system. The output of this interferometer is a series of reference signals which are in phase with a reference wave and have a pulse repetition frequency which corresponds to the pulse repetition frequency over the entire transmission system. Due to the thorough determination of the two interferometers, the reference pulses have a smaller phase noise than the individual synchronization pulses. These reference pulses are fed to the pendulum oscillator 68 via a beam splitter 67, in order to prevent the amplitude of the: * pulses between successive synchronizers. The output signal of this Pendelossillafeore 68 is successively fed to the beam splitters 69 and 70 to localize the three

009849/1709 /

generated signals a, b and c. The phases and amplitudes of these signals are matched to the values required for phase recovery of the incoming pulses. The appropriate phases are generated at the desired locations by appropriate selection of optical path lengths. The first locally generated signal a is derived from the transmitted part of the incident light from the pendulum oscillator 68 at the beam splitter JO and superimposed on the output signal of the pendulum oscillator 63 with the beam splitter 71. The resulting signal is fed to the pendulum oscillator 73 via the beam splitter 72. The superposition of the first locally generated signal a converts the phase separation between unequal symbols from 3 \ to Tf / 2. The pendulum oscillator 73 serves as a threshold value circuit in order to compensate for the amplitude fluctuations which are caused by the phase fluctuations in the signals of the pendulum oscillator 63. The second locally generated signal b is derived from the reflected part of the light from the pendulum oscillator 68 which strikes the beam splitter 70. This signal b is superimposed via the beam splitter 74 with the output signal of the pendulum oscillator. The resulting signal passes through the beam splitter 75 to the pendulum oscillator 76. This superimposition of the second locally generated signal is used to reduce or cancel the phase fluctuations in the signals received by the pendulum oscillator 73, these phase fluctuations being converted into amplitude fluctuations that are generated by the as Threshold circuit operating pendulum oscillator 76 are suppressed. As already mentioned, a disruptive effect of this superposition of the second locally generated signals is the phase reversal of the characters. The third, locally generated signal c is derived from the transmitted light component of the pendulum oscillator 68 via the light splitter 69 and superimposed with the output signal of the pendulum oscillator 76 via the beam splitter 77. The superposition of this third locally generated signal transforms the phase separation between dissimilar symbols of Jf / 2

009849/1709 ./.

Λ »

AH. Reeves 92-3-2-1

return. The resulting signal represents the output signal of the pulse repeater at output 78.

One-way optical lines are not shown in FIG. 6, which prevent undesired feedback. To be good To obtain separation, such one-way lines are required in at least three places in the pulse repeater. One One-way line is absolutely necessary in order to prevent light from being reflected back into the Fabry-Perot interferometer 66. Another one-way line is required to decouple the two interferometers 65 and 66, so that the operation

will.

of interferometer 65 does not affect another one-way line must be switched on before the first element of a pulse repeater or after the last element because it it is unlikely that the optical path length between two successive pulse repeaters will be on one sufficiently small tolerance can be kept to ensure that light pulses between the pulse repeaters in the opposite direction do not occur at the wrong time and a pendulum oscillator of the upstream pulse repeater can drive.

The final receiver of the transmission system is constructed similarly like the pulse repeater according to FIG. 6. The output signal at the output 78 is recorded via a photodetector. The third, locally generated signal c has a different one Phase and amplitude, which are chosen so that the superimposition of this signal at the beam gap step 77 © ine cancellation the impulses of one type of character caused by destructive interference, while the impulses of the other Character type are only subject to a phase shift.

15 claims

0 sheets drawings * 6 fig.

009849/1709

Claims (15)

.H. Reeves - M.M. Ramsay - A.E. Mounter P. Laurence 92-3-2-1 claims 1. Transmission system with phase-modulated pulses, characterized in that the binary characters ("1" and ¹¹ O ") of the information are represented by different phase positions of the carrier wave of the pulses with respect to a reference wave of the same frequency, the reference signals at regular time intervals as synchronizing pulses transmitted over the same channel. *. Transmission system according to Claim 1, characterized in that the modulated signals for different characters ("1" and ^B O ^M ) have a predetermined phase separation (eg Jf or 3F / 2). 3. Transmission system according to claim 1 and 2, characterized in that the frequency of the carrier wave is in the optical or quasi-optical frequency range. 4. Transmission system according to claim 1 to 3, characterized in that the reference signals transmitted as synchronizing pulses always have the same phase and that the modulated signals between the synchronizing pulses form a code symbol »which provides a periodic symbol reversal. Vo / Gr May 25, 1970 009849/1709 .H. Reeves 92-3-2-1 5. Transmission system according to claim 1 to 4, characterized in that several pulse repeaters are switched on in the transmission path, which convert the phase fluctuations of the incoming signals into amplitude fluctuations, which are then eliminated via a threshold value circuit. 6. transmission system naoh "claim 5, characterized in that that a pulse sweater and the end recipient © inen or several pendulum oscillators and phase recovery facilities contain “the at least a part of the Fhasensohwiegen the ankonrosnd ©” impulses in amplitude fluctuations convert ^ al © then via a Scliwellwe ^ fc = circuit can be eliminated. 7k transmission system aaeh claim 6 _9, characterized in that the Signalimpuls® are renewed without phase reversal between unequal characters ("I * ¹ and ^W O ^W ). 8. Transmission system according to Claims 5 to J, characterized in that a pulse repeater comprises the following stages a) a Schwellwertsehaltimg to cancel the Amplitude fluctuations of the incoming signal & ls, b) a phase change stage to bring about the phase separation between unequal characters ("1" and ⁿ O ⁿ ) with the help of a pulse-modulated carrier wave of suitable amplitude and Ruti® a ^ jf the value ^ f / 2 su, β) another 8 @: b $ «lXwsrtech & ltung, 009849 / Λ709 .H. Reeves 92-3-2-1 d) another phase change stage, which increases the relative phase separation of 3f / 2 between unequal Characters ("1" and "0") by adding one reverses the pulse-modulated carrier wave of the corresponding amplitude and phase, e) a third threshold circuit and f) a third phase change stage, the original phase separation (egvfl) between unequal characters ("I" and "O ¹ V) by addition with a pulse-modulated carrier wave of appropriate amplitude and phase. 9. Transmission system according to claim 8, characterized in that a pulse repeater as a final receiver after the third threshold value circuit comprises an inversion which converts the phase-modulated pulses into amplitude-modulated pulses, whereby the signal pulses of one meaning (e.g. ^w 0 ") by addition with a pulse-modulated carrier wave of corresponding amplitude and phase are extinguished, and that this reversing stage is followed by a photodetector stage. 10 · Transmission system according to claims 8 and 9 *, characterized in that pendulum absorbers are used as visual threshold circuits. 009849/1709 AH. Reeves 92-3-2-1 11. Transmission system according to claim 1 to 10, characterized in that a portion of the incoming pulse train is fed to a first Fabry-Perot interferometer (65) whose basic resonance frequency corresponds to the pulse repetition frequency of the Synehronisierimpulse that parts of the locally generated reference wave from this first interferometer with start a periodicity which corresponds to the pulse repetition frequency of the Synehronisierimpulse, and that these parts of the reference wave are fed to a second Fabry-Perot interferometer whose basic resonance frequency corresponds to the pulse repetition frequency of the entire transmission system. 12. Transmission system according to claim 2 to 10 * characterized in that that a portion of the incoming pulse train controls a pendulum oscillator (63) which has a feedback loop to Mitnate © by the synchronization impulses and its oscillation frequency is slightly smaller than the pulse repetition rate of the synchronization pulses that the feedback loop is a Fabry-Perot interferometer contains whose fundamental resonance frequency is the pulse repetition frequency corresponds to the synchronization impulses and allows parts of the locally generated Besugswelle to pass through, and that part of these output signals of the interferometer is used as an auxiliary control signal for the oscillator, while the other part is a second Fabry-Perot interferometer is supplied, whose ßrundresonanzfrequenz the pulse repetition frequency of the entire Corresponds to the transmission system. 09849/1709 --23 - AH. Reeves 92-3-2-1 13. Transmission system according to claim 1 to 12, characterized in that the pendulum oscillator is constructed from a single semiconductor body. 14. Transmission system according to claim I3, characterized in that the semiconductor body has an output surface which carries a non-reflective coating which consists of a multilayered stack with great reflectivity on a permeable dielectric with a layer thickness of half a wavelength, the refractive index of which is close to the real Proportion of the refractive index of the active material of the semiconductor body is adapted. 15. Transmission system according to claim 14, characterized in that the dielectric layer is connected to the outer surface of the semiconductor body. 009849/1709