DE10222170B4 - Quantum optical interference device and method for operating such - Google Patents

Abstract

Quantum optical interference device comprising
A photon source (1) which emits pairs of photons, each consisting of a first photon of medium wavelength λ ₁ and a second photon of medium wavelength λ ₂ ,
- An interferometer with two arms, between which an optical path length difference (D), wherein
The first photon of a photon pair is able to pass through the first arm and the second photon of the photon pair is able to pass through the second arm,
- In the first arm, a first optical delay means (11) is arranged, after passing the first photon is either in a first forward or delayed in relation to this by a first delay d1 first rear probability wave packet (51, 52), which given is by d1 = (n · λ _1/2) + x1, wherein x1 is a value from the interval ₁ -λ /4...+λ _1/4 and n is an integer,
- In the second arm, a second optical delay device (111) is arranged, after passing through ...

Description

Technical area:

The The invention relates to a quantum optical interference device and a method of operating such.

State of the art:

The Information technology demands ever new methods of data processing and data transmission. One of these methods is data processing by quantum computers; Such computers are more intense in recent times the subject Research and development efforts become. It could be shown that quantum computers have some problems computer science are able to solve much faster than conventional ones electronic computers. Examples of such problems are the Prime factorization big Numbers or the calculation of the discrete logarithm with the algorithm from Shor, which on conventional electronic computers is inefficient and therefore much computing time takes up.

The solution these problems by means of conventional electronic computer therefore presents difficulties in practice which is the basis of various cryptological procedures for transmission secret data forms. Quantum computers are therefore potentially suitable to break these cryptological procedures. Not least therefore, they have recently become a center of research interest become.

Furthermore Quantum computers are capable of performing many other tasks needed To reduce time. The asymptotic runtime can be done by means of of the quantum computer are often reduced to a period of time which the square root of the runtime of corresponding algorithms for conventional equivalent to electronic computer.

One well-known example of this is the Grover search algorithm. A classic algorithm needed in the Means n / 2 queries (steps) to a specific record using a search criterion in a list of n records Find. However, Grover's quantum algorithm requires only a number of Steps that correspond to the root of the number n.

quantum mechanical approaches are not only of great interest in computer development, but also in data transmission technology, and there especially in the field of cryptography.

Known Methods of quantum cryptography are based on the one-time-key method (One Time Pad, Vernam Cypher). This procedure can be done with conventional Computers are performed and is proven for sure. requirement for the procedure is that the key secret, random and as long as the message to be transmitted itself is. encoded is done with a bitwise logical XOR operation. The transmitter and the receiver must have the same key for this, So that the secret transmission this key is crucial. This is exactly what quantum cryptography does (Quantum Key Distribution): The bits of the key are bitwise into individual Photons encoded; a listening attempt on these photons, however, this is due to quantum mechanical Change principles, and this change can be from the sender and the receiver be detected.

These Examples show that the Quantum mechanics in information technology is gaining importance. Since logical gates are an essential foundation of computer and computer technology Represent information technology, are quantum mechanical logic gates of great Interest.

A Application of cryptography consists in the identification of two Parties, which only together and possibly only simultaneously an action against a third party, which does not come into the possession of the key. conventional Procedures in which the keys not directly to the examining Party transfer are usually based on the fact that the calculation of the discrete Logarithm with the algorithm of Shor on conventional electronic computers is inefficient (see, for example, Schneier, Bruce: Applied Cryptography, 2nd, John Wiley & Sons, 1996, ISBN: 0-471-11709-9. Especially for But quantum computers have efficient algorithms for this Problem, so that obstructed by use of quantum computers can not and the security of the mentioned identification procedures not more guaranteed is.

Various methods for generating photon pairs are known in the art. One method is based on nonlinear optics. It utilizes a quantum optical effect based on optical parametric fluorescence which results in the formation of so-called "multiphoton Fock states." For this purpose, photons from a laser, referred to below as primary photons, are irradiated into a crystal suitable for nonlinear optics. For example, the crystal may consist of beta-barium borate, potassium deuterium phosphate, or lithium niobate, and the primary photon is likely to be optically parame- tered when passing through the crystal converted trical fluorescence into a set of two secondary, "entangled" photons whose total energy corresponds to the energy of the primary photon. The wavelength of each secondary photon is therefore larger than that of the primary photon.

In Literature becomes the secondary Photon with greater energy as "signal photon", the one with the smaller energy called "Mitläufer" or "Idler". A details Description of said effect provides the publication "Quantum Phenomena in the world of light "from J. Brendel, Series Physik Volume 28 Thun [ua], 1994, Frankfurt (Main), Univ., Diss., P. 41 ff. [ISBN: 3-8171-1384-6.] The Photon Packages can without further on conventional Way initiated in optical fibers and telecommunications technology are processed. The achieved efficiencies and intensities at The generation of such photon packets are so far very low. However, there are already applications for such today Light that gets along with these low intensities, e.g. in the area of quantum cryptography.

A another method for generating photon pairs is to use a two-photon laser as a light source. A two-photon laser is e.g. in the publication "Polarization Instabilities in a Two-Photon Laser "by O. Pfister et al., Physical Review Letters, Vol. 86, No. 20, pp. 4512-4515, May 2001.

From the DE 198 23 849 A1 For example, a method and apparatus for generating selective single photons or pairs of photons in an optical channel is known wherein a two-photon state (photon pair) is generated and the photon pair is spatially separated to obtain the quantum mechanical correlation if the photons are collinearly emitted. One photon each is coupled into an optical channel, wherein one channel contains an interferometer with a variable optical path length difference and the other channel contains an optical delay path. Both channels are spatially merged by a beam splitter. The optical path lengths and the delay path are tuned such that the probability K for coincidences between the outputs of the beam splitter is optionally about K = 0 or about K = 1 or about a predetermined intermediate value, where K = 0 a photon pair in one of the output channels of the beam splitter and K = 1 corresponds to two single photons in both output channels of the beam splitter.

Of the Invention is the object of a device and a To provide methods for quantum optical execution of a logical XOR operation e.g. suitable in a quantum computer and for the identification of two Opposite parties apply to a third party.

These The object is achieved by a quantum optical interference device with the features of Claim 1. The object is further achieved by a method with the features of claim 35.

• – eine Photonenquelle, welche Photonenpaare abgibt, deren jedes aus einem ersten Photon von mittlerer Wellenlänge λ₁ und einem zweiten Photon von mittlerer Wellenlänge λ₂ besteht,
• – ein Interferometer mit zwei Armen, zwischen welchen ein optischer Weglängenunterschied D besteht, wobei
• – das erste Photon eines Photonenpaares den ersten Arm und das zweite Photon des Photonenpaares den zweiten Arm zu durchlaufen imstande ist,
• – in dem ersten Arm eine erste optische Verzögerungseinrichtung angeordnet ist, nach deren Passieren sich das erste Photon entweder in einem ersten vorderen oder in einem gegenüber diesem um eine erste Verzögerung d1 verzögerten ersten hinteren Wahrscheinlichkeitswellenpaket befindet, welche gegeben ist durch d1 = (n·λ₁/2) + x1, wobei x1 ein Wert aus dem Intervall –λ₁/4...+λ₁/4 und n eine ganze Zahl ist,
• – in dem zweiten Arm eine zweite optische Verzögerungseinrichtung angeordnet ist, nach deren Passieren sich das zweite Photon entweder in einem zweiten vorderen oder in einem gegenüber diesem um eine zweite Verzögerung d2 verzögerten zweiten hinteren Wahrscheinlichkeitswellenpaket befindet, welche gegeben ist durch d2 = (m·λ₂/2) + x2, wobei x2 ein Wert aus dem Intervall –λ₂/4...+λ₂/4 und m eine ganze Zahl ist,
• – einen Strahlenkoppler mit einem ersten und einem zweiten Kopplereingang und einem ersten und einem zweiten Kopplerausgang, wobei
• – das erste Photon nach Durchlaufen des ersten Arms durch den ersten Kopplereingang und das zweite Photon nach Durchlaufen des zweiten Arms durch den zweiten Kopplereingang in den Strahlenkoppler einfällt,
• – die beiden Photonen das Photonenpaares in dem Strahlenkoppler interferieren, und
• – jedes der beiden Photonen des Photonenpaares den Strahlenkoppler sowohl durch den ersten als auch durch den zweiten Kopplerausgang verlassen kann, so daß entweder die Photonen des Photonenpaares den Strahlenkoppler gemeinsam durch denselben Kopplerausgang verlassen – Antikoinzidenz des Photonenpaares – oder getrennt voneinander durch verschiedene Kopplerausgänge verlassen – Koinzidenz des Photonenpaares –,
• – einen ersten Detektor, welcher das erste bzw. das zweite Photon zu registrieren imstande ist, sofern das erste bzw. zweite Photon den Strahlenkoppler durch den ersten Kopplerausgang verlassen hat,
• – einen zweiten Detektor, welcher das erste bzw. das zweite Photon zu registrieren imstande ist, sofern das erste bzw. zweite Photon den Strahlenkoppler durch den zweiten Kopplerausgang verlassen hat, sowie
• – eine Koinzidenzschaltung, welche ein Koinzidenzsignal auslöst, wenn der erste Detektor das erste Photon des Photonenpaares und der zweite Detektor das zweite Photon des Photonenpaares registriert oder umgekehrt, d.h. bei Feststellung einer Koinzidenz des Photonenpaares, und ein Antikoinzidenzsignal auslöst, wenn der erste oder der zweite Detektor beide Photonen des Photonenpaares registriert, d.h. bei Feststellung einer Antikoinzidenz des Photonenpaares,

wobei der optische Weglängenunterschied D so gewählt ist, daß mit einer signifikant höheren Wahrscheinlichkeit als 50% eine logische XOR-Operation ausgeführt wird, welche darin besteht, daß die Koinzidenzschaltung das Koinzidenzsignal auslöst, wenn eine der beiden Zahlen n oder m eine gerade und die andere eine ungerade Zahl ist, und das Antikoinzidenzsignal auslöst, wenn beide Zahlen n und m gerade Zahlen oder beide Zahlen n und m ungerade Zahlen sind.A quantum optical interference device according to the invention comprises

• A photon source emitting photon pairs each consisting of a first photon of medium wavelength λ ₁ and a second photon of medium wavelength λ ₂ ,
• - An interferometer with two arms, between which an optical path length difference D exists, wherein
• The first photon of a photon pair is able to pass through the first arm and the second photon of the photon pair is able to pass through the second arm,
• - In the first arm, a first optical delay device is arranged, after passing the first photon is either in a first forward or in a relative to this delayed by a first delay d1 first rear probability wave packet, which is given by d1 = (n · λ _1/2) + x1, wherein x1 is a value from the interval ₁ -λ /4...+λ _1/4 and n is an integer,
• A second optical delay device is arranged in the second arm, after which the second photon passes either in a second forward or in a second rear likelihood wave packet delayed by a second delay d2, which is given by d2 = (mλ) _2/2 ) + x2, where x2 is a value from the interval -λ ₂ /4...+λ _2/4 and m is an integer,
• - A beam coupler having a first and a second coupler input and a first and a second coupler output, wherein
• The first photon passes through the first coupler input after passing through the first arm and the second photon passes through the second coupler input into the beam coupler after passing through the second arm,
• - The two photons interfere with the photon pair in the beam coupler, and
• - Each of the two photons of the photon pair can leave the beam coupler through both the first and through the second coupler output, so that either the photons of the photon pair leave the beam coupler together through the same coupler output - anti-coincidence of the photon pair - or separated from each other by different coupler exits - coincidence of the photon couple -,
• A first detector capable of registering the first and second photons, if the first or second photon has left the beam coupler through the first coupler output,
• - A second detector which is capable of registering the first and the second photon, provided that the first and second photon has left the beam coupler through the second coupler output, and
• A coincidence circuit which triggers a coincidence signal when the first detector registers the first photon of the photon pair and the second detector registers the second photon of the photon pair or vice versa, ie upon detection of a coincidence of the photon pair, and an anti-coincidence signal if the first or the second Detector detects both photons of the photon pair, ie upon detection of an anticoincidence of the photon pair,

wherein the optical path length difference D is chosen to perform with a significantly higher probability than 50% a logical XOR operation, which consists of the coincidence circuit triggering the coincidence signal if one of the two numbers n or m is even and the other is an odd number and triggers the anti-coincidence signal if both numbers n and m are even numbers or both numbers n and m are odd numbers.

deviant From this, the coincidence circuit can be arranged to have no Anti-coincidence signal triggers. The logical XOR operation then is that the coincidence circuit triggers the coincidence signal, if one of the two numbers n or m is one straight and the other is an odd number, with an absence of the coincidence signal can be interpreted as the occurrence of anticoincidence.

alternative For this purpose, the coincidence circuit can be set up so that it does not Coincidence signal triggers. The logical XOR operation then is that the coincidence circuit triggers the anti-coincidence signal, if both numbers are n and m are even numbers or both numbers are n and m are odd numbers, with the absence of the anticoincidence signal can be interpreted as the occurrence of a coincidence.

The both latter embodiments the device according to the invention are particularly useful if it is known at what times crossings of photon pairs through the interferometer.

• – einer Photonenquelle, welche Photonenpaare abgibt, deren jedes aus einem ersten Photon von mittlerer Wellenlänge λ₁ und einem zweiten Photon von mittlerer Wellenlänge λ₂ besteht,
• – einem Interferometer mit zwei Armen, zwischen welchen ein optischer Weglängenunterschied D besteht, wobei
• – das erste Photon eines Photonenpaares in den ersten Arm und das zweite Photon des Photonenpaares in den zweiten Arm eingekoppelt wird,
• – in dem ersten Arm eine erste optische Verzögerungseinrichtung angeordnet ist, nach deren Passieren sich das erste Photon entweder in einem ersten vorderen oder in einem gegenüber diesem um eine erste Verzögerung d1 verzögerten ersten hinteren Wahrscheinlichkeitswellenpaket befindet, welche gegeben ist durch d1 = (n·λ₁/2) + x1, wobei x1 ein Wert aus dem Intervall –λ₁/4...+λ₁/4 und n eine ganze Zahl ist,
• – in dem zweiten Arm eine zweite optische Verzögerungseinrichtung angeordnet ist, nach deren Passieren sich das zweite Photon entweder in einem zweiten vorderen oder in einem gegenüber diesem um eine zweite Verzögerung d2 verzögerten zweiten hinteren Wahrscheinlichkeitswellenpaket befindet, welche gegeben ist durch d2 = (m·λ₂/2) + x2, wobei x2 ein Wert aus dem Intervall –λ₂/4...+λ₂/4 und m eine ganze Zahl ist,
• – und mit einem Strahlenkoppler mit einem ersten und einem zweiten Kopplereingang und einem ersten und einem zweiten Kopplerausgang, wobei
• – die beiden Photonen das Photonenpaares nach Durchlaufen des ersten bzw. zweiten Interferometerarmes in einen Strahlenkoppler eingekoppelt und dort zur Interferenz gebracht werden, und
• – jedes der beiden Photonen des Photonenpaares den Strahlenkoppler sowohl durch einen ersten als auch durch einen zweiten Kopplerausgang verlassen kann, so daß entweder die Photonen des Photonenpaares den Strahlenkoppler gemeinsam durch denselben Kopplerausgang verlassen – Antikoinzidenz des Photonenpaares – oder getrennt voneinander durch verschiedene Kopplerausgänge verlassen – Koinzidenz des Photonenpaares –, wird
• – ein Koinzidenzsignal dann ausgelöst, wenn das erste Photon des Photonenpaares den Strahlenkoppler durch den ersten Kopplerausgang und das zweite Photon des Photonenpaares den Strahlenkoppler durch den zweiten Kopplerausgang verläßt oder umgekehrt, d.h. bei Koinzidenz des Photonenpaares,
• – ein Antikoinzidenzsignal dann ausgelöst, wenn das erste und das zweite Photon des Photonenpaares den Strahlenkoppler gemeinsam durch den ersten oder durch den zweiten Kopplerausgang verlassen, d.h. bei Koinzidenz des Photonenpaares,
• – und der optische Weglängenunterschied D so gewählt, daß mit einer signifikant höheren Wahrscheinlichkeit als 50% eine logische XOR-Operation ausgeführt wird, welche darin besteht, daß das Koinzidenzsignal ausgelöst wird, wenn eine der beiden Zahlen n oder m eine gerade und die andere eine ungerade Zahl ist, und das Antikoinzidenzsignal ausgelöst wird, wenn beide Zahlen n und m gerade Zahlen oder beide Zahlen n und m ungerade Zahlen sind.

In a method according to the invention for operating a quantum optical interference device with

• A photon source emitting photon pairs each consisting of a first photon of medium wavelength λ ₁ and a second photon of medium wavelength λ ₂ ,
• - An interferometer with two arms, between which an optical path length difference D consists, wherein
• - the first photon of a photon pair is coupled into the first arm and the second photon of the photon pair is coupled into the second arm,
• - In the first arm, a first optical delay device is arranged, after passing the first photon is either in a first forward or in a relative to this delayed by a first delay d1 first rear probability wave packet, which is given by d1 = (n · λ _1/2) + x1, wherein x1 is a value from the interval ₁ -λ /4...+λ _1/4 and n is an integer,
• A second optical delay device is arranged in the second arm, after which the second photon passes either in a second forward or in a second rear likelihood wave packet delayed by a second delay d2, which is given by d2 = (mλ) _2/2 ) + x2, where x2 is a value from the interval -λ ₂ /4...+λ _2/4 and m is an integer,
• - And with a beam coupler having a first and a second coupler input and a first and a second coupler output, wherein
• - The two photons coupled the photon pair after passing through the first and second Interferometerarmes in a beam coupler and there brought to the interference, and
• - Each of the two photons of the photon pair can leave the beam coupler both through a first and through a second coupler output, so that either the photons of the photon pair leave the beam coupler together through the same coupler output - anti-coincidence of the photon pair - or leave each other separated by different coupler outputs - coincidence of the photon pair -, will
• A coincidence signal is triggered when the first photon of the photon pair leaves the beam coupler through the first coupler output and the second photon of the photon pair leaves the beam coupler through the second coupler output, or vice versa, ie if the photon pair coincides,
• An anticoincidence signal is triggered when the first and the second photon of the photon pair leave the beam coupler in common through the first or the second coupler output, ie in the case of coincidence of the photon pair,
• - And the optical path length difference D so It is chosen to perform with a significantly higher probability than 50% a logical XOR operation, which consists of triggering the coincidence signal if one of the two numbers n or m is even and the other is an odd number, and the anti-coincidence signal is triggered if both numbers n and m are even numbers or both numbers n and m are odd numbers.

deviant thereof is according to a Variant of the invention, no anti-coincidence signal is triggered. The logical XOR operation is then that the coincidence circuit the Coincidence signal triggers, if one of the two numbers n or m is one straight and the other is an odd number, where a lack of coincidence signal can be interpreted as the occurrence of anticoincidence.

In an alternative variant, no coincidence signal is triggered. The logical one XOR operation is then that the coincidence circuit the anti-coincidence signal triggers, if both numbers are n and m are even numbers or both numbers are n and m are odd numbers, with the absence of the anticoincidence signal can be interpreted as the occurrence of a coincidence.

The two latter variants of the method are in particular then makes sense if it is known at which times passages of Photon pairs take place through the interferometer.

Especially the coincidence signal can be a high signal and the anticoincidence signal be a low signal or vice versa.

Of the release of the coincidence signal may be a set bit, logical one, and that of the anti-coincidence signal is a bit not set, logical Zero, to be assigned. The XOR operation is then that when straightened Number n and odd number m or conversely a logical one, at each even or odd numbers n, m outputs a logical zero becomes. However, alternatively, the triggering of the coincidence signal an unset bit, logic zero, and that of the anticoincidence signal a set bit, logical one, are assigned. Then exists the XOR operation is that at even number n and odd number m or conversely a logical one Zero, for each even or odd number n, m is a logical one One is issued; such a variant of an XOR operation can also be called an inverse XOR operation be designated.

The inventive method is thus based on two-photon interference. According to the invention optical path length difference D and the integers n and m are chosen so that the first and the second Photon interfere in the beam coupler. Classic would be expected that probability for this, that this first photon the beam coupler through the first coupler output and the second photon the beam coupler through the second coupler output leaves or conversely, exactly 50%. Due to quantum optical effects, the probability depends for coincidence however, from the optical path length difference D and of the first and second delay d1, d2 and thus in particular from the numbers n and m. Therefore, the probability depends for coincidence in particular, whether one of the two numbers n, m is even and the other is an odd number or both numbers are n, m each are even numbers or odd numbers.

The dependence the probability of coincidence from the optical path length difference D is in each of these cases relatively complicated. Wearing one said probability against the optical path length difference as a curve, so this curve shows a plurality of minima and Maxima, i. of extreme values. One gets due to the photon pair interference a so-called fourth order interference pattern. The curve as well as the shape and location of extreme values changes when the first or second delay d1, d2 are changed.

These Interference effects are exploited according to the invention. If appropriate Choice of optical path length difference D is the probability for Coincidence significantly higher then 50%, if n is an even number and m is an odd number, or vice versa, where a main maximum exists, in which for a given value of the optical path length D a coincidence probability of 100% is achieved.

Of the Beam coupler may e.g. a polarizing beam splitter or be a fusion coupler. According to one preferred embodiment the beam coupler is arranged so that in the classical case, i. without interference in its interior, a photon, which in the first Input of the coupler, the beam coupler with a probability of substantially each leaves 50% through the first or the second coupler output, and a photon which enters the second coupler input, the Beam coupler also with a probability of substantially each leaving 50% through the first or the second coupler output, i. that in the essentially no coupler input and no coupler output preferred are.

According to a preferred variant, which has the advantage of improved separation of the said extreme values is the value x1, a value from the interval ₁ -λ /8...+λ _1/8, preferably from the In interval -λ ₁ /1h...+λ ₁ / 16th, and the value x2, a value from the interval -λ ₂ /8...+λ _2/8, preferably in the interval -λ _2/16 ... + λ ₂ / ib.

Just Numbers n, m can as logical input variables 1 and odd numbers n, m are interpreted as logical input variables 0 of the XOR operation be, or vice versa. The coincidence signal can be used as the output of the XOR operation be used, where triggered coincidence signal as a logical 1 and a triggered one Anti-coincidence signal can be interpreted as a logical 0. The XOR operation in this case is that equal inputs 0.0 or 1.1 with a significantly higher Probability as 50% to a logical output 0 and unequal input values 0.1 or 1.0 1 with a significantly higher probability than 50% lead to a logical output 1.

According to one embodiment each of the first and second optical elements is arranged to that the Numbers n and m can only assume the values 0 or 1 respectively.

The Probability of that this Coincidence signal triggered if n is an even number and m is an odd number or vice versa, can by suitable adaptation of the optical path length difference D far over 50%, namely up to a maximum of 100%. According to a variant of the invention becomes the optical path length difference S so chosen that the Probability for Coincidence higher than 75%, preferably higher than 99% is when n is an even and m is an odd number or vice versa. In this case, the coincidence signal becomes more likely than 75%, or preferably with a higher Probability as 99%, triggered when one of the two numbers n or m one straight and the other of the two numbers n or m one odd number is.

A certain statistical uncertainty arises because even then, if n is an even number and m is an odd number or vice versa, the probability for Coincidence even with optimal matching of the optical path length difference D though approximately, but not 100% mathematically exactly. Further statistical Uncertainty arises from the fact that on average a certain Proportion of photons absorbed when passing through the interferometer is, as the transmittance of the interferometer of course not can reach 100% full. The probability that a particular Photon pair, for which holds that n one is even and m is an odd number, or vice versa, the coincidence signal indeed triggers, Therefore, in practice, it can easily reach values above 95%. but always at least slightly below of 100%.

Around to meet the statistical uncertainties is, according to a Variant of the invention a counter when coincidence is incremented, it is decremented in the case of anticoincidence vice versa, where the coincidence signal is only triggered when a certain predetermined first count is exceeded, and the Anti-coincidence signal is only triggered when a specific predetermined second meter reading is fallen below, or vice versa. For this purpose, the coincidence circuit a counter which it increments upon detection of coincidence and decremented upon detection of anticoincidence, or vice versa, wherein the coincidence circuit triggers the coincidence signal only when a certain predetermined first count is exceeded, and the Anti-coincidence signal only triggers when a given given second meter below is, or vice versa.

According to one modified variant, a counter is incremented at coincidence is decremented at Antikoinzidenz or vice versa, where the coincidence signal only triggered if, within a predetermined first count period, a certain predetermined first count exceeded is, and the anti-coincidence signal is only triggered if within a given second counting period a certain predetermined second meter reading is fallen below, or vice versa. For this purpose, the coincidence circuit a counter which it increments upon detection of coincidence and upon detection of a coincidence decrements, or vice versa, wherein the coincidence circuit the Coincidence signal only triggers if within a given first counting time period a certain predetermined first meter reading exceeded is, and the anti-coincidence signal only triggers when within a predetermined second counting period falls below certain predetermined second count, or vice versa.

By a sufficiently high specification of the meter reading, which for triggering the Coincidence signal or the Antikoinzidenzsignals leads may At the same time an arbitrarily high statistical security can be achieved.

It different photon sources are known, which photon pairs submit.

According to a preferred embodiment of the invention, a primary laser of suitable wavelength and a nonlinear optical crystal are used as the photon source, the primary laser irradiating the crystal with a beam of primary photons and optical parametrization in the crystal generates fluorescence photon pairs, which emerge from the crystal and are each in a two-photon Fock state, wherein the first and the second delay are each chosen smaller than the coherence length of the primary photons. In particular, this process can be carried out in such a way that the two photons of a pair of photons emerge from the crystal in different directions, so that it is possible, with only a small amount of equipment, to place the first photon of each pair in the first arm and the second photon in the second arm of the interferometer. In this case, the primary laser can be a continuous light taser or a pulse laser. The primary laser is also referred to as a "pump laser". For example, the crystal may be beta barium borate, potassium deuterium phosphate, or lithium niobate.

The Photons of a photon pair thus generated are in several ways correlated and crossed. The corresponding light paths, i. the associated Interferometer arms are often called signal and idler arms. The Coherence length of the first and second photons can be when using such photon pair sources typically 10 ... 500 μm, the wavelength of the first and second photons may e.g. each 1.3 microns.

According to one another embodiment the photon source is a two-photon laser through which the Photon pairs are generated. According to another embodiment the photon source is one in which quadrupole transitions or Cascade transitions take place through which the photon pairs are generated. According to another embodiment is the photon source one in which the Coulomb blockade effect occurs by which the photon pairs are generated.

The first optical delay device can a first retardation plate and the second optical delay device a second delay plate include.

in the quantum physical wave image, the first photon represents a probability wave packet in the spatial space, which are split by the first delay device is placed in a first front and a first rear probability wave packet. The first front and the first rear probability wave packet are a delay d1 shifted against each other so that the total probability of residence of the first photon in space now has two maxima, which by the distance d1 in the direction of movement of the first photon from each other are spaced. Similarly, the overall residence probability of the second photon in space after passing the second delay means two maxima, namely a front and a rear maximum, which by a distance d2 spaced apart in the direction of movement of the second photon are.

The first or second optical delay device can a uniaxial or multiaxial birefringent first or second Crystal, wherein none of the optical axes of the first or second crystal to the propagation direction of the first and second Photons in the first and second crystal arranged in parallel is. The relative size of the two Maxima hangs for both Photons each of the orientation between the Polrarisationsrichtung of the photon and the optical axes of the crystal. Especially can the first and second retardation plates, respectively each be a birefringent crystal, its slow and its fast axis in each case perpendicular to the propagation direction of the first and second photons and with the polarization direction each of the first and second photons has an angle between 40 ° and 50 °, preferably 45 °, form. In the case of the preferred angle of 45 ° are the front and the rear Maximum of the probability of residence of the first or second Photons same size.

The birefringent crystals are preferably uniaxial and so on arranged that both their slow as well as their fast axis each perpendicular to the propagation direction of the first and second photons.

in the Wave image becomes the location of a photon through a probability wave described, wherein the photon in a probability wave packet located. After passing the birefringent crystal, whose slow and fast axis are aligned as indicated above, the photon is parallel with a certain probability polarized to fast axis and with a complementary thereto probability polarized parallel to the slow axis. If the photon is parallel polarized to the fast axis, it passes through the birefringent Crystal faster than when polarized parallel to slow axis is. The probability wave packet entering the birefringent crystal splits as it passes through it into two parts, namely in a front and a delayed against this rear Part. In the birefringent crystal, therefore, the two run up these major axes projected portions of the photon due to the Double calculation apart.

To Passing the birefringent crystal is therefore the Photon either in a front or in front of this one delayed rear probability shaft package.

The first photon is therefore after passing the first birefringent Crystal, whose slow and fast axis are aligned as stated above, either in a first front or in a first across from this one for a first delay delayed first rear probability wave packet. Likewise is the second photon after passing the second birefringent crystal, its slow and fast axis are also aligned like stated above, either in a second front or in one across from this by a second delay delayed second rear probability wave packet.

Of the first and second delay plate is according to one preferred refinement of this variant in each case a polarizer upstream, so that the first and the second photon before hitting the first and second, respectively Retarder plate polarized become.

The birefringent crystals can be polarizers be upstream. Alternatively, such a photon source becomes which already uses first and second photons each with a fixed polarization direction, which is e.g. in many Photon sources are the case in which a primary laser and a non-linear optical crystal are used, wherein in the crystal is generated by optical parametric fluorescence photon pairs become.

Around the above-described quantum-optical XOR operation successively for different Execute number pairs n, m to be able to have to the delays d1, d2 can be changed over time be. According to one embodiment comprises hence the first optical delay device a first optical control with a first control input, via which the first delay d1 is controllable so that the Number n is an even or odd number, if at the first control input a first and a second voltage value is applied, and the second optical delay device a second optical control having a second control input via which the second delay d2 is controllable so that the Number m is an even or odd number, if at the second control input a third or a fourth voltage value is applied. The first delay d1 will over the first control input controlled so that the number n is a straight or odd number is, if at the first control input the first or the second voltage value is applied, and the second is applied via the second control input controlled so that the number m is a straight or odd number is, if at the second control input the third resp. the fourth voltage value is applied.

Of the Number n is thus assigned a voltage value, which to the first Control input is applied and thereby affects the first delay d1. certain Voltage values correspond to even values of the number n, other voltage values odd numbers of the number n. The number n will thus be at a distance d1 between the front and the back maximum of the probability of residence of the first photon or, in other words, in "coded" the first interferometer arm. If you interpret, As stated above, even values for n as logical "1" and odd values for n as logical "0", or vice versa, can be in this way, any bit code in the first interferometer encode.

In Analogous to this, according to the in Presented embodiment the invention of the number m assigned a voltage value, which at the second control input is applied and thereby the second delay d1 affected. The number m is thus at a distance d2 between the front and the rear maximum of the probability of residence of the second Photons "coded". If you interpret, as stated above, even values for m as logical "1" and odd values for m as logical "0", or vice versa, that's how it works in this way, any bitcode even in the second interferometer encode.

The called bitcodes according to the invention the above explained XOR operation together connected become. For this purpose, the method according to another variant is repeated several times executed in succession, being the first delay d1 is controlled in time so that the number n is a first predetermined Sequence of even and odd numbers, according to the first bit sequence, and synchronously, the second delay d2 so controlled in time will that the Number m is a second predetermined sequence of even and odd numbers goes through according to the second bit sequence, so that the change of the resulting trips of the coincidence signal and the anti-coincidence signal of the logical XOR the first and the second sequence corresponds, if even numbers n or m respectively as logical one and odd numbers n and m respectively each interpreted as logical zero or vice versa.

According to a modification of this variant, in this case the first delay d1 and the second delay d2 are controlled in time so that exactly one element of the first and exactly one element of the second sequence is omitted for each photon pair entering the interferometer or passing through each of the interferometers. According to another modification of this variant, the first delay d1 and the second delay d2 are controlled in time so that each incident on the interferometer or on each passing through the interferometer Photon pair at least one element of the first and one element of the second sequence is omitted.

According to one Further modification of this variant, the first delay d1 and the second delay d2 timed so that on each entering the interferometer or on each the interferometer passing photon pair averages at least one element of the first and one element of the second sequence is omitted.

The Bits of the two bit codes become the last three modifications in each case bitwise or on average bitwise in each individual Photons encoded. To intercept a bit is the intercepting party thus exactly one photon or at most a photon or on average at most a photon available. This photon then stands for an XOR operation no longer or only in changed Form available; A listening attempt This is because of these photons change based on quantum mechanical principles, and this change can be from the sender and the receiver be detected.

According to one Advantageous variant is always the next bit the respective bit codes in the first and second interferometer when either the coincidence signal or the anti-coincidence signal triggered has been. To do this in sync for to allow both interferometer arms can the triggering of the coincidence signal or anti-coincidence signal, e.g. in the manner of a "hand-shaking" operation over one Feedback line reported back to the two parties become. The encodings can be made so that on each incident into the interferometer or to each the interferometer passing photon pair exactly one bit or on average one bit each of the two bitcodes is omitted. The bits of the two bit codes are bitwise or in this case Means coded bit by bit into individual photons; the hereby for quantum mechanical reasons associated significant advantages in terms of privacy have already been explained above.

One Soleil compensator is an optical retardation element with two wedge-shaped crystals, wherein the delay changed when one of the two wedge-shaped crystals opposite the others in a certain direction. These Displacement can be achieved by means of an electrically controlled piezo-crystal respectively. As a first or second optical control element can therefore a with a first and second piezo crystal provided first or second Soleil compensator can be used, wherein the first or second delay by applying the first or second voltage value to the first Piezo crystal or by applying the third or fourth voltage value the second piezoelectric crystal is electrically controlled. The input quantities n, m The XOR operation is thus by appropriate adjustment of the compensators predetermined. The compensators can also be designed as Babinet compensators.

According to one Other embodiments of the Invention are the first and second optical delay device a Mach-Zehnder interferometer or a Michelson interferometer or a Fabry-Perot interferometer or an etalon or an echelon each with two Interferometerzweigen, where the gait difference between the interferometer branches mechanically and / or electro-optically and / or is magneto-optically adjustable.

According to one preferred variant is the first optical delay device a first polarizer and / or the second optical delay device followed by a second polarizer, so that the first photon after passing the first optical delay device and / or the second photon after passing the second optical delay means each polarized. The advantage here is that only such first or second photons get into the beam coupler, which have a certain polarization direction.

Especially can Advantageously, the first photon after passing the first optical delay means and / or the second photon after passing the second optical delay means each polarized so that the polarization direction of the first photon upon its arrival in the beam coupler less than 5 °, preferably less than 1 ° of the polarization direction of the second photon upon its arrival differs in the beam coupler. For this purpose, the first and the second polarizer be oriented so that the the polarization direction of the first photon when it arrives in the beam coupler by less than 5 °, preferably by less than 1 ° from the Polarization direction of the second photon on its arrival in the beam coupler differs. This variant has the advantage an improved separation of the already explained above extreme values.

The polarizers can be designed in particular as polarizing beam splitters. In one embodiment of the invention, therefore, the first polarizer is a first polarizing beam splitter having a first and a second beam splitter output and the second polarizer is a second polarizing beam splitter having a third and a fourth beam splitter output, wherein the first beam splitter output in the first Coupler input and the second beam splitter output in a first register device, which is capable of registering the first photon, and the third beam splitter output terminates in the second coupler input and the fourth beam splitter output terminates in a second recorder capable of registering the second photon.

These Variant is advantageous for the following reason: if the first photon of a photon pair through the first polarizing beam splitter leaves the second beam splitter output, or if the second photon of the photon pair through the second polarizing beam splitter leaves the fourth beam splitter output, the photon pair in question goes loss of a photon and thus can no longer lead to a coincidence, but also no longer lead to an anticoincidence. If, however, the leftover Photon this photon pair reaches one of the detectors and this, which is common in practice the case is unable to distinguish whether a single Photon alone or both photons of a photon pair invaded are, it becomes erroneous detected an anticoincidence and thus the result of XOR operation falsified. These cases can be recognized as such with the help of the variant in question.

By an advantageous further refinement of the latter variant these are cases Not only recognized, but it also becomes an erroneous statement anticoincidence avoided. According to this refinement are the first and the second recorder are connected to the coincidence circuit and put it temporarily in a non-functional state when the first recorder is the first one Photon or the second recorder registered the second photon. This is an erroneous, detection of anticoincidence based on a single photon only locked out.

In a preferred variant is therefore the first and / or the second optical delay device a polarizing beam splitter having a first and a second Downstream of the beam splitter output, wherein the first beam splitter output in the beam coupler and the second beam splitter output in a Recorder opens; which the first and second photon are able to register and with the Coincidence circuit, the same by the recorder temporarily is set to a nonfunctional state when the recorder is the first one or second photon registered.

Of course you can in place of the first and second registration devices single recorder be used, which corresponds to both the second and the fourth beam splitter output is optically connected.

According to one another embodiment of the invention the first and / or the second optical delay means a partially transparent plate.

The first or second optical delay device can be a first or second electro-optical or magneto-optical Phase modulator comprising an adjustment of the first or second delay allowed. In this case, the first and / or the second delay advantageously by the first and second electro-optical or magneto-optical Phase modulator affected become. The phase modulators can e.g. based on the Kerr effect. To control the first or second delay may be at the first and second phase modulator, a first or second electrical control signal are applied. The input variables of the XOR operation are thus beyond the Control signals can be specified.

According to one preferred variant, the first and second delay are greater than the coherence length of the first or second photons selected. This variant has the advantage of improved separation of the explained above Extreme values.

According to one Refinement of this variant with even further improved separation the extreme values become the first and second delay greater than twice, preferably larger than three times the coherence length of the first or second photons selected.

In a preferred embodiment the device according to the invention Therefore, the first and second delay means are arranged so that the first or second delay is larger as the coherence length of the first or second photons. According to others Refinements of this embodiment are the first and second delay means arranged so that the first or second delay greater than twice, preferably larger than three times the coherence length of the first and second photons, respectively, so that the front probability wave packets with the respective associated practically do not overlap rear probability wave packets, but completely separated , thereby improving the separation of the extreme values explained above becomes.

According to one Another preferred variant of the invention, the first delay d1 and the second delay d2 chosen so that she at least the coherence length of the first photons or at least to distinguish the coherence length of the second photon. This variant also has the advantage of improved separation the above explained Extreme values.

According to one another variant, the first delay (d1) and the second Delay (d2) always chosen that the Case n = m does not occur.

According to one Variant becomes the optical path length difference D so chosen that yourself the mean value of the two optical path lengths, which is the first front or the first rear probability wave packet in the first one Arm through, from the mean of the two optical path lengths, which the second front and the second rear probability wave packet, respectively go through in the second arm to differ at most a maximum value, which is given by three times the coherence length of the first photon or by three times the coherence length of the second photons. Preferably, the maximum value is then through the Threefold the coherence length of the given first photons, if this is smaller than the coherence length of the second photon and otherwise by three times the coherence length of the given second photons.

According to one preferred variant, the optical path length difference D is chosen so that the Average of the two optical path lengths, which is the first front or the first rear probability wave packet in the first one Arm through, from the mean of the two optical path lengths, which the second front and the second rear probability wave packet, respectively in the second arm, by at most such a maximum value which is given by a quarter of the coherence length of the first photons or by a quarter of the coherence length of the second photon. Preferably Here, the maximum value is then by a quarter of the coherence length of given first photons, if this is smaller than the coherence length of the second photons, and otherwise by a quarter of the coherence length of the second Given photons.

one or both arms of the interferometer may be complete or at least partially be formed by optical waveguides. By such a configuration the operation of a device according to the invention in practice facilitated. The interferometer may e.g. as a Mach-Zehnder interferometer be educated.

The intensity the current of first photons through the first arm and the intensity of the second Photons through the second arm can be replaced by one in the first or second arm interposed or upstream of the interferometer Absorbers are reduced. For this purpose, between the photon source and the interferometer or between the primary laser and the nonlinear one optical crystal or in at least one of the arms of the interferometer an absorber be interposed, which is a part of it absorbs impinging photons. By such intermediary Absorber can in particular the intensity of the photon pair current so customized be that at the coding of bitcodes in the interferometer arms on average only one photon on each bit causing certain illegitimate manipulations, especially an unnoticed monitor the bitcodes coded into the interferometer arms, of quantum mechanical establish impossible become.

Further can the polarization and / or the phase position and / or the frequency of the first and second photons by one in the first and second, respectively Arm intermediate optical component to be changed. To this end can be between the photon source and the interferometer or between the primary laser and the nonlinear optical crystal or at least one the arms of the interferometer an optical component interposed which is the polarization and / or the phase angle and / or the frequency of the optical component passing through photons able to change is.

According to one another embodiment has the device according to the invention not two detectors, namely the first and the second detector, but comes with only one of these detectors. According to this embodiment Therefore, the device has the first and the second instead Detector, a single detector on which either downstream of the first or the second coupler output and both Photons of the photon pair is able to register and that Coincidence signal then triggers if it registers only one of the two photons of the photon pair, and the anticoincidence signal then triggers when it has both photons of the photon pair registered. There is a coincidence in it recognized that the Single detector registers only one of the photons of the photon pair, and an anticoincidence is recognized by the fact that it contains both photons of the photon pair registered. The single detector must be able to distinguish whether only one of the photons of the photon pair is incident or both.

Application exists the invention further in a method with the features of Claim 64.

• a) Die erste Partei A erhält einen ersten geheimen Bitschlüssel aS, umfassend eine Folge von ersten Schlüsselbits,
• b) die zweite Partei B erhält einen zweiten geheimen Bitschlüssel bS, umfassend eine Folge von zweiten Schlüsselbits,
• c) der Kontrolleur C erhält einen dritten geheimen Bitschlüssel cS, welcher gegeben ist durch die bitweise logische XOR-Verknüpfung zwischen dem ersten und dem zweiten Bitschlüssel, so daß cS = aS XOR bS gilt,
• d) die erste Partei A setzt den ersten Bitschlüssels aS in eine zeitliche Abfolge von Werten für die ganze Zahl n um, indem die erste Partei A die erste Verzögerung d1 zeitlich so steuert, daß gesetzte Bits des ersten Bitschlüssels aS in geradzahlige Werte für die Zahl n und nicht gesetzte Bits des ersten Bitschlüssels aS in ungeradzahlige Werte für die Zahl n umgesetzt werden bzw. umgekehrt,
• e) die zweite Partei B setzt synchron hierzu den zweiten Bitschlüssel bS in eine zeitliche Abfolge von Werten für die ganze Zahl m um, indem die zweite Partei B die zweite Verzögerung d2 zeitlich so steuert, daß alle gesetzten Bits des zweiten Bitschlüssels bS in geradzahlige Werte für die Zahl m und alle nicht gesetzten Bits des zweiten Bitschlüssels bS in ungeradzahlige Werte für die Zahl m umgesetzt werden bzw. umgekehrt, so daß jedes solchermaßen umgesetzte Bit des ersten Bitschlüssels aS mit dem synchron hierzu umgesetzten Bit des zweiten Bitschlüssels bS ein Bitpaar bildet und zu jedem Bitpaar genau ein Wert für die Zahl n sowie genau ein Wert für die Zahl m gehört,
• f) der Kontrolleur C erfaßt synchron hierzu die Auslösungen des Koinzidenzsignals und des Antikoinzidenzsignals und setzt diese um in eine Ergebnis-Bitfolge cE, indem er jeder Auslösung des Koinzidenzsignals ein gesetztes Bit und jeder Auslösung des Antikoinzidenzsignals ein nicht gesetztes Bit zuweist,
• g) die Ergebnis-Bitfolge cE wird mit dem dritten Bitschlüssel cS verglichen, h) das durch die erste und die zweite Partei gebildete Parteienpaar A, B wird als identifiziert eingestuft, sofern mindestens ein bestimmter vorgegebener Grad an Übereinstimmung zwischen der Ergebnis-Bitfolge cE und dem dritten Bitschlüssel cS vorliegt.

Thus, the invention further consists in a method for identifying a party pair consisting of a first party and a second party, to a controller, using the method explained above or using the device explained above, comprising the following steps:

• a) The first party A receives a first secret Bit key aS, comprising a sequence of first key bits,
• b) the second party B receives a second secret bit key bS comprising a sequence of second key bits,
• c) the controller C receives a third secret bit key cS, which is given by the bitwise logical XOR operation between the first and the second bit key, so that cS = aS XOR bS,
• d) the first party A converts the first bit key aS into a time series of values for the integer n by the first party A timing the first delay d1 so that set bits of the first bit key aS are converted into even numbers for the number n and unset bits of the first bit key aS are converted into odd numbers for the number n and vice versa,
• e) the second party B synchronously converts the second bit key bS into a time series of values for the integer m by having the second party B time-control the second delay d2 such that all set bits of the second bit key bS are even-numbered for the number m and all non-set bits of the second bit key bS are converted into odd-numbered values for the number m or vice versa, so that each thus converted bit of the first bit key aS with the synchronized converted bit of the second bit key bS forms a bit pair and for each pair of bits exactly one value for the number n and exactly one value for the number m belongs,
• f) the controller C synchronously detects the resolutions of the coincidence signal and the anti-coincidence signal and converts them into a result bit sequence cE by assigning a set bit to each triggering of the coincidence signal and assigning an unset bit to each triggering of the anti-coincidence signal,
• g) the result bit sequence cE is compared with the third bit key cS, h) the party pair A, B formed by the first and the second party is classified as identified if at least a certain predetermined degree of agreement between the result bit string cE and the third bit key cS is present.

A Application of this method consists in the pairwise identification two parties a third, this third party is not in the possession of both involved keys arrives. For example, the method can then be used advantageously if only two parties are entitled to an action at the same time are.

conventional Identification procedures where the keys are not transmitted directly to the verifying party are usually based on the fact that the discrete logarithm to calculate a number with classical computers only inefficiently is [Schneier: Applied Cryptography]. But there are just for quantum computers efficient algorithms for this problem that these identification methods become uncertain to let.

The Inventive identification method has the advantage that it even with the use of quantum computers no loss of his Security suffers. Across from approaches, which are based on single-photon interference, has the identification method according to the invention the advantage of insensitivity to phase fluctuations between the two poor people involved.

The first and second party as well as the controller can be people, as well Computer or other electronic devices. The implementation of the first Bitschlüsssels aS into a temporal sequence of values for the integer n and the Implementation of the second Bitschlüsssels bS take place in a time sequence of values for the integer m synchronous to each other, i. the first bit of the first bit key aS becomes synchronous with the first bit of the second bit key bS, the second bit of the first bit key aS synchronous with the second one Bit of the second bit key bs etc. implemented. To achieve such synchronicity, a synchronization device may be used which e.g. a central trigger signal or clock signal to both parties emits.

According to one Variant, the first party A controls the first delay d1 and the second party B times the second delay d2 so that on each entering the interferometer or on each the interferometer passing photon pair exactly one bit pair or at least one Bit pair is omitted.

According to one In another variant, the first party A controls the first delay d1 and the second party B times the second delay d2 so that on each entering the interferometer or on each the interferometer passing photon pair on average at least one bit pair is omitted.

that is, that the first party A the first delay d1 and the second party B control the second delay d2 in time, that on each incident into the interferometer or to each the interferometer passing photon pair exactly one bit or at most one bit or on average at the most one bit of the first or second bit key is omitted. The bit flow in bits per second, band rate, is in this case at least as large or in the Medium at least as big as the photon flux in Photons per second.

Application exists the invention further in a method with the features of Claim 67.

• a) Die erste Partei A erhält einen ersten geheimen Bitschlüssel aS, umfassend eine Folge von ersten Schlüsselbits, und modifiziert diesen zu einer ersten Prüf-Bitfolge aP, indem die erste Partei A eine erste Mehrzahl X von Bits des ersten Bitschlüssels aS, welche sich an willkürlich oder zufällig gewählten ersten Prüfbit-Positionen innerhalb des ersten Bitschlüssels aS befinden, durch willkürlich oder zufällig gewählte erste Prüfbits ersetzt,
• b) die zweite Partei B erhält einen zweiten geheimen Bitschlüssel bS, umfassend eine Folge von zweiten Schlüsselbits, und modifiziert diesen zu einer zweiten Prüf-Bitfolge bP, indem die zweite Partei B eine zweite Mehrzahl Y von Bits des zweiten Bitschlüssels bS, welche sich an willkürlich oder zufällig gewählten zweiten Prüfbit-Positionen innerhalb des zweiten Bitschlüssels bS befinden, durch willkürlich oder zufällig gewählte zweite Prüfbits ersetzt,
• c) der Kontrolleur C erhält einen dritten geheimen Bitschlüssel cS, welcher gegeben ist durch die bitweise logische XOR-Verknüpfung zwischen dem ersten und dem zweiten Bitschlüssel, so daß cS = aS XOR bS gilt,
• d) die erste Partei A setzt die erste Prüf-Bitfolge aP in eine zeitliche Abfolge von Werten für die ganze Zahl n um, indem die erste Partei A die erste Verzögerung d1 zeitlich so steuert, daß gesetzte Bits der ersten Prüf-Bitfolge aP in geradzahlige Werte für die Zahl n und nicht gesetzte Bits der ersten Prüf-Bitfolge aP in ungeradzahlige Werte für die Zahl n umgesetzt werden bzw. umgekehrt,
• e) die zweite Partei B setzt synchron hierzu die zweite Prüf-Bitfolge bP in eine zeitliche Abfolge von Werten für die ganze Zahl m um, indem die zweite Partei B die zweite Verzögerung d2 zeitlich so steuert, daß gesetzte Bits der zweiten Prüf-Bitfolge bP in geradzahlige Werte für die Zahl m und nicht gesetzte Bits der zweiten Prüf-Bitfolge bP in ungeradzahlige Werte für die Zahl m umgesetzt werden bzw. umgekehrt, so daß jedes solchermaßen umgesetzte Bit der ersten Prüf-Bitfolge aP mit dem synchron hierzu umgesetzten Bit der zweiten Prüf-Bitfolge bP ein Bitpaar bildet und zu jedem Bitpaar genau ein Wert für die Zahl n sowie genau ein Wert für die Zahl m gehört, wobei die erste Partei (A) die erste Verzögerung (d1) und die zweite Partei (B) die zweite Verzögerung (d2) zeitlich so steuern, daß auf jedes in das Interferometer einfallende oder auf jedes das Interferometer durchlaufende Photonenpaar genau ein Bitpaar oder mindestens ein Bitpaar oder im Mittel mindestens ein Bitpaar entfällt,
• f) der Kontrolleur C erfaßt synchron hierzu die Auslösungen des Koinzidenzsignals und des Antikoinzidenzsignals und setzt diese um in eine dritte Prüf-Bitfolge cP, indem er jeder Auslösung des Koinzidenzsignals ein gesetztes Bit und jeder Auslösung des Antikoinzidenzsignals ein nicht gesetztes Bit zuweist,
• g) die dritte Prüf-Bitfolge cP wird mit dem dritten Bitschlüssel cS verglichen,
• h) das durch die erste und die zweite Partei gebildete Parteienpaar A, B wird als identifiziert eingestuft, sofern mindestens ein bestimmter vorgegebener Grad an Übereinstimmung zwischen der dritten Prüf-Bitfolge cP und dem dritten Bitschlüssel cP vorliegt,
• i) mindestens ein Teil der ersten Prüfbit-Positionen wird mit mindestens einem Teil der zweiten Prüfbit-Positionen verglichen, woraus sich eine Menge von Z Übereinstimmungs-Positionen ergibt, an welchen jeweils eine der ersten Prüfbit-Positionen mit einer der zweiten Prüfbit-Positionen zusammenfällt, so daß zu jeder Übereinstimmungs-Position genau ein Bit der ersten Prüf-Bitfolge aP, ein Bit der zweiten Prüf-Bitfolge bP und ein Bit der dritten Prüf-Bitfolge cP gehört,
• j) für jede Übereinstimmungs-Position oder einen Teil derselben wird durch logische XOR-Verknüpfung zwischen dem zugehörigen Bit der ersten Prüf-Bitfolge aP und dem zugehörigen Bit der zweiten Prüf-Bitfolge bP ein Übereinstimmungs-Bit bestimmt,
• k) mindestens ein Teil der Übereinstimmungs-Bits wird mit jeweils zugehörigen Bits der dritten Prüf-Bitfolge cP verglichen, und
• l) der erste und der zweite Bitschlüssel werden als abgehört eingestuft, falls der Verfahrensschritt k) ergibt, daß mindestens eine vorgegebene kritische Zahl oder mindestens ein vorgegebener kritischer Anteil von Übereinstimmungs-Bits von dem jeweils zugehörigen Bit der dritten Prüf-Bitfolge cP abweicht.

Thus, the invention further consists in a method for identifying a party pair consisting of a first party and a second party, to a controller, using the method explained above or using the device explained above, comprising the following steps:

• a) The first party A receives a first secret bit key aS comprising a sequence of first key bits, and modifies it to a first check bit string aP by the first party A a first plurality X of bits of the first bit key aS, which abuts arbitrarily or randomly selected first check bit positions within the first bit key aS are replaced by arbitrarily or randomly selected first check bits,
• b) the second party B receives a second secret bit key bS comprising a sequence of second key bits and modifies it to a second check bit string bP by the second party B a second plurality Y of bits of the second bit key bS, which abuts located arbitrarily or randomly selected second check bit positions within the second bit key bS, replaced by arbitrarily or randomly selected second check bits,
• c) the controller C receives a third secret bit key cS, which is given by the bitwise logical XOR operation between the first and the second bit key, so that cS = aS XOR bS,
• d) the first party A converts the first check bit sequence aP into a time sequence of values for the integer n by the first party A timing the first delay d1 so that set bits of the first check bit string aP are even-numbered Values for the number n and unset bits of the first check bit sequence aP are converted into odd numbers for the number n or vice versa,
• e) the second party B synchronously converts the second check bit string bP into a time series of values for the integer m by the second party B timing the second delay d2 such that set bits of the second check bit string bP are converted into even-numbered values for the number m and unset bits of the second check bit sequence bP into odd-numbered values for the number m, respectively, vice versa, so that each bit thus converted is the first check bit sequence aP with the bit of the second one synchronized therewith Test bit sequence bP forms a bit pair and for each pair of bits exactly one value for the number n and exactly one value for the number m belongs, where the first party (A) the first delay (d1) and the second party (B) the second Timing the delay (d2) such that each photon pair entering the interferometer or passing through each interferometer has exactly one bit pair or at least one bit pair, or at least one on average Bit pair omitted,
• f) the controller C synchronously detects the resolutions of the coincidence signal and the anticoincidence signal and converts them into a third check bit sequence cP by assigning a set bit to each triggering of the coincidence signal and assigning an unset bit to each triggering of the anticoincidence signal,
• g) the third check bit sequence cP is compared with the third bit key cS,
• h) the party pair A, B formed by the first and the second party is deemed to be identified, provided that there is at least a certain predetermined degree of agreement between the third check bit string cP and the third bit key cP,
• i) at least a portion of the first check bit positions is compared to at least a portion of the second check bit positions, resulting in a set of Z match positions at each of which one of the first check bit positions coincides with one of the second check bit positions such that exactly one bit of the first check bit sequence aP, one bit of the second check bit sequence bP and one bit of the third check bit sequence cP belong to each matching position,
• j) for each match position or a part thereof, a match bit is determined by logical XORing between the associated bit of the first check bit string aP and the associated bit of the second check bit string bP,
• k) at least a part of the match bits is compared with respective bits of the third check bit string cP, and
• l) the first and the second bit key are classified as intercepted, if the method step k) shows that at least one predetermined critical number or at least one predetermined critical portion of match bits deviates from the respectively associated bit of the third test bit sequence cP.

Due to the condition that the first party A controls the first delay d1 and the second party B times the second delay d2 such that each photon pair entering the interferometer or passing through each interferometer will have exactly one bit pair or at least one bit pair or mean If at least one bit pair is omitted, one intercepting party has exactly one photon or at most one photon, or on average only one photon available to intercept a bit. This photon is then no longer available for an XOR operation. By the above condition, it is thus advantageously achieved that interception of the test bit sequences coded into the interferometer arms is falsified by an unauthorized party for quantum mechanical reasons, resulting in the result of the XOR operation associated with the respective photon pair, thereby discovering the eavesdropping attempt can be.

The Number of check bits in the first and in the second check bit sequence can be the same or different from each other. According to a variant of the method are each between 10% and 50% of the first test bit sequence aP and the second Check bit sequence bP Check bits.

The Implementation of the test bit sequences aP, bP in time sequences of values for the integers n and m occur synchronously with each other, i. the first bit of the first check bit string aP becomes synchronous with the first bit of the second check bit sequence bP, the second bit of the first check bit sequence aP synchronously with the second bit of the second check bit sequence bP, etc. implemented. To such a synchronicity can be achieved, a synchronization device can be used e.g. a central trigger signal or clock signal to both parties emits. In particular, for synchronization of the method steps d), e) and f) an electrical or electronic synchronization device be used with which the first and second party as well the controller are connected.

According to one advantageous variant is the number of first key bits, the first check bits, the second key bits and the second check bits each chosen so large that the probability that less as a number of L match positions is smaller than a given limit probability, where L is an arbitrary is positive integer. The average expected number of match positions grows both with the average density of check bits within the check bit strings as well as with the length the same.

The number L is greater than 10, preferably greater than 100, selected according to a variant of the invention. For a given average density of the check bits within the check bit sequences, a given average length L for the test bit sequences results for a given number L. Likewise, by appropriate choice of said densities and lengths, an arbitrarily small probability can be achieved that there are fewer than a number of L match positions. The limit probability is set according to a variant smaller than 10 ^-3 , preferably smaller than 10 ^-6 .

If in step i), a number of match positions result, which is smaller than a predetermined minimum value, is determined according to a Variant of the party formed by the first and the second party A, B are not classified as identified, but the bit-key aS, bS, cS with new bitcodes replaced and the process with the same again.

A Possibility, the first and second delay To control d1, d2 in time so that the interferometer on each passing photon pair exactly one bit pair is eliminated, exists in that the Both parties A, B always the next bit of the first or second Bitschlüssels convert to integers n or m if either the coincidence signal or triggered the anti-coincidence signal has been. To enable this procedure, the triggering of the Coincidence signal or anti-coincidence signal e.g. in the manner of a "hand-shaking" operation over one Feedback line reported back to the two parties become.

The Bits of the first and those of the second bit key in this case bitwise or on average bitwise in each individual Photons encoded; the hereby connected for quantum mechanical reasons considerable advantages in terms of privacy have already been explained above. The feedback can be advantageous at the same time for the synchronization of the parties A and B serve.

In Modification to the above explained Method, the third secret bit cS by the bitwise logical NXOR link, i.e. through the inverse XOR operation between the first and second bit-keys (aS, bS) such that cS = aS NXOR bS applies, where the controller C synchronously to the triggers of the Coincidence signal and the anti-coincidence signal detected and this into a result bit sequence cE by triggering each triggering of the Coincidence signal is a bit not set and any triggering of the Anti-coincidence signal assigns a set bit.

Summary the drawing, in which show:

1 a schematic representation of an embodiment of a quantum optical interference device according to the invention,

2 an example of the splitting probability of a photon into two separate wave packets,

3 Examples of the dependence of the coincidence probability on the optical path length difference,

4 a schematic representation of another embodiment of a quantum optical interference device according to the invention,

5 Examples of bit keys for illustrating an application of the invention, and

6 and 7 Examples of check bit strings, check bit positions, and match positions to illustrate another application of the invention.

1 shows a schematic representation of an embodiment of a quantum optical interference device according to the invention with a photon source 1 which emits pairs of photons, each consisting of a first photon of medium wavelength λ ₁ and a second photon of average wavelength λ ₁ . The photon source 1 consists of a primary laser 2 and a nonlinear optical crystal 3 ,

For generating photon pairs is in the device of 1 exploits a quantum optical effect, which is based on optical parametric fluorescence and leads to the formation of so-called "multiphoton Fock states." For this purpose, a beam of primary photons 2A from the primary laser 2 in the crystal suitable for non-linear optics 3 irradiated. The crystal 3 may for example consist of beta-barium borate, potassium deuterium phosphate or lithium niobate. The primary photons 2A become when passing through the crystal 3 with a certain probability converted by optical parametric fluorescence in a package of two secondary photons, hereinafter called the first and second photon whose total energy corresponds to the energy of the primary photon. The wavelength of each secondary photon is therefore larger than that of the primary photon. The energy distribution between the secondary photons within the photon packet is not always the same, but statistically distributed and given by a probability distribution. In particular, the secondary photons can have the same energy, which means that both have half the wavelength of the primary photon.

In particular, this process can be carried out in such a way that the two photons of a photon pair emerge from the crystal in different directions, so that it is easily possible to couple the first photon of each pair into the first arm and the second photon into the second arm of an interferometer. as it is in 1 is shown schematically. The first arm is through a first optical delay device 11 and a first mirror 15 , the second arm by a second optical delay device 111 and a second mirror 115 educated. The light paths between these components may be formed by optical fibers, which in 1 but not shown. Both arms of the interferometer unite in a beam coupler 30 ,

There is an optical path length difference D between the two arms, which is adjusted by adjusting the first or second mirror 15 . 115 is changeable.

In the first arm is a first optical delay device 11 after passing through the first photon is either in a first forward or in a relation to this delayed by a first delay d1 first rear probability wave packet. The first optical delay device 11 is so arranged that the first delay d1 has the value d1 = (n · λ _1/2) + x1, wherein x1 is a value from the interval ₁ -λ /4...+λ _1/4 and n is an integer is.

Likewise, in the second arm is a second optical delay device 111 after passing through the second photon of the photon pair is either in a second forward or in a relation to this delayed by a second delay d2 second rear probability wave packet. The second optical delay device 111 is set up such that the second delay d2 is the value d2 = (m * λ _2/2 ) + x ₂ , where x _{2 is} a value from the interval -λ _2/4 ... + λ _3/4 and m is an integer ,

The splitting of the probability of residence of a photon into a forward and a separate rear probability wave packet will be discussed below with reference to FIG 2 explained with reference to an example.

Preferably, the intervals for the values of x1, x2 chosen narrower, so that x1 -λ example, a value from the interval ₁ /16...+λ ₁ / 16th and x2, a value from the interval -λ _2/16 ... + λ is _2/16. Furthermore, the first and the second delay d1, d2 are preferably each chosen smaller than the coherence length of the primary photon. In addition, the first and second delay d1, d2 are preferably selected to be greater than three times the coherence length of the first and second photons.

The beam coupler 30 has a first and a second coupler input 31 . 131 and a first and a second coupler output 32 . 132 , the first arm of the interferometer opens into the first coupler input 31 such that the first photon passes through the first coupler input after passing through the first arm 31 and the beam coupler passes. Likewise, the second arm opens into the other coupler input 131 such that the second photon passes through the second coupler input after passing through the second arm 131 in the beam coupler 30 incident.

The interferometer is arranged so that the first and the second photon in the beam coupler 30 interfere. The beam coupler 30 is set up so that after the interference of each of the two Photons of the photon pair the beam coupler 30 both through the first and the second coupler output 32 . 132 can leave.

The photons of the photon pair can thus the beam coupler 30 after the interference either together through the same coupler output 32 . 132 which is referred to hereinafter as the anticoincidence of the photon pair, or they can separate the beam coupler from each other through different coupler outputs 32 . 132 which is referred to below as the coincidence of the photon pair.

The first coupler output 32 opens into a first detector 33 , which is capable of registering the first and the second photon, respectively, provided that the first and second photon, respectively, the beam coupler 30 through the first coupler output 32 has left. Likewise, the second coupler output opens 132 in a second detector 133 , which is capable of registering the first and the second photon, respectively, provided that the first and second photon, respectively, the beam coupler 30 through the second coupler output 132 has left.

Both detectors 33 . 133 are with a coincidence circuit 40 which triggers a coincidence signal when the first detector 33 the first photon of the photon pair and the second detector 133 the second photon of the photon pair registered or vice versa, ie, upon detection of a coincidence of the photon pair. Furthermore, the coincidence circuit triggers 40 an anti-coincidence signal when one of the detectors 33 . 133 both photons of the photon pair registered, ie upon detection of an anticoincidence of the photon pair. The coincidence signal and the anti-coincidence signal are output 40A issued.

However, due to quantum optical effects, the likelihood of coincidence depends on the optical path length difference D and on the first and second delays d1, d2 ( 3 ). Therefore, the probability of coincidence depends in particular on whether one of the two numbers n, m is even and the other is an odd number, or whether both numbers n, m are even numbers or odd numbers.

In this case, it is possible to choose the optical path length difference D between the two arms for the purpose of utilizing the aforementioned quantum optical effects such that, according to the invention, a significantly higher probability than 50% is performed by a logical XOR operation consisting in the coincidence circuit 40 triggers the coincidence signal if one of the two numbers n or m is even and the other an odd number, and triggers the anticoincidence signal with the same probability if both numbers n and m are even numbers or both numbers n and m are odd numbers.

For example, the first and the second delay d1, d2 can be selected such that n is an even number, eg n = 20, and m is an odd number, eg n = 71. The logical XOR operation is then that the coincidence signal is triggered with a significantly higher probability than 50%. With a suitable choice of the optical path length difference D, a probability of approximately 100% can be achieved. The anti-coincidence signal is triggered with a complementary probability, which differs only slightly from zero in the present example. If the triggering of the coincidence signal is interpreted as logical 1, then the quantum-optical XOR operation according to the invention consists in the fact that the combination of the input values 20 and 71 , which in the present example (n = 20, m = 71) by the device of 1 takes place, with a significantly higher probability than 50% of the baseline 1 results.

In another example, the first and second delays d1, d2 are chosen such that n is an even number, eg, n = 20, and m is also an even number, eg, n = 70. The logical XOR operation then is that the anticoincidence signal is triggered with a significantly higher probability than 50%. Again, a probability of approximately 100% can be achieved. The coincidence signal is triggered with a complementary probability, which differs only slightly from zero in this example. If the triggering of the anticoincidence signal is interpreted as logical 0, then the quantum-optical XOR operation according to the invention consists in linking the input values 20 and 70, which in the present example (n = 20, m = 70), by the device of FIG 1 takes place, with a significantly higher probability than 50% of the baseline 0 results.

Different expressed is in the latter two examples by the inventive method a quantum optical XOR operation performed by which with a significantly higher probability than 50% of that initial value is achieved, which is a conventional logical XOR operation equivalent.

2 illustrates an example of the splitting of the residence probability of the first photon into two separate wave packets 51 . 52 , In the quantum physical wave pattern, a photon represents a probability wave packet in space. The probability wave packet 50 of the first photon is due to the first delay approximately facility 11 from 1 split into a first front and a first back probability wave packet 51 . 52 , The first delay device 11 exists in the example of 2 from a delay plate 12 and one of these downstream first polarizer 14 , The first delay plate 12 is in the example of 2 a single-axis birefringent crystal whose slow and fast axes L, S are each perpendicular to the propagation direction of the first photon and with the polarization direction of the same each form an angle of about 45 °. In the first delay plate 12 The two parts of the first photon projected onto these principal axes diverge on the basis of the double calculation. According to a preferred embodiment, the thickness of the first retarder plate 12 chosen so large that the first delay d1 is greater than the coherence length of the first photon.

After passing the first retarder plate 12 If the first photon then consists of two practically non-overlapping parts, but have different polarization. With sufficient thickness of the retarder plate 12 The photon now consists of two practically non-overlapping parts. The downstream polarizer 14 projects these two shares in the example of 2 back to the original polarization.

The first front and the first rear probability wave packet 51 . 52 are shifted by a delay d1 against each other, so that the total probability of residence of the first photon in space now has two maxima which are spaced apart by the distance d1 in the direction of movement of the first photon.

In completely analogous manner, the probability of residence of the second Photons are split into a second front and a second rear probability wave packet (not shown) which around the second delay d2 are spaced from each other.

3 shows four examples of the dependence of the coincidence probability on the optical path length difference. The curves of 3 are so-called interference patterns of the fourth order, are based on calculations and all relate to the special case that the values x1 and x2 are each equal to zero. In this case, the first and second deceleration d1, d2 in the case of even values for the numbers n and m are each an integer multiple of the wavelength, for odd numbers for the numbers n and m, respectively an integer multiple of the wavelength plus a half wavelength ,

The curve top left in 3 For example, if n is an even number and m is an odd number, the upper-right curve for the case where n is an odd and m is an even number is the lower-left graph for the case where n and m are both odd numbers and the lower right corner for the case that n and m are both even numbers. One sees from the curves of 3 in that the dependence of the probability of coincidence on the optical path length difference D in each of these cases is relatively complicated and each curve has a plurality of extreme values.

In particular, each of the four curves of 3 a major extreme 61 . 62 . 63 . 64 of finite width, where the coincidence probability is in the respective center of the major extreme values 61 . 62 . 63 . 64 either 100% (fully constructive interference, upper curves of 3 ) or 0% (completely destructive interference, lower curves of 3 ) reached. The location of the main extreme values 61 . 62 . 63 . 64 , with respect to the optical path lengths difference D, agrees in all four figures: the centers of the two main maxima 61 . 62 and the centers of the two main minima 63 . 64 each correspond to the special case that the average of the two optical path lengths which pass through the first front and the first rear probability wave packet in the first arm, with the average of the two optical path lengths, the second front and the second rear probability wave packet in the second Go through arm, exactly matches. This special case corresponds in the two upper curves of 3 in the case of a coincidence probability of 100%, in the two lower curves of 3 a coincidence probability of 0%. This behavior corresponds to the logical exclusive-or (XOR).

However, even with an optical path length difference D, which deviates from this special case, it is in the range of the main extreme values 61 . 62 . 63 . 64 a coincidence probability significantly different from 50%; for example, an even value will result in the number n and an odd value for the number m (upper left graph of 3 ) with a significantly higher probability than 50% to coincidence and thus to the triggering of the coincidence signal, if the optical path length difference D in the region of the main maximum 61 is selected. At the cost of statistical uncertainty, therefore, the non-central areas of the extremes can be left 61 . 62 . 63 . 64 to perform a quantum optical XOR operation.

Another example: Even values for both numbers n, m (lower right curve of 3 ) result in the same choice of optical path length difference D (which in the lower right corner of FIG 3 the area of the main minimum 64 ent speaks) with a significantly lower probability than 50% to coincidence, therefore with a significantly higher probability than 50% to anti-coincidence and triggering of the anti-coincidence signal.

For the quantum-optical XOR operation, in practice, only the range of the main extreme values 61 . 62 . 63 . 64 readily usable, but not the ranges of the other extreme values of the curves of 3 , According to a preferred variant of the invention, therefore, the optical path length difference D is chosen such that the mean value of the two optical path lengths, which are the first front and the first rear probability wave packet, respectively 51 . 52 in the first arm, from the average of the two optical path lengths which pass through the second front and the second rear probability wave packet in the second arm, respectively, differs at most a maximum value given by three times the coherence length of the first photon or by the Threefold the coherence length of the second photon.

The associated Anti-coincidence probability is given by the complementary value of the coincidence probability, i.e. by the difference 100% minus coincidence probability.

4 illustrates a schematic representation of another, over that of 1 modified embodiment of a quantum optical interference device according to the invention. The modification to the device of 1 is that the first and the second optical delay device 11 . 111 in the embodiment of 4 each of three components, namely a first and second retarder plate 12 . 112 , a first and second electro-optical phase modulator 13 . 113 and a first and second polarizer, respectively 14 . 114 exist and are controllable.

The function of the first and second retarder plate 12 . 112 and the respective downstream polarizers 14 respectively. 114 corresponds to that already exemplified with reference to 2 was explained. The delay plates 12 . 112 generate a fixed first and second delay, respectively, due to the plate thickness of the retardation plates 12 respectively. 112 is determined. To make these delays controllable and thus changeable, is between the delay plate 12 respectively. 112 and the polarizer 14 respectively. 114 in each case an electro-optical phase modulator 13 respectively. 113 interposed. The phase modulators 13 . 113 each have a control input 13E respectively. 113E over which the delay is variable by applying a control voltage. The first and second delays d1, d2 can hereby be controlled such that the numbers n and m each optionally assume an even or an odd value.

The device of 4 So is a quantum optical XOR gate, which has two electrical input channels 13E . 113E and an electrical output channel 40A has. This quantum-optical XOR gate performs XOR operations (interpreting even values for n and m as logical and odd values for n and m, respectively, as logical zero or vice versa) such that the output value (coincidence signal interpreted as logical 1, anticoincidence signal interpreted as logical 0) with a significantly higher probability than 50% corresponds to the corresponding output value of a classical XOR gate.

Through the polarizers 14 . 114 photons are lost, which are no longer available in the further course. If you use instead of the polarizers 14 . 114 polarizing beam splitters with a recorder on the out reflected output (not shown) so these cases can be detected immediately. The polarizers 14 . 114 Therefore, they can be designed, in particular, as polarizing beam splitters, with the photons selected out of the interferometer arms being incident into a recording device which is capable of registering these photons. The recorder may be connected to the coincidence circuit and temporarily render it inoperable when the recorder registers a photon. This precludes an erroneous determination of anticoincidence based on only a single photon left over from a pair of photons.

The polarizers 14 . 114 do not necessarily need to be linear polarizers; Rather, they can also be designed as circular or elliptical polarizers.

By means of a compensator, the distance of the front probability wave packet from the rear probability wave packet and thus the internal phase between the two parts of the photon can be changed. Instead of the phase modulators 13 . 113 Therefore, for example, also Soleil compensators can be used, the delay is adjustable, for example, with the aid of electrically driven piezo crystals.

The invention is very advantageously applicable to the identification of a pair of parties, consisting of a first party A and a second party B, to a controller C. This is based on 5 exemplified. The first party A receives a first secret bit key aS, consisting of a sequence of first key bits. In the example of 5 this sequence comprises nine key bits which contain the positions P1 ... P9 within the take first bit key aS, ie the first key bit is located at the first position P1, the second key bit at the second position P2, etc. (top row in 5 ).

The second party B receives a second secret bit key bS, consisting of a sequence of second key bits whose number in the example of 5 is also nine and which take the positions P1 ... P9 within the second bit key bS (second row in 5 ). Checker C receives a third secret bit key cS, which is given by the bitwise logical XOR connection between the first and the second bit key, so that cS = aS XOR bS (fourth row in FIG 5 ). These bit keys aS, bS, cS must have been distributed in advance.

In order to identify the party pair A, B according to the invention, the first party A converts the first bit key aS into a temporal sequence of values for the integer n. The first party A controls the first delay d1 in time so that set bits of the first bit key aS , in 5 each characterized by a "1", converted into odd numbers for the number n, and unset bits of the first bit key as, in 5 each characterized by a "0" to be converted into even numbers for the number n.

For this purpose, in particular the device of 4 to be used. Party A controls the at the entrance 13E applied input voltage for this purpose so that the first bit key aS is successively converted into such a sequence of values of the input voltage that the first delay d1 for each set bit of the first bit key aS an odd number for the number n and for each unset bit of the first bit key aS corresponds to an even value for the number n. The bits of the first bit key are thus converted into different delays d1 via the input voltage and via the number n or, in other words, "coded" into the first interferometer arm.

In this case, the party A can, for example, convert set bits of the first bit key aS into the value n = 1 and unset bits of the first bit key aS into the value n = 0, so that only two different values for the number n occur. For the corresponding control of the input voltage, advantageously only the change between two different voltage values is required. With a suitable choice of the thickness of the first retarder plate 12 For example, one of these voltage values may be zero, so that the coding of the first bit key aS in the first interferometer arm only requires a corresponding switching on and off of the input voltage by the party A.

Another possibility is the thickness of the first retarder plate 12 and the sensitivity of the first phase modulator 13 to choose such that a high signal as an input voltage an odd number for the number n, a low signal as an input voltage an even number for the number n entails. In this case, the coding of the first bit key aS into the first interferometer arm requires a corresponding switching of the input voltage between "high" and "low" by the party A. To adapt to the sensitivity of the phase modulator 13 can be between the entrance 13E and the phase modulator 13 a suitable matching circuit be interposed.

Completely the same applies to the coding of the second bit key bS in the second interferometer by the second party B. The second party B synchronously to the first party A the second bit key bS in a temporal sequence of values for the integer m by the second party B the at the entrance 113E applied input voltage controls so that the second bit key bS is converted in synchronism with the first bit key aS in such a sequence of values of the input voltage at the input 113E in that the second delay d2 for each set bit of the second bit key bS corresponds to an odd-numbered value for the number m and for each unset bit of the second bit-key bS to an even-numbered value for the number m.

"Synchronous" means that the Implementation of the two bit keys aS, bS takes place in time so that a bitwise XOR operation the first and the second bit key can be done.

each thus converted bit of the first bit key aS thus forms with the synchronized thereto converted bit of the second bit key bS a pair of bits, with each of these pairs of bits being exactly one value for the number n and exactly one value for the number m belongs.

The controller C synchronously detects the firing of the coincidence signal and the anticoincidence signal and converts them into a result bit string cE (third row in FIG 5 ) assigning a set bit to each triggering of the coincidence signal and assigning an unset bit to each triggering of the anti-coincidence signal. In the example of 5 Therefore, the result bit sequence cE also comprises nine bits.

Subsequently, the result bit sequence cE is compared with the third bit key cS. The party pair A, B formed by the first and the second party is classified as identified only if at least a certain predetermined degree There is a match between the result bit sequence cE and the third bit key cS. In the example of 5 the result bit sequence cE and the third bit key cS are completely identical; The party pair A, B formed by the first and the second party is therefore considered to be identified here.

An identification process according to the invention can thus take place as follows: The photon pair source generates photon pairs, from which in each case one photon arrives at the party A and the party B. Parties A and B each have an optical delay element 11 . 111 ( 1 . 4 ), with which they each encode a bit from their key aS or bS into the respective photon. A decoder, Controller C, contains the beam splitter and detectors; he directly measures the signal aS XOR bS, which he compares with his key cS. If they match, the identification process was successful.

In The practice can be the length the bitkey and of course the number of available positions far greater than nine elected become. Furthermore, the first and second bit keys aS, not necessarily the same length; at different Length of the same is for the identification of the party pair A, B then only the overlap area relevant to the first and the second bit key.

The invention is according to a further variant also very advantageously applicable for identifying the party pair A, B against the controller C, wherein it can be determined whether an attempt has been made to listen to the first or second bit key unlawfully. This is based on the 6 and 7 exemplified, wherein 6 to explain the identification of the pair of parties A, B and 7 to explain the finding of a listening attempt serves.

According to this variant, the first party A receives in turn, as already in the reference to 5 explained method, a first secret bit key aS comprising a sequence of first key bits. Party A now replaces a first plurality X of bits of the first bit key aS located at random or randomly selected first check bit positions within the first bit key aS with arbitrarily or randomly selected first check bits. As a result, the first bit key aS is modified to a first check bit sequence aP.

This is in the first row of 6 exemplified: based on the first bit key aS of 5 ; however, its second, third, fifth and sixth bits have each been replaced by a check bit. In this example, X = 4. These four first check bits are located at the consecutively numbered four first check bit positions PPa1, PPa2, PPa3 and PPa4 (second row in FIG 7 ), which corresponds to the four positions P2, P3, P5 and P6 (bottom rows in 6 and 7 ).

The result of these substitutions is the first check bit sequence aP (each first row of 6 and 7 ). The check bits are in 6 and 7 indicated by bold letters and framing.

There not only the positions of the check bits, but also their content arbitrarily or random chosen The first check bit sequence aP usually differs not at all, but only on a part of the first check bit positions, namely in the present example only at the two check bit positions PPa1 and PPa3, from the first bit key aS. However, this is for the further procedure irrelevant.

Likewise, the second party B in turn receives the second secret bit key bS and modifies this to a second check bit sequence bP (second row in 6 and third row in 7 by the second party B replacing a second plurality Y of bits of the second bit key bS located at arbitrarily or randomly selected second check bit positions within the second bit key bS with arbitrarily or randomly selected second check bits.

This is in the second row in 6 exemplified: based on the second bit key bS of 5 ; its first, third, sixth, seventh and eighth bits, positions P1, P3, P6, P7 and P8, were each replaced by a check bit. The result of these five substitutions is the second check bit sequence bP. In this example, Y = 5. The five second check bits are located at the consecutively numbered five second check bit positions PPb1, PPb2, PPb3, PPb4 and PPb5 (fourth row in FIG 7 ). As a rule, the second check bit sequence bP does not differ at all, but only at a part of the second check bit positions, namely in the present example only at check bit positions PPb1, PPb3, PPb7 and PPa8, from the second bit key bS. However, this is irrelevant to the further process.

In turn receives the controller C the third secret bit key cS, which is given by the bitwise logical XOR connection between the first and second bit-keys aS, bS, such that cS = aS XOR bS applies.

Now, the first party A converts the first check bit string aP into a time series of values for the integer n by having the first party A time-control the first delay d1 such that set bits of the first check bit string aP are odd-numbered Values for the number n and unset bits of the first check bit sequence aP are converted into even values for the number n, ie, the first party A encodes the first check bit string aP into the first interferometer arm. Here, the first party A may proceed as above with reference to 5 explained, where, for example, the device of 4 can be used.

As well the second party B sets the second check bit sequence synchronously bP into a time sequence of values for the integer m um by the second party B controls the second delay d2 in time, that put Bits of the second check bit sequence bP in odd values for the number m and unset bits of the second check bit string bP into even values for the number m will be implemented.

Thus, each bit thus converted forms the first check bit sequence aP with the bit of the second check bit sequence bP synchronized therewith, one bit pair for each bit pair, and exactly one value for the number m. It should again be emphasized that in this case the parties A and B, in contrast to the reference to 5 explained methods do not encode the bit key aS, bS, but the test bit sequences aP, bP in the Interferometerarme.

in this connection the first party A control the first delay d1 and the second party B the second delay d2 in time so that on each incident into the interferometer or to each the interferometer passing photon pair exactly one bit pair or at least one Bit pair or on average at least one bit pair is omitted. hereby will ensure that at the coding of the test bit sequences aP, bP into the interferometer arms exactly one photon pair, at most a photon pair or on average at most on each bit pair is omitted. This turn means that on every bit of the first check bit string aP is exactly one of the first photons or at most one of the first photons or on average at most one of the first photons is omitted, and that on every bit of the second check bit string bP exactly one of the second photons or at most one of the second photons or on average at most one of the second photons is omitted. The bit flow in Bit per second, baud rate, is in this case as large or at least so big or on average at least as big as the photon flux in Photons per second.

Around to listen to a bit, is the listening party under this condition exactly one photon or at most one photon or im Means at most a photon available. This photon then stands for one XOR operation is no longer available or only available in modified form. By the said condition is thus advantageously achieved that a listening to the in the Interferometerarme coded test bit sequences by a this unauthorized party, e.g. by the controller C, from quantum mechanical establish the result of the XOR operation belonging to the respective photon pair is falsified, whereby the listening attempt can be discovered, as explained below.

The controller C, in synchronism with the above-described encoding of the check bit strings aP, bP, detects in the interferometer arms the triggers of the coincidence signal and the anti-coincidence signal and converts them into a third check bit string cP (third row in FIG 6 and seventh row in 7 ) assigning a set bit to each triggering of the coincidence signal and assigning an unset bit to each triggering of the anti-coincidence signal. The third check bit sequence cP is generally not identical to the result bit sequence cE of 5 because not the bit keys aS, bS, but rather the check bit strings aP, bP have been coded into the interferometer arms. Subsequently, the third test bit sequence cP is compared with the third bit key cS. The pair of parties A, B formed by the first and the second party A, B is classified as identified if at least a certain predetermined degree of agreement between the third check bit sequence cP and the third bit key cP is determined.

The remaining process steps that follow now serve to check whether one of the check bit sequences aP, bP has been intercepted, either by the controller C itself, by one of the parties A or B or by another party. For example, controller C could illegally not only receive the output of the coincidence circuit 40 ( 4 ), but also the first and / or second check bit sequence aP, bP, so that they are no longer secret.

The first check bit positions are compared to the second check bit positions (second and fourth rows of 7 ); In this case, it is determined at which positions P1... P9 both a first check bit and a second check bit are present, ie one of the first check bit positions PPa1, PPa2, PPa3, PPa4 with one of the second check bit positions PPb1, PPb2, PPb3, PPb4, PPb5 coincides, this is in the example of 7 at the positions P3 and P6 the case, since at the position P3 both the check bit PPa2 and the check bit PPb2 is present and at the position P6 both the check bit PPa4 and the check bit PPb3 is present. These positions P3, P6 are referred to below as match positions and labeled ÜP1, ÜP2 (fifth row in FIG 7 ). In the present example, therefore, there is a set of two match positions ÜP1, ÜP2, which are at the positions P3 and P6. At the other positions P1, P2, P4, P5, P7, P8 and P9 there is either only one check bit or no check bit at all; these positions are therefore no match positions.

To Each match position ÜP1, ÜP2 is thus included exactly one bit each of the first check bit sequence aP, the second Testing bit sequence bP and the third check bit sequence cP.

After the match positions ÜP1, ÜP2 are determined, a match bit is successively determined for each of them by logically XORing between the corresponding bit of the first check bit string aP and the corresponding bit of the second check bit string bP. For the first coincidence position ÜP1, the matching bit 1 XOR 0 = 1, for the second coincidence position ÜP2 the matching bit 0 XOR 0 = 0 (sixth row in FIG 7 ).

Now become the match bits with the respective associated Bits of the third check bit sequence cP compared. The first and second bit-keys aS, bS are classified as eavesdropped, if this comparison shows that at least one predetermined critical Number Z or at least a predetermined critical portion AZ of the match bits from the respectively associated Bit of the third check bit sequence cP deviates. In the present example, the controller has C in the third check bit string cP at the coincidence position ÜP1 (= position P3) the value 1 (coincidence signal), at the coincidence position ÜP2 (= position P6) the value 0 (anti-coincidence signal). None of the match bits give way So from the respectively associated Bit of the third check bit sequence cP off. Therefore, the first and second bit-keys aS, bS do not become classified as eavesdropped, but you can continue to be considered secret. Otherwise, it may be considered that wrench aS, bS to discard and replace with new bit-key.

Of the Checker C or any other unauthorized party is at This method is also not able by deviating methods of measurement, the participating bit key aS, bS to determine directly and without errors. An attempt to do this can be uncovered by parties A and B at random positions instead of a bit from her key aS or bS is a random one Encode bit. Encode A and B at random at the same time such check bit, so they are able to do so in retrospect publicly with that measured by C. To compare result. A and B give this first (via a public Channel) list and compare the positions of their check bits two lists with each other. For the Positions where A and B have coded check bits at the same time C his measurement results, then from A and B over the public Channel with the check bits be compared. Now tried C in the measurement instead of the XOR signal the bitstreams from A and B to determine directly, so he introduced errors and can for the test positions do not always give the correct answer, so this attempt is discovered. The keys a and b are then considered compromised and must be discarded become.

Industrial Applicability:

The Invention is industrially applicable e.g. in the field of optoelectronics, optical data processing, optical communication and cryptography.

1

photon source

2

primary laser

2A

beam of primary photons

3

nonlinear optical crystal

11 111

first, second optical delay device

12,112

first, second birefringent plate

13 113

first, second optical control

13E, 113E

Control inputs from 13 . 113

15 115

First, second mirror

30

beam coupler

31 131

First, second coupler input

32 132

First, second coupler output

33 133

First, second detector

40

coincidence circuit

40A

Output from 40

51

first front probability wave packet

52

first rear probability wave packet

61,62,63,64

Main extremes

d1 d2

first, second delay

D

difference the optical path length

n, m

whole numbers

λ ₁ , λ ₂

Wavelength of the first, second photons

Claims (74)

A quantum optical interference device, comprising - a photon source ( 1 ), which emits pairs of photons, each of which consists of a first photon of medium wavelength λ ₁ and a second photon of average wavelength λ ₂ , - an interferometer with two arms, between which there is an optical path length difference (D), where - the first photon of a photon pair the first arm and the second photon of the photon pair is able to pass through the second arm, - in the first arm a first optical delay device ( 11 ) is arranged, after passing the first photon either in a first front or in a relation to this a first delay d1 delayed first rear probability wave packet ( 51 . 52 ) Is, which is given ₁ /4...+λ _1/4 and n is by d1 = (n · λ _1/2) + x1, wherein x1 is a value from the interval -λ an integer, - in which second arm a second optical delay device ( 111 After passing, the second photon is located either in a second forward or in a second rear likelihood wave packet delayed by a second delay d2 therefrom, which is given by d2 = (m * λ 2/2) + x ₂ , where x2 is a value from the interval -λ ₂ /4..+λ _2/4 and m is an integer, - a beam coupler ( 30 ) with a first and a second coupler input ( 31 . 131 ) and a first and a second coupler output ( 32 . 132 ), wherein - the first photon after passing through the first arm through the first coupler input ( 31 ) and the second photon after passing through the second arm through the second coupler input ( 131 ) in the beam coupler ( 30 ), - the two photons the photon pair in the beam coupler ( 30 ), and - each of the two photons of the photon pair has the beam coupler ( 30 ) through both the first and the second coupler output ( 32 . 132 ) can leave, so that either the photons of the photon pair, the beam coupler ( 30 ) together through the same coupler output ( 32 . 132 ) - anticoincidence of the photon pair - or separated by different coupler outputs ( 32 . 132 ) - coincidence of the photon pair -, - a first detector ( 33 ), which is capable of registering the first and / or the second photon, provided that the first and / or second photon is capable of registering the beam coupler ( 30 ) through the first coupler output ( 32 ), - a second detector ( 133 ), which is capable of registering the first and / or the second photon, provided that the first and / or second photon is capable of registering the beam coupler ( 30 ) through the second coupler output ( 132 ), and - a coincidence circuit ( 40 ), which triggers a coincidence signal when the first detector ( 33 ) the first photon of the photon pair and the second detector ( 133 ) detects the second photon of the photon pair or vice versa, ie upon detection of a coincidence of the photon pair, and triggers an anti-coincidence signal when the first or the second detector ( 33 . 133 ) detects both photons of the photon pair, ie upon detection of anticoincidence of the photon pair, the optical path length difference (D) being chosen to perform a significantly higher probability than 50%, a logical XOR operation consisting in using the coincidence circuit ( 40 ) triggers the coincidence signal if one of the two numbers n or m is even and the other is an odd number, and triggers the anti-coincidence signal if both numbers n and m are even numbers or both numbers n and m are odd numbers. Device according to Claim 1, characterized in that the coincidence circuit ( 40 ) as a modification to claim 1 no anti-coincidence signal triggers. Device according to Claim 1, characterized in that the coincidence circuit ( 40 ) in a modification to claim 1 does not trigger a coincidence signal. Device according to claim 1, characterized in that the value of x1 -λ a value from the interval ₁ /8...+λ ₁ / 8th or -λ from the interval ₁ /16...+λ ₁ / 16th, and the value x2, a value from the interval -λ ₂ /8...+λ _2/8 or -λ from the interval ₂ 116 ... + λ _2/16's. Device according to Claim 1, characterized in that the coincidence circuit ( 40 ) comprises a counter and increments it upon detection of a coincidence, and decrements the counter upon detection of anticoincidence, or vice versa, the coincidence circuit ( 40 ) triggers the coincidence signal only when a certain predetermined first count is exceeded, and the anti-coincidence signal only triggers when a certain predetermined second count is exceeded, or vice versa. Device according to Claim 1, characterized in that the coincidence circuit ( 40 ) comprises a counter and increments it upon detection of a coincidence, and decrements the counter upon detection of anticoincidence, or vice versa, the coincidence circuit ( 40 ) triggers the coincidence signal only if within a predetermined first counting period a certain predetermined first count is exceeded, and the anti-coincidence signal triggers only if within a predetermined second counting time a certain predetermined second count is exceeded, or vice versa. Device according to Claim 1 or 2, characterized that the optical path length difference (D) so chosen is that the Probability for Coincidence higher than 75% is or higher than 99% is when n is an even and m is an odd number or vice versa. Apparatus according to claim 1, characterized in that the photon source ( 1 ) a primary laser ( 2 ) and a nonlinear optical crystal ( 3 ), wherein the primary laser ( 2 ) the crystal ( 3 ) with a beam of primary photons ( 2A ) and in the crystal ( 3 ) is capable of generating pairs of photons by means of optical parametric fluorescence, which crystals 3 ) and are each in a two-photon Fock state, and the first and second delays (d1, d2) are each set smaller than the coherence length of the primary photons. Apparatus according to claim 8, characterized in that the primary laser ( 2 ) is a continuous light laser or a pulse laser. Apparatus according to claim 1, characterized in that the photon source ( 1 ) is a two-photon laser, by which the photon pairs are generated. Apparatus according to claim 1, characterized in that the photon source ( 1 ) is one in which quadrupole transitions or cascade transitions occur through which the photon pairs are generated. Apparatus according to claim 1, characterized in that the photon source ( 1 ) is one in which the Coulomb blockade effect occurs by which the photon pairs are generated. Device according to one of Claims 1 to 12, characterized in that the first optical delay device ( 11 ) a first retardation plate ( 12 ) and the second optical delay device ( 111 ) a second retardation plate ( 112 ). Apparatus according to claim 13, characterized in that - the first retardation plate ( 12 ) is a birefringent crystal, whose slow and fast axes are each perpendicular to the propagation direction of the first photon and each form an angle of 45 ° or an angle between 40 ° and 50 ° with the polarization direction of the first photon, and / or - the second Retardation plate ( 112 ) is a birefringent crystal whose slow and fast axes are each perpendicular to the propagation direction of the second photon and each form an angle of 45 ° or an angle between 40 ° and 50 ° with the polarization direction of the second photon. Device according to Claim 14, characterized in that the first and second retardation plates ( 12 . 112 ) in each case a polarizer is connected upstream. Device according to one of claims 1 to 15, characterized in that - the first optical delay device ( 11 ) a first optical control ( 13 ) with a first control input ( 13E ), over which the first delay (d1) is controllable so that the number n - is an even number, if at the first control input ( 13E ) a first voltage value (U1) is present, and - an odd number, if at the first control input (U1) 13E ) a second voltage value (U2) is applied, - and the second optical delay device ( 111 ) a second optical control ( 113 ) with a second control input ( 113E ) over which the second delay (d2) is controllable so that the number m - is an even number, if at the second control input ( 113E ) a third voltage value (U3) is applied, and - an odd number, if at the second control input (U3) 113E ) a fourth voltage value (U4) is present. Apparatus according to claim 16, characterized in that - the first optical control element ( 13 ) is a first Soleil compensator provided with a first piezoelectric crystal, wherein the first delay (d1) is electrically controllable by applying the first or second voltage value (U1, U2) to the first piezoelectric crystal, and / or - the second optical control ( 113 ) is a second Soleil compensator provided with a second piezoelectric crystal, wherein the second delay (d2) is electrically controllable by applying the third or fourth voltage value (U3, U4) to the second piezoelectric crystal. Device according to one of claims 1 to 17, characterized in that the first and / or the second optical delay device ( 11 . 111 ) is a Mach-Zehnder interferometer or a Michelson interferometer or a Fabry-Perot interferometer or an echelon with two interferometer branches, wherein the path difference between the interferometer branches is mechanically and / or electro-optically and / or magneto-optically adjustable. Device according to one of Claims 1 to 18, characterized in that the first and the second optical delay devices ( 11 . 111 ) in each case a polarizer is connected upstream. Device according to one of Claims 1 to 19, characterized in that the first optical delay device ( 11 ) a first polarizer ( 14 ) and / or the second optical delay device ( 111 ) a second polarizer ( 114 ) is connected downstream. Apparatus according to claim 20, characterized ge indicates that the first and second polarizers ( 14 . 114 ) are oriented so that the polarization direction of the first photon upon its arrival in the beam coupler ( 30 ) by less than 5 °, or less than 1 °, from the polarization direction of the second photon upon its arrival in the beam coupler ( 30 ) is different. Device according to Claim 20, characterized in that the first polarizer is a first polarizing beam splitter having a first and a second beam splitter output and the second polarizer is a second polarizing beam splitter having a third and a fourth beam splitter output, the first beam splitter Output to the first coupler input ( 31 ) and the second beam splitter output terminates in a first register, which is capable of registering the first photon, and the third beam splitter output into the second coupler input ( 131 ) and the fourth beam splitter output terminates in a second recorder capable of registering the second photon. Apparatus according to claim 22, characterized in that the first and the second recording device with the coincidence circuit ( 40 ) and temporarily rendering it inoperable when the first recording device registers the first photon or the second recording device registers the second photon. Device according to one of Claims 1 to 23, characterized in that the first and / or the second optical delay device ( 11 . 111 ) comprises a partially transparent plate. Device according to one of Claims 1 to 24, characterized in that - the first optical delay device ( 11 ) comprises a first electro-optical or magneto-optical phase modulator, and / or - the second optical delay device ( 111 ) comprises a second electro-optical or magneto-optical phase modulator. Device according to claim 1, characterized in that that - the first delay (d1) is larger as the coherence length of the first photons, and / or - the second delay (d2) is larger as the coherence length of the second photons. Device according to claim 26, characterized in that that - the first delay (d1) greater than twice or larger than three times the coherence length of the first photons is, and / or The second delay (d2) greater than twice or larger than three times the coherence length of the second photons is. Device according to one of claims 1 to 27, characterized that the the first delay (d1) and the second delay (d2) at least the coherence length of the first photons or at least to distinguish the coherence length of the second photon. Device according to one of Claims 1 to 28, characterized in that the optical path length difference (D) is chosen such that the mean value of the two optical path lengths comprising the first front and the first rear probability wave packet ( 51 . 52 ) in the first arm, from the average of the two optical path lengths which pass through the second front and second rear probability wave packets in the second arm, differs at most a maximum value given by three times the coherence length of the first photon or by the Threefold the coherence length of the second photon. Device according to claim 29, characterized in that that the Maximum value in deviation from claim 29 is given by a Quarter of the coherence length of the first photons or by a quarter of the coherence length of the second photon. Device according to one of claims 1 to 30, characterized that one or both arms of the interferometer at least partially by optical fibers are formed. Device according to one of claims 1 to 31, characterized in that between the photon source ( 1 ) and the interferometer or between the primary laser ( 2 ) and the nonlinear optical crystal ( 3 ) or in at least one of the arms of the interferometer an absorber is interposed, which absorbs a portion of the incident photons on it. Device according to one of claims 1 to 32, characterized in that between the photon source ( 1 ) and the interferometer or between the primary laser ( 2 ) and the nonlinear optical crystal ( 3 ) or in at least one of the arms of the interferometer, an optical component is interposed, which is capable of changing the polarization and / or the phase position and / or the frequency of the optical component passing through photons. Apparatus according to claim 1, characterized in that it, instead of the first and the second detector ( 33 . 133 ) has a single detector which is connected to either the first or the second coupler output ( 32 . 132 ) and able to register both photons of the photon pair and the koinzi denzsignal triggers when it registers only one of the two photons of the photon pair, and the anti-coincidence signal then triggers when it registers both photons of the photon pair. Method for operating a quantum optical interference device with - a photon source ( 1 ), which emits pairs of photons, each of which consists of a first photon of medium wavelength λ ₁ and a second photon of average wavelength λ ₂ , - an interferometer with two arms, between which there is an optical path length difference (D), where - the first photon of a photon pair in the first arm and the second photon of the photon pair is coupled into the second arm, - in the first arm, a first optical delay device ( 11 ), after which the first photon passes either in a first forward or in a first posterior probability wave packet delayed by a first delay d1 (FIG. 51 . 52 ) Is, which is given ₁ /4...+λ _1/4 and n is by d1 = (n · λ _1/2) + x1, wherein x1 is a value from the interval -λ an integer, - in which second arm a second optical delay device ( 111 After passing, the second photon is located either in a second forward or in a second rear likelihood wave packet delayed by a second delay d2 therefrom, which is given by d2 = (m * λ 2/2) + x ₂ , where x2 is a value from the interval -λ _2/4 ... + λ _2/4 and m is an integer, - and with a beam coupler ( 30 ) with a first and a second coupler input ( 31 . 131 ) and a first and a second coupler output ( 32 . 132 ), wherein - the first photon of the photon pair, after this photon has passed through the first interferometer arm, and the second photon of the photon pair, after this photon has passed through the second interferometer, each in a beam coupler ( 30 ) and brought into interference there, - each of the two photons of the photon pair the beam coupler ( 30 ) by both a first and a second coupler output ( 32 . 132 ) can leave, so that either the photons of the photon pair, the beam coupler ( 30 ) together through the same coupler output ( 32 . 132 ) - anticoincidence of the photon pair - or separated by different coupler outputs ( 32 . 132 ) - coincidence of the photon pair -, where - a coincidence signal is triggered when the first photon of the photon pair, the beam coupler ( 30 ) through the first coupler output ( 32 ) and the second photon of the photon pair the beam coupler ( 30 ) through the second coupler output ( 132 ) or vice versa, ie coincidence of the photon pair, - an anti-coincidence signal is triggered when the first and the second photon of the photon pair, the beam coupler ( 30 ) together through the first or the second coupler output ( 32 . 132 ), ie coincidence of the photon pair, - and the optical path length difference (D) is chosen so that with a significantly higher probability than 50%, a logical XOR operation is carried out, which is that the coincidence signal is triggered when a the two numbers n or m are even and the other an odd number, and the anticoincidence signal is triggered if both numbers n and m are even numbers or both numbers n and m are odd numbers. Process according to claim 35, characterized in that that in Variation to claim 35 no anti-coincidence signal is triggered. Process according to claim 35, characterized in that that in Variation to claim 35 no coincidence signal is triggered. A method according to claim 35, characterized in that the value x 1 -λ a value from the interval ₁ /8...+λ ₁ / 8th or -λ from the interval ₁ /16...+λ ₁ / 16th, and the value of x2 is a value from the interval -λ ₂ /8...+λ _2/8 or -λ from the interval ₂ /16...+λ _2/16. Process according to claim 35, characterized in that that - a counter at Coincidence is incremented and decremented in anticoincidence, wherein the coincidence signal is triggered only when a certain predetermined first count exceeded is, and the anti-coincidence signal is only triggered if a certain predetermined second meter reading is undershot, - or the other way around. Process according to claim 35, characterized in that that - a counter at Coincidence is incremented and decremented in anticoincidence, wherein the coincidence signal is triggered only if within a predetermined first count period a certain predetermined first count is exceeded, and the Anti-coincidence signal is only triggered when within a predetermined second count period falls below a certain predetermined second count, - or the other way around. A method according to claim 35, characterized ge indicates that the optical path length difference (D) is chosen such that the probability of coincidence is significantly higher than 75%, or higher than 99%, when n is even and m is an odd number or vice versa. Process according to claim 35, characterized in that as photon source ( 1 ) a primary laser ( 2 ) and a nonlinear optical crystal ( 3 ), the primary laser ( 2 ) the crystal ( 3 ) with a beam of primary photons ( 2A ) and irradiated in the crystal ( 3 ) generates pairs of photons by means of optical parametric fluorescence, which are emitted from the crystal ( 3 ) and each are in a two-photon Fock state, and the first and second delays (d1, d2) are each set smaller than the coherence length of the primary photons. Process according to claim 42, characterized in that as a primary laser ( 2 ) a continuous light laser or a pulse laser is used. Method according to one of Claims 35 to 43, characterized in that - the first optical delay device ( 11 ) a first optical control ( 13 ) with a first control input ( 13E ), over which the first delay (d1) is controlled so that the number n - is an even number, if at the first control input ( 13E ) a first voltage value (U1) is present, and - an odd number, if at the first control input (U1) 13E ) a second voltage value (U2) is applied, - and the second optical delay device ( 111 ) a second optical control ( 113 ) with a second control input ( 113E ), over which the second delay (d2) is controlled so that the number m - is an even number, if at the second control input ( 113E ) a third voltage value (U3) is applied, and - an odd number, if at the second control input (U3) 113E ) a fourth voltage value (U4) is present. Method according to claim 44, characterized in that - as the first optical control element ( 13 ) using a first piezoelectric crystal provided first Soleil compensator, wherein the first delay (d1) by applying the first or second voltage value (U1, U2) is electrically controlled to the first piezoelectric crystal, and / or - than second optical control ( 113 ) using a second piezoelectric crystal second Soleil compensator is used, wherein the second delay (d2) is electrically controlled by applying the third or fourth voltage value (U3, U4) to the second piezoelectric crystal. Method according to one of claims 35 to 45, characterized in that - the first photon after passing through the first optical retarder by a first polarizer ( 14 ) and / or - the second photon after passing the second optical retarder through a second polarizer ( 114 ) is polarized. A method according to claim 46, characterized in that the first photon after passing through the first optical retarder and / or the second photon after passing through the second optical retarder are each polarized so that the polarization direction of the first photon upon arrival in the beam coupler ( 30 ) by less than 5 ° or less than 1 ° from the polarization direction of the second photon upon its arrival in the beam coupler ( 30 ) is different. The method of claim 46 or 47, characterized in that the first polarizer is a first polarizing beam splitter having a first and a second beam splitter output and the second polarizer is a second polarizing beam splitter having a third and a fourth beam splitter output, the first Beam splitter output in the first coupler input ( 31 ) and the second beam splitter output terminates in a first register, which is capable of registering the first photon, and the third beam splitter output into the second coupler input ( 131 ) and the fourth beam splitter output terminates in a second recorder capable of registering the second photon. Method according to Claim 48, characterized in that the first and the second recording device are equipped with the coincidence circuit ( 40 ) and temporarily put it in a non-functional state when the first recording device registers the first photon or the second recording device registers the second photon. Method according to one of claims 35 to 49, characterized that - the first delay (d1) by a first electro-optical or magneto-optical phase modulator affected will, and / or - the second delay (d2) by a second electro-optical or magneto-optical Phase modulator affected becomes. A method according to claim 35 or 50, characterized that - the first delay (d1) greater than the coherence length of the first photons chosen will, and / or - the second delay (d2) greater than the coherence length of the second photons chosen become. Process according to claim 51, characterized in that that - the first delay (d1) greater than twice or larger than three times the coherence length of the first photons chosen will, and / or - the second delay (d2) greater than twice or larger than three times the coherence length of the second photons chosen becomes. A method according to claim 51 or 52, characterized that the first delay (d1) and the second delay (d2) be chosen so that she at least the coherence length of the first photons or at least to distinguish the coherence length of the second photon. Method according to one of claims 35 to 53, characterized that the first delay (d1) and the second delay (d2) always chosen like that be that Case n = m does not occur. Method according to one of Claims 35 to 54, characterized in that the optical path length difference (D) is chosen such that the mean value of the two optical path lengths comprising the first front and the first rear probability wave packet ( 51 . 52 ) in the first arm, from the average of the two optical path lengths which pass through the second front and second rear probability wave packets in the second arm, differs at most a maximum value given by three times the coherence length of the first photon or by the Threefold the coherence length of the second photon. Device according to Claim 55, characterized that the Maximum value in deviation from claim 55 is given by a Quarter of the coherence length of the first photons or by a quarter of the coherence length of the second photon. Method according to one of claims 35 to 56, characterized that - the intensity of the current from first photons through the first arm through one in the first intermediate or one upstream of the interferometer Absorber is reduced, and / or - the intensity of the current from second photons through the second arm through one in the second arm intermediate or one upstream of the interferometer Absorber is reduced. Method according to one of claims 35 to 57, characterized that - the polarization and / or the phase position and / or the frequency of the first photon is changed by an optical component interposed in the first arm, and or - the Polarization and / or the phase position and / or the frequency of the second Photons by an interposed in the second arm optical Component changed becomes. Method according to one of Claims 35 to 58, characterized that this Method is performed several times in succession, the first delay (d1) is controlled in time so that the Number n is a first predetermined sequence of even and odd numbers goes through and synchronously, the second delay (d2) is controlled in time so that the Number m is a second predetermined sequence of even and odd numbers goes through so that the Change of the resulting triggers of the Coincidence signal and the anti-coincidence signal of the logical XOR operation of first and second sequences, if even numbers n, m each as logical one and odd numbers n, m each as logical zero or vice versa. Method according to claim 59, characterized that the first delay (d1) and the second delay (d2) be timed so that on each in the interferometer incident or passing through each of the interferometer photon pair exactly one element of the first and one element of the second sequence is omitted. Method according to claim 59, characterized that the first delay (d1) and the second delay (d2) be timed so that on each in the interferometer incident or passing through each of the interferometer photon pair at least one element of the first and one element of the second sequence eliminated. Method according to claim 59, characterized that the first delay (d1) and the second delay (d2) be timed so that on each in the interferometer incident or passing through each of the interferometer photon pair on average at least one element of the first and one element of the second consequence deleted. Method according to one of Claims 35 to 62, characterized in that either the first or the second coupler output ( 32 . 132 a single detector is connected downstream, which is capable of registering both photons of the photon pair, a coincidence being recognized by the fact that the single detector registers only one of the photons of the photon pair, and an anticoincidence is detected by registering both photons of the photon pair. A method of identifying a party pair (A, B) consisting of a first party (A) and a second party (B) to a controller (C) using the method according to one of claims 35 to 63 or using the device according to one of claims 1 to 34, comprising the following steps: a) the first party (A) receives a first secret bit key (aS) comprising a sequence of first key bits, b) the second party (B) receives a second secret bit key (bS) comprising a sequence of second key bits; c) the controller (C) receives a third secret bit key (cS) which is given by the bitwise logical XOR operation between the d) the first bit (a) converts the first bit key (aS) into a time sequence of values for the integer n by the first and second bit keys (aS, bS) such that cS = aS XOR bS first party (A) times the first delay (d1) so that set bits of the first bit key (aS) are converted into even numbers for the number n and unset bits of the first bit key (aS) into odd numbers for the number n or, conversely, e) the second party (B) synchronously translates the second bit key (bS) into a time series of values for the integer m by the second party (B) timing the second delay (d2) to control that all set bits of the second bit key (bS) are converted into even values for the number m and all unset bits of the second bit key (bS) into odd values for the number m, or vice versa, so that each bit thus converted is the first bit key (aS) forms a bit pair with the bit of the second bit key (bS) converted in synchronism therewith and exactly one value for the number n belongs to each bit pair and exactly one value belongs to the number m; f) the controller (C) detects synchronously to do this, trigger the coincidence signal and the anticoincidence signal and convert them into a result bit sequence (cE) by giving each trigger of the coincidence signal a set bit and triggering the ant g) the result bit string (cE) is compared with the third bit key (cS); h) the party pair (A, B) formed by the first and second parties (A, B) is said to be identified, provided that there is at least a certain predetermined degree of agreement between the result bit sequence (cE) and the third bit key (cS). Method according to claim 64, characterized in that that the first party (A) the first delay (d1) and the second party (B) time-control the second delay (d2), that on each incident into the interferometer or to each the interferometer passing photon pair exactly one bit pair or at least one Bit pair is omitted. Method according to claim 64 or 65, characterized that the first party (A) the first delay (d1) and the second party (B) time-control the second delay (d2), that on each entering the interferometer or on each the interferometer passing photon pair on average at least one bit pair is omitted. A method of identifying a party pair (A, B) consisting of a first party (A) and a second party (B) to a controller (C) using the method of any of claims 35 to 63 or using The apparatus according to one of claims 1 to 34, comprising the following steps: a) the first party (A) receives a first secret bit key (aS) comprising a sequence of first key bits, and modifies it to a first check bit string (aP) in that the first party (A) assigns a first plurality (X) of bits of the first bit key (aS) which are located at arbitrary or randomly selected first check bit positions (PPa1, PPa2, ..., PPaX) within the first bit key ( aS) are replaced by randomly or randomly selected first check bits (aP1 ... aPX), b) the second party (B) receives a second secret bit key (bS) comprising a sequence of second key bits and modifies this to a second one a check bit sequence (bP), in which the second party (B) assigns a second plurality (Y) of bits of the second bit key (bS) which are connected to arbitrarily or randomly selected second check bit positions (PPb1, PPb2,. PPbY) are within the second bit key (bS), replaced by arbitrarily or randomly selected second check bits (bP1 ... bPY), c) the controller (C) receives a third secret bit key (cS) which is given by the bitwise logical XORing between the first and second bit-keys (aS, bS) such that cS = aS XOR bS, d) the first party (A) sets the first check bit string (aP) into a time sequence of values for the integer n by the first party (A) timing the first delay (d1) so that set bits of the first check bit sequence (aP) are converted into even numbers for the number n and unset bits of the first check bit sequence ( aP) are converted into odd numbers for the number n, or vice versa e) the second party (B) synchronously converts the second check bit string (bP) into a time series of values for the integer m by the second party (B) timing the second delay (d2), that set bits of the second check bit sequence (bP) are converted into even values for the number m and unset bits of the second check bit sequence (bP) into odd values for the number m, or vice versa, so that each bit thus converted is the first Test bit sequence (aP) with the synchronously converted bit of the second test bit sequence (bP) forms a pair of bits and for each pair of bits exactly one value for the number n and exactly one value for the Number m, the first party (A) controlling the first delay (d1) and the second party (B) timing the second delay (d2) so that each photon pair entering the interferometer or passing through each one of the interferometer accurately F) the controller (C) synchronously detects the triggering of the coincidence signal and the anti-coincidence signal and converts them into a third test bit sequence (cP) by triggering the coincidence signal g) the third check bit string (cP) is compared with the third bit key (cS), h) the one formed by the first and second parties (A, B) Party pair (A, B) is considered identified if at least a certain predetermined degree of agreement between the third check bit string (cP) and the third I) at least a portion of the first check bit positions (PPa1, PPa2, ..., PPaX) is compared with at least a portion of the second check bit positions (PPb1, PPb2, ..., PPbY), resulting in a set of Z match positions (ÜP1, ÜP2, ..., ÜPZ) at which one of the first check bit positions (PPa1, PPa2, ..., PPaX) with one of the second check bit positions ( PPb1, PPb2, ..., PPbY), so that for each match position (ÜP1, ÜP2, ..., ÜPZ) exactly one bit (aPÜ1, aPÜ2, ..., aPÜZ) of the first check bit sequence ( aP), a bit of the second check bit string (bP) and a bit of the third check bit string (cP) belongs, j) for each match position (ÜP1, ÜP2, ..., ÜPZ) or a part thereof is passed through logical XOR operation between the associated bit of the first test bit sequence (aP) and the associated bit of the second test bit sequence (bP) determines a match bit (cÜ1, cÜ2, ..., cÜZ), k) at least e in part of the match bits (cÜ1, cÜ2, ..., cÜZ) is compared with respective bits of the third check bit string (cP), and l) the first and second bit keys (aS, bS) are categorized as eavesdropping if the method step k) shows that at least one predetermined critical number (Z) or at least one predetermined critical portion (AZ) of match bits (cÜ1, cÜ2, ..., cÜZ) from the respectively associated bit of the third check bit Bit sequence (cP) deviates. Process according to claim 67, characterized in that that each between 10% and 50% of the first check bit sequence (aP) and the second Testing bit sequence (bP) check bits are. Method according to claim 64 or 67, characterized that to Synchronization of process steps d), e) and f) an electrical or electronic synchronization device is used, with which the first and second party (A, B) and the inspector (C) are connected. Method according to one of Claims 67 to 69, characterized that the Number of first key bits, the first check bits (aP1 ... aPX), the second key bits and the second check bits (bP1 ... bPY) each is chosen so big that the probability for this, that less as a number of L match positions (ÜP1, ÜP2, ..., ÜPZ), is less than a given limit probability, where L a freely selectable is positive integer. Method according to claim 70, characterized in that that the Number L greater than 10, or greater than 100, chosen becomes. A method according to claim 70 or 71, characterized in that the limit probability is set to be less than 10 ^-3 , or less than 10 ^-6 . Process according to claim 67, characterized in that that - that through the first and the second party (A, B) formed party pair (A, B) is not classified as identified, The bit-keys (aS, bS, cS) with new bitcodes be replaced, and - the Process is repeated with the same, if in the Verfahrensschsschritt i) a number of match position which is smaller than a predetermined minimum value. Method according to claim 64 or 67, characterized that in Modification to Claim 64 or 67 The Third Secret Bit-Key (cS) is given by the bitwise logical NXOR link, i. through the inverse XOR operation between the first and second bit-keys (aS, bS) such that cS = aS NXOR bS applies, and the controller (C) synchronously to the triggers of the Coincidence signal and the anti-coincidence signal detected and this into a result bit sequence (cE) by triggering each triggering of the Coincidence signal is a bit not set and any triggering of the anti-coincidence signal assigns a set bit.