GB2582556A - Improvements to QKD network architectures - Google Patents

Abstract

A system and method for quantum key distribution (QKD) by the transmission of a plurality of optical pulses, such as single-photon pulses, from a transmitter 2 to a plurality of receivers 3-6 via a guide 47. The guide is configured to allocate a proportion of the optical pulses to each of its ports and a modification means is provided to be able to change the proportions at each port. The modifications may be applied continuously by a controller 60. The guide may be a single optical variable splitter or a plurality of splitters connected in series. The receivers may send an indication of pulse receipt rate to the controller such that the pulse proportions at the outputs are reconfigured in response. The advantages of the invention are i) a sufficiently high pulse rate may be supplied to a receiver that needs to generate a large number of keys, ii) a larger proportion of pulses may be sent to a remote receiver that suffers high levels of attenuation due to it being at a greater distance from the transmitter.

Description

Improvements to QKD network architectures QKD (Quantum Key Distribution) is a known method of highly-secure communication which enables two parties to produce a shared secret key. In the broadest sense, QKD networks involve a transmitting unit (referred to as Alice) which generates a single photon pulse, encodes it with a randomly-chosen value of 0 or 1 and applies a modulation to it. In a known example system this is done by polarising the pulse using a randomly-chosen polarisation basis. The pulse is then transmitted to a receiving unit (referred to as Bob) which attempts to measure the encoded value. In the example system this is done by passing the pulse through a polariser with a randomly-chosen polarisation basis. The process is repeated for multiple pulses. Alice and Bob then exchange information regarding the pulses and, in doing so, arrive at a shared secret key.

QKD arrangements are expensive and one way of reducing costs is to have a single Alice which transmits pulses to multiple Bobs. Such arrangements have a branched optical path so that each Bob receives a proportion of the pulses. Optical splitters have been used to provide the branched optical path. A problem with splitting the signal is that some of the Bobs may not receive a sufficiently high pulse rate to meet their key generation demands. This is particularly true if, say, one Bob needs to produce a larger number of keys than the others. It may also be true if one Bob is located further from the splitter than the others (as the pulses directed to it will suffer more attenuation), or if there are a large number of Bobs. In such situations the pulse receipt rate at the Bobs may not be high enough to keep the session with Alice active. If this happens, lengthy re-establishment procedures are required in order to re-establish the session.

It would be desirable to overcome or mitigate some and/or all of the above-mentioned and/or other disadvantages of the prior art.

According to a first aspect of the invention there is provided a system of transmitting optical pulses from a transmitter to first and second receivers for the purposes of quantum key distribution, the system comprising: a transmitter configured to transmit a plurality of optical pulses; first and second receivers configured to receive some or all of the plurality of optical pulses; a guide for guiding the plurality of optical pulses, the guide having first and second output ports, the first output port being in optical communication with the first receiver and the second output port being in optical communication with the second receiver, the guide being configured to output a proportion of the optical pulses at the first output port and a proportion of the optical pulses at the second output port, modification means for modifying the guide so as to change the proportion of the optical pulses output at the first output port.

According to a second aspect of the invention there is provided a method of transmitting optical pulses between a transmitter and first and second receivers for the purposes of quantum key distribution, the method comprising guiding a plurality of optical pulses to either a first output port or a second output port, the first output port being in optical communication with the first receiver and the second output port being in optical communication with the second receiver; the guide being configured to output a proportion of the optical pulses at the first output port and a proportion of the optical pulses at the second output port, modifying the guide so as to change the proportion of the optical pulses output at the first output port.

The inventive method enables the construction of a system which is able to meet the needs of particular receivers in real time. Such needs may arise because a particular receiver may need to agree a larger number of keys than other receivers. Alternatively, one of the receivers may be located a greater distance from the transmitter then the other and so pulses transmitted to it suffer greater attenuation. The invention enables a larger proportion of the pulses to be provided to the output port that leads to that receiver, so that the receiver receives the pulse rate it requires.

Furthermore, in prior art arrangements having splitters that are connected in series, receivers may not receive a sufficiently high pulse rate to maintain a connection, requiring lengthy re-connection procedures. The present invention enables the pulse rate to be equalised between proximate and remote receivers.

The method may further comprise transmitting optical pulses from a transmitter to one or more further receivers. The pulses may be randomly encoded with one or zero. A modulation may be applied to the pulses. The step of applying a modulation to the pulses may comprise polarising the pulses using a randomly chosen polarisation basis. Alternatively, the step of applying a modulation to the pulses comprise modulating the phase or the position of the pulse. After being encoded and modulated the pulses may be transmitted in accordance with QKD principles.

The value encoded onto the pulse and/or the modulation basis applied to the pulse and/or the time of transmission of the pulse may be recorded. The modulation basis may be the polarisation basis. Furthermore, the transmitter may be connected for communication with each of the receivers by a respective non-quantum channel. The non-quantum channel may be a metallic wire, or an optical fibre or free space. The transmitter may send some or all of the recorded information to one or more of the receivers using the respective non-quantum channel.

The value that has been encoded onto the pulses (i.e. a one or a zero) may be measured at each of the receivers using randomly chosen modulation basis which may be a randomly chosen polarisation basis. For each received pulse, the measured value and/or the polarisation basis used for the measurement and/or the arrival time of the pulse may be recorded. Each of the receivers may send the measured value and/or the arrival time of the pulse to the transmitter using its respective non-quantum channel.

The optical pulses may be single-photon pulses. Single-photon pulses may be generated by a single-photon generator. The guide may comprise an optical splitter. The optical splitter may be separate to the transmitter and may be spaced apart from the transmitter.

The splitter may be connected to the transmitter by an optical channel which may be an optical fibre. The optical fibre may also comprise the non-quantum channel. The splitter may comprise the first and second output ports and may comprise further output ports. Each output port may be capable of outputting single-photon pulses received from the transmitter. The first output port may be connected to the first receiver by an optical channel which may be an optical fibre. The second output port may be connected to the second receiver by an optical channel which may be an optical fibre. Further output ports may be connected to respective further receivers by respective further optical fibres.

Although in preferred embodiments the guide comprises a splitter, the skilled person would understand that it would be possible to provide a guide which did not comprise a splitter but which provides substantially the same functionality. A possible example of this would be an arrangement which provided a plurality of pulses to a first output for a short time period, then provided a plurality of pulses to a second output for a short time period, and repeated this cycle over a long period. Over time such an arrangement would direct a proportion of the input pulse stream to each of the outputs.

In some embodiments the guide comprises first and second splitters. In such embodiments the first splitter may comprise the first output port which may be connected to the first receiver. The second splitter may comprise the second output port which may be connected to the second receiver. The second splitter may be connected to an output port of the first splitter such that the second splitter receives pulses output by the first splitter. The guide may comprise one or more further splitters optically connected in series. The one or more further splitters may each be connected to a respective receiver. Each of the one or more further splitters may output a portion of the pulses input to it to its connected receiver, and the remainder to an input of the next splitter in the series.

The modification means may be automated and may comprise a control logic. The control logic may determine a modification to make to the guide. The control logic may apply the determined modification to the guide. The determination step may comprise determining a modification which equalises the pulse rate at each of the receivers.

Alternatively the determination step may comprise determining a modification which provides a higher pulse rate at one or more of the receivers than at one or more other of the receivers. The determination step may comprise determining a modification which ensures that each of the receivers receives sufficiently high pulse rate such that the session is maintained, i.e. re-establishment procedures are not required. The determination step may take the pulse rate demands of one or more of the receivers into account. One or more of the receivers may send the control logic an indication of the pulse rate demands at that receiver.

In embodiments in which the guide comprises one splitter, the step of modifying the guide may comprise modifying the split ratio of the splitter. In embodiments in which there are more than one splitter, the step of modifying the guide may comprise modifying the split ratio of some or all of the splitters. This may comprise applying an appropriate electrical impulse to the splitter or splitter. The split ratio of the splitter or splitters may be modifiable in a continuous fashion or may be modifiable in discrete steps.

A preferred embodiment of the invention will now be described in detail, for illustration purposes only, with reference to the appended drawings, in which Figure 1 is a schematic view of a basic QKD architecture comprising a single Alice and a single Bob; Figure 2 is a schematic view of a known QKD architecture comprising a single Alice and multiple Bobs; Figure 3 is a schematic view of a different known QKD architecture comprising a single Alice, multiple Bobs and multiple splitters; Figure 4 is a schematic view of a first embodiment of the invention; Figure 5 is a schematic view of a second embodiment of the invention.

The present invention concerns improvements in Quantum Key Distribution (QKD). QKD is a method of encryption involving distributing an encryption key from a first quantum node (known as Alice) to a second quantum node (known as Bob). Fig. 1 shows a schematic view of a simplified QKD arrangement in which Alice is shown at 100 and Bob at 200. Although multiple QKD protocols are in use, the present explanation relates to the commonly-used BB84 protocol. In particular, this explanation concerns a type of BB84 in which the modulation is applied to the pulses using polarisation. According to that protocol, Alice 100 randomly generates a bit (either 0 or 1) and also randomly chooses one of two polarisation bases: rectilinear and diagonal. Alice 100 then sends a photon that has been encoded with the chosen bit and the chosen polarisation basis to Bob 200 via a quantum channel 300, such as an optical fibre. Bob 200 randomly selects one of the two polarisation bases and measures the photon using its chosen basis. If Bob 200 uses the same basis as Alice 100 then the bit value measured by Bob 200 will match that applied to the photon by Alice 100. After repeating the process with a large number of photons, Alice 100 and Bob 200 perform a key agreement stage. In particular, Alice 100 informs Bob 200, via a non-quantum communication channel 400 (such as a copper cable), which of the two bases Alice 100 applied to each photon, along with the time of transmission by Alice 100 of each photon. Bob 200 then informs Alice 100 which of the two bases Bob 200 used when measuring each photon along with the time at which Bob 200 received each photon. Alice 100 and Bob 200 then discard their bit values for which Alice 100 and Bob 200 used different bases, and keep the remaining bit values. The remaining bit values constitute a secret key that Alice 100 and Bob 200 both have and which they can use to encrypt messages sent between them over the non-quantum channel 400.

Fig. 2 shows a known QKD architecture. In particular, Alice 2 is connected to four different Bobs 3,4,5,6. An optical splitter 7 is provided in the optical path from Alice 2 to the four Bobs 3,4,5,6. Alice 2 is connected to the splitter 7 by an optical fibre 10. Each of the four Bobs 3,4,5,6 is connected to the splitter 7 by a respective optical fibre 50. Each of the Bobs 3,4,5,6 is also connected to Alice 2 by its own classical (i.e. non-quantum) channel (not shown).

In use, Alice 2 sends a series of photons to the splitter 7, each photon having been encoded with a random bit and a random polarisation base as described above. Each photon passes through the splitter 7 and on to one of the Bobs 3,4,5,6. The splitter 7 diverts each photon to one of the Bobs 3,4,5,6 at random. Therefore, on average, 25% are directed towards each of the four Bobs. Each Bob measures the photons using a randomly-chosen polarisation basis, and, once it has received enough photons to enable a secret shared key to be established with Alice 2, it does so using the key agreement stage described above. The key agreement stage involves Alice 2 and each one of the Bobs 3,4,5,6 exchanging their respective lists of polarisation bases and takes place over the classical channel. Alice 2 and each one of the Bobs use the lists to establish a shared secret key which can be used to encrypt communications between them.

As can be seen in Fig. 2, Bob 6 is located further from the splitter 7 than the other Bobs 3,4,5. As the photons sent to Bob 6 travel further than the photons sent to the other Bobs 3,4,5, the photons transmitted to Bob 6 suffer a greater attenuation, meaning that more photons are lost. The key exchange rate between Alice and Bob 6 may therefore be lower than the key exchange rate between Alice and the other Bobs 3,4,5. This is undesirable as it slows down the establishment of a secure communication link between Bob 6 and Alice 2. Furthermore, if the photon receipt rate at bob 6 falls below a threshold rate, it is necessary to re-initiate the connection, which is time consuming.

A further disadvantage of the arrangement of Fig. 2 is where one of the Bobs, say Bob 4, needs to establish more keys with Alice over a given time period than do the other Bobs. As the splitter splits the photons approximately equally, it may take a long time to establish all the keys at Bob 4, while photons are sent unnecessarily to other Bobs after they have finished establishing their keys.

Fig. 3 shows an alternative known architecture for QKD. This may be advantageous over the architecture of Fig. 2 because Fig. 3 all the Bobs are able to be served using only a single main optical fibre (plus much shorter branches). As in Fig. 2, the four Bobs 23,24,25,26 are each also connected to Alice 22 by a respective classical channel that is not shown in Fig. 3. Elements 27, 28 and 29 are 50:50 splitters. In use, Alice sends a stream of photons, encoded and polarised in the manner described above, to splitter 27. Splitter 27 directs approximately 50% of the photons towards Bob 23. Bob 23 detects the photons that reach it and performs the key agreement and exchange steps described above using its classical channel to Alice 22. The remaining approximately 50% of the photons pass through splitter 27 towards splitter 28. Splitter 28 directs approximately 50% of the photons arriving at it towards Bob 24. Bob 24 detects the photons that reach it and performs the key agreement and exchange steps described above using its classical channel to Alice 22. Splitter 28 also directs approximately 50% of the photons arriving at it towards splitter 29, which directs these photons approximately equally between Bob 25 and Bob 26, which each perform the key agreement and exchange steps described above using its respective classical channel to Alice 22.

Due to the architecture of this arrangement, Bob 23 receives approximately double the number of photons as Bob 24, which in turn receives approximately double the number of photons as Bobs 25 and 26. This can make key generation difficult for Bobs 25 and 26. The number of photons reaching Bobs 25 and 26 is further significantly reduced if Bobs 25 and 26 are located remotely to the reminder of the arrangement. This is due to fibre loss. Furthermore, as mentioned above, if the photon receipt rate at a particular Bob falls below a threshold rate, it is necessary to re-initiate the connection to Alice 22, which is time consuming. Furthermore, a user may wish to add a further Bob to the arrangement of Fig. 3. This may be done, e.g. by replacing Bob 26 with a splitter and connecting that splitter to Bob 26 and the new Bob. The additional splitter with further reduce the photon receipt rate at Bob 26 and at the new Bob. As noted above, this is likely to slow the process of key forming and may result in disconnection and consequent lengthy re-initiation procedures.

The known arrangements shown at Figs 2 and 3 have the disadvantages described above. Fig 4 shows an arrangement in accordance with an aspect of the invention. Fig 4 has many features in common with Fig 2. In particular, Alice 2 is connected to four different Bobs 3,4,5,6. An optical splitter 47 is provided in the optical path from Alice 2 to the four Bobs 3,4,5,6. Alice 2 is connected to the splitter 47 by an optical fibre 10. Each of the four Bobs 3,4,5,6 is connected to the splitter 47 by a respective optical fibre 50. Each of the Bobs 3,4,5,6 is also connected to Alice 2 by its own classical (i.e. non-quantum) channel (not shown).

The principal difference between Fig 4 and Fig 2 is that in Fig 4 the splitter 47 is a variable splitter. Variable splitters are known and so their structure will not be described in detail here. A variable splitter is an optical splitter, the split ratio of which can be changed in real time. That is to say, the proportion of photons that the splitter provides at each output can be varied. In the example of Fig 2, the splitter 7 split the incoming signal equally between its outputs, such that approximately 25% of the photons output by the splitter 7 were directed towards each Bob. As noted above, photons travelling to Bob 6 are more highly attenuated than those travelling to Bobs 3,4 and 5 as Bob 6 is more remote. In the arrangement of Fig 4, the split ratio of splitter 47 can be varied such that a greater proportion of photons are directed towards Bob 6 than are directed towards Bobs 3,4 and 5 in order to compensate for the increased attenuation. So, for example, variable splitter 47 may be adjusted such that 20% of the photons output from splitter 47 are directed towards each of Bobs 3,4 and 5 and the remaining 40% of the photons output from splitter 47 are directed towards Bob 6.

Furthermore, it may be that another of the Bobs, say, Bob 4, wishes to agree a larger number of keys with Alice 22 than do the other Bobs 3,5 and 6. In other words, Bob 4 has a higher key exchange rate demand than do Bobs 3,5 or 6. In such a situation, Bob 4 sends controller 60 an indication of the key exchange rate that it requires with Alice 22, along with the photon receipt rate that it is currently experiencing. As in the example above, the split ratio of splitter 47 can be varied such that a greater proportion of photons are directed towards the Bob that needs it, in this case Bob 4, than are directed towards the other Bobs 3,4 and 5. So, for example, variable splitter 47 may be adjusted such that 20% of the photons output from splitter 47 are directed towards each of Bobs 3,5 and 6 and the remaining 40% of the photons output from splitter 47 are directed towards Bob 4.

The adjustment of variable splitter 47 described above is carried out by a controller 60 in the following way. The controller 60 is provided with an electrical connection (shown in Fig. 4 by a dashed line) to each of Bobs 3,4,5 and 6. In use, each of the Bobs 3,4,5 and 6 send the controller 60 an indication of the photon receipt rate that it is currently experiencing along with an indication of the photon receipt rate that it requires (due to its current key exchange rate demands). If the ratio of the actual photon receipt rate to required photon receipt rate for a particular Bob, e.g. Bob 6, falls significantly below that of the other Bobs 3,4,5, the controller 60 sends an instruction to splitter 47 to increase the ratio of photons being output towards Bob 6 and decrease the ratio of photons being output towards Bobs 3,4 and 5.

Fig. 5 shows an embodiment of the invention which has many features in common with Fig. 3. In particular, Alice 32 is connected to a single optical channel. The optical channel is provided with branched connections to four Bobs 33,34,35,36. The four Bobs 33,34,35,36 are also each connected to Alice 32 by a respective classical channel that is not shown in Fig. 5. The principal difference between Fig. 5 and Fig. 3 is that in Fig. the splitters 37, 38 and 39 are variable splitters.

As in Fig. 3, Alice sends a stream of photons, encoded and polarised in the manner described above, to the four Bobs 33,34,35,36 via respective splitters. As the split ratio in Fig. 5 can be adjusted, the split ratio of the splitters need not be 50%. The split ratio of each splitter can be set in order to equalise the photon receipt rate at the Bobs 33,34,35,36. This enables key formation at each of the Bobs to take place at approximately the same rate -with no individual Bobs falling behind in their key formation rate.

The adjustment of variable splitters described above is carried out by a controller 70 in the following way. The controller 70 is provided with an electrical connection (shown in Fig. 5 by dashed lines) to each of splitters 37, 38, 39. In a manner analogous to that described above in relation to Fig. 4, each of the Bobs 33,34,35,36 sends the controller 70 an indication of the photon receipt rate that it is currently experiencing along with an indication of the photon receipt rate that it requires (due to its current key exchange rate demands). If the ratio of the actual photon receipt rate to required photon receipt rate for a particular Bob, e.g. Bob 36, falls significantly below that of the other Bobs, the controller 70 determines appropriate split ratios to apply to each of the splitters 37, 38, 39 in order to increase the photon receipt rate at Bob 36 while providing an appropriate photon receipt rate at the other Bobs, in view of their own indicated demands.

This technique could also be of benefit if, for example, the key formation rate requirement at a particular Bob increases to a higher rate. This extra demand can be met in real time by the controller 70 determining and applying appropriate split ratios to apply to each of the splitter 37, 38, 39 to meet the increased photon rate demand of the particular Bob, while also ensuring the photon rate demands of the other Bobs are also met.

This technique could also be of benefit if a user adds an extra Bob to the arrangement.

This may be done, e.g. by replacing Bob 36 with a splitter and connecting that splitter to both Bob 36 and the new Bob. The additional splitter will further reduce the photon receipt rate at Bob 26 and at the new Bob, which may slow the process of key forming and may result in disconnection and consequent lengthy re-initiation procedures. However, in use, each of new Bob and Bob 36 send the controller 70 an indication of the photon receipt rate that it is currently experiencing along with an indication of the photon receipt rate that it requires (due to its current key exchange rate demands). If the ratio of the actual photon receipt rate to required photon receipt rate for either the new Bob or Bob 36 falls significantly below that of the other Bobs 33, 34, 35, the controller 70 determines appropriate split ratios to apply to each of the splitters 37, 38, 39 in order to increase the photon receipt rate at the new Bob and Bob 36 while providing an appropriate photon receipt rate at the other Bobs 33, 34, 35 in view of their own indicated demands. This reduces the likelihood of slow key forming or disconnection at the new Bob or Bob 36.

Claims (15)

1. Claims 1. A system for transmitting optical pulses from a transmitter to first and second receivers for the purposes of quantum key distribution, the system comprising: a transmitter configured to transmit a plurality of optical pulses; first and second receivers configured to receive some or all of the plurality of optical pulses; a guide for guiding the plurality of optical pulses, the guide having first and second output ports, the first output port being in optical communication with the first receiver and the second output port being in optical communication with the second receiver, the guide being configured to output a proportion of the optical pulses at the first output port and a proportion of the optical pulses at the second output port; modification means for modifying the guide so as to change the proportion of the optical pulses output at the first output port.
2. 2. A system as claimed in claim 1, further comprising one or more further receivers configured to receive some or all of the plurality of optical pulses.
3. 3. A system as claimed in claim 1 or claim 2, wherein the guide comprises an optical splitter.
4. 4. A system as claimed in any preceding claim, wherein the splitter is connected to the transmitter by an optical fibre.
5. 5. A system as claimed in any preceding claim, wherein the splitter comprises the first and second output ports.
6. 6. A system as claimed in any preceding claim, wherein the guide comprises first and second splitters.
7. 7. A system as claimed in claim 6, wherein the second splitter comprises the second output port.
8. 8. A system as claimed in claim 6 or claim 7, wherein the guide comprises one or more further splitters.
9. 9. A system as claimed in claim 8, wherein the one or more further splitters are optically connected in series.
10. 10. A system as claimed in any preceding claim, wherein the modification means comprises a control logic.
11. 11. A system as claimed in claim 10, wherein the receivers are connected for communication to the control logic. 10
12. 12. A method of transmitting single-photon pulses between a transmitter and first and second receivers for the purposes of quantum key distribution, the method comprising: using a guide to guide a plurality of optical pulses to either a first output port or a second output port, the first output port being in optical communication with the first receiver and the second output port being in optical communication with the second receiver, the guide being configured to output a proportion of the optical pulses at the first output port and a proportion of the optical pulses at the second output port; modifying the guide so as to change the proportion of the optical pulses output at the first output port.
13. 13. A method as claimed in claim 12, wherein the step of modifying the guide comprises modifying the proportion of optical pulses guided to the first output port relative to the second output port.
14. 14. A method as claimed in claim 13, wherein the proportion of optical pulses guided to the first output port relative to the second output port is modified in a continuous fashion
15. 15. A method as claimed in any of claims 12 to 14, wherein the method further comprises sending an indication of a pulse receipt rate required by the first and second or receivers.