A Photonic Integrated Circuit

Abstract

A photonic integrated circuit 4 comprises a demultiplexer 14, first 181 second 182 and third 22 phase controlling elements, and a multiplexer 20. The demultiplexer is configured to receive a multiplex

Description

Technical Field 5 Embodiments described herein relate to photonic integrated circuits. Background In optical systems, information may be stored in the phase of an optical signal. Such an optical signal is transmitted between distant nodes using an optical channel, for example an optical 10 fibre. The optical channel may introduce phase drift, which contributes to phase noise. Compensation of such phase noise can be critical to the performance of these systems. Many optical communication systems use wavelength-division multiplexing, i.e. multiple information signals carried by separate wavelengths along the same optical channel. For 15 example, wavelength-division multiplexing can be used to increase the number of separate information signals that can be transmitted along a given optical channel. The phase noise introduced by the optical channel may affect each of the multiple information signals. In a quantum communication system, information is sent between a transmitter and a receiver 20 by encoded single quanta, such as single photons. A bit of information can be encoded upon a property of the photon, such as its polarization, phase, time or energy. Quantum communication systems may be used to implement quantum key distribution (QKD), which enables the sharing of cryptographic keys between two parties. Some QKD methods 25 require an accurate measurement of the phase of the photon received by the receiving party. Phase drifts caused by the optical channel can deteriorate the performance of phase-based QKD methods. Typically, the single photons that carry the transmitted information cannot be used to reliable determine the phase drift. Instead, using wavelength-division multiplexing, a strong reference signal may be combined into the optical channel and used to monitor 30 changes of the optical channel. Brief Description of the Drawings Figure 1 is a schematic of an example optical system in accordance with an embodiment. Figure 2 is a schematic of a variation on the optical system of Figure 1 in accordance with an 35 embodiment. Figure 3 is a schematic of a further variation on the optical system of Figure 1 in accordance with an embodiment. 09 05 24 Figure 4 is a schematic of an example optical communication system in accordance with an embodiment. Figure 5 is a schematic of a QKD network in accordance with an embodiment. Figure 6 is a further embodiment of an optical system in accordance with an embodiment. 5 Detailed Description In an embodiment, a photonic integrated circuit is provided. The photonic integrated circuit is formed on an integrated chip and comprises a demultiplexer, a first, second and third phase controlling element, and a multiplexer. The demultiplexer is configured to receive a multiplexed 10 optical signal comprising a first signal having a first wavelength and a second signal having a second wavelength different to the first wavelength. The demultiplexer is further configured to demultiplex the multiplexed optical signal to extract the first signal and the second signal. The first phase controlling element is configured to provide a first phase shift to the first signal to produce a first modified signal. The second phase controlling element is configured to provide 15 a second phase shift to the second signal to produce a second modified signal. The multiplexer is configured to multiplex the first modified signal and the second modified signal to produce a modified multiplexed signal. The third phase controlling element is configured to provide a third phase shift to the modified multiplexed signal to produce a further modified multiplexed signal, and to output the further modified multiplexed signal. 20 The photonic integrated circuit can be used to shift the phase of wavelength multiplexed optical signals in a flexible manner. The photonic integrated circuit enables shifting the phase of all multiplexed signals collectively as well as shifting the phase of the signals individually. This enables a variety of applications of the photonic integrated circuit. In an embodiment, the 25 photonic integrated circuit may be part of an optical system or optical network, and may be used to perform phase compensation such that phase noise introduced by optical channels is corrected. In an embodiment, the photonic integrated circuit may be used to reduce phase noise in an optical communication system. 30 Providing all components of the photonic integrated circuit on a single integrated chip translates into a chip-based device that can have a much smaller footprint, lower cost and higher manufacturing yield compared to conventional systems comprised of discrete, off-chip components. The photonic integrated circuit enables a compact and inexpensive device that is suitable for volume production. Further the photonic integrated circuit can be monolithically 35 integrated with other photonic integrated devices on a single chip. In contrast conventional systems typically are not compatible with monolithic integration. By enabling scalability, 09 05 24 volume production and monolithic integration compatibility the photonic integrated circuit allows for extended architectures over conventional technologies. Further the photonic integrated circuit may be polarization selective and polarization 5 maintaining. In contrast conventional systems typically suffer from polarization drift or polarization dispersion. In an embodiment, an optical Mach-Zehnder interferometer comprises a first arm and a second arm, and the first arm comprises the photonic integrated circuit. In this case, the optical 10 Mach-Zehnder interferometer with the first and second arm, and the photonic integrated circuit may be formed on a single integrated chip. In an embodiment, an optical ring interferometer comprises the photonic integrated circuit. 15 In an embodiment the photonic integrated circuit may be used to stabilise an optical phase in an optical phased array. The first and the second phase controlling elements may be thermal phase shifters and the third phase controlling element may be an electro-optic phase modulator. 20 Alternatively, the first, the second and the third phase controlling elements may be either thermal phase shifters or electro-optic phase modulators. The demultiplexer and the multiplexer may be integrated (de)multiplexer, for example arrayed 25 waveguide gratings, angled multimode interferometer, or distributed wavelength multiplexer. In an embodiment, an optical system is provided comprising a transmitter and a receiver. The transmitter comprises an encoding unit configured to encode information using phase on a first optical information signal, the first information signal having a single first wavelength. The 30 transmitter is configured to output a reference signal. The reference signal has a reference wavelength which is different to the first wavelength. The transmitter further comprises a first multiplexer configured to multiplex the first information signal and the reference signal to produce a multiplexed signal and output the multiplexed signal to a communication channel. The receiver comprises a photonic integrated circuit formed on an integrated chip. The 35 photonic integrated circuit comprises a first demultiplexer, a first, a second and a third phase controlling element, and a second multiplexer. The first demultiplexer is configured to receive the multiplexed first signal, and demultiplex the multiplexed signal to extract the first phase 09 05 24 controlling element and the first information signal. The first phase controlling element is configured to provide a first phase shift to the reference signal to produce a modified reference signal. The second phase controlling element is configured to provide a second phase shift to the first information signal to produce a modified information signal. The second multiplexer is 5 configured to multiplex the modified reference signal and the modified information signal to produce a modified multiplexed signal. The third phase controlling element is configured to provide a third phase shift to the modified multiplexed signal to produce a further modified multiplexed signal, and to output the further modified multiplexed signal. The receiver further comprises a second demultiplexer configured to demultiplex the further modified multiplexed 10 signal received from the transmitter to extract the first information signal and the reference signal. The receiver further comprises a decoder configured to decode the phase information in the first information signal, and a phase compensation unit configured to estimate the phase change of the first information signal caused by the communication channel from the first reference signal and to compensate, using the photonic integrated circuit, the phase change 15 of the first information signal caused by the communication channel. In the above, phase noise affecting the information signal can be compensated based on the reference signal. By using a reference signal which does not carry the sensitive information of the first information signal, the reference signal can be bright, i.e. of high enough intensity to 20 provide a good signal-to-noise ratio, and can be used to provide feedback to correct for the phase drift in the information signal. Also, by using a reference signal which has a different wavelength to the information signal, it is possible to wavelength multiplex the information signal and reference signal and thus the rate of information transmitted to the receiver is not reduced by additionally sending the reference signal. 25 Further in this embodiment, the optical system may also comprise a second transmitter, wherein the second transmitter comprises an encoding unit configured to encode information using phase on a second optical information signal, the second information signal having said first wavelength, the transmitter being configured to output a second reference signal having 30 the reference wavelength. The transmitter further comprises a multiplexer configured to multiplex the second information signal and the second reference signal to produce a multiplexed second signal and output the multiplexed second signal to a communication channel. The decoder is configured to decode the phase information in the first and second information signals. The phase compensation unit is configured to estimate the phase change 35 of the first and second information signals caused by the communication channel from an interference of the first and second reference signals, and to control the third phase shift such that the phase change of the first and second information signals caused by the 09 05 24 communication channel is reduced. The phase compensation unit is further configured to estimate a remaining phase change of the first and second information signals caused by the communication channel from an interference of the first and second information signals, and to control the second phase shift such that the phase change of the first and second 5 information signals caused by the communication channel is further reduced. The above is an example of a three node network where the first transmitter is a first node, the second transmitter is a second node and the receiver is in a third node. Such a network can be used in twin field QKD “TF-QKD” where the first and second transmitters both send 10 signals to the receiver and the receiver performs first order optical interference between the signals from the first and second transmitters. By disclosing the results of the interference, it is possible for the two transmitters to establish a secret key. In the above, the first and the second phase controlling elements may be thermal phase 15 shifters and the third phase controlling element may be an electro-optic phase modulator. The third phase controlling element can be used to implement fast feedback, i.e. to stabilise fast drift introduced by optical channels. For example, the fast feedback system can be implemented by monitoring the interference of the bright reference. From the interference of the bright pulse, an error signal is derived and used to adjust the phase shift applied by the 20 third phase controlling element. The fast feedback system may provide near-instantaneous correction of the phase drift. In some cases, the fast feedback may not perfectly correct the phase drift for the information signal. Any such remaining phase drift can be compensated by the first and the second phase controlling element. 25 The optical system may be configured as a quantum communication system which is configured to distribute a key between the first sending unit and the second sending unit. In broad terms, optical system 1, illustrated in Figure 1, may be used to control the phase of wavelength-division multiplexed optical signals which are exchanged between two nodes of 30 an optical network. In one application, the optical system 1 may be used to control the phase of the multiplexed optical signals such that phase noise introduced by an optical channel is compensated. The optical system 1 includes a transmitter 2, a photonic integrated circuit (PIC) 4, and 35 receiver 6. 09 05 24 The transmitter 2 is configured to provide a multiplexed optical signal, and comprises a first multiplexer 10 with first to fifth input channels 81 to 85. The first to fifth input channels 81 to 85 of the multiplexer 10 are configured to receive respectively first to fifth input signals. The first to fifth input signals have respectively first to fifth wavelengths that are different from one 5 another. The first multiplexer 10 is configured to perform wavelength-division multiplexing such as to multiplex the first to fifth input signals onto a single optical channel 12. In this embodiment, the first multiplexer 10 is configured to join the first to fifth input signals together such that they may be transmitted over the same optical channel 12. The optical channel 12 may be an optical fibre. 10 The PIC 4 is configured to receive the multiplexed optical signal transmitted from the transceiver 2 over the optical channel 12, and to provide controlled phase shifts to the first to fifth input signals before transmitting the phase-shifted signals to the receiver 6, as will be described below. 15 The PIC 4 comprises a demultiplexer 14 and first to fifth photonic waveguides 161 to 165 which comprise respectively first to fifth phase controlling elements I81 to 185. The PIC 4 also includes a second multiplexer 20, and a further phase controlling element 22. The demultiplexer 14 is configured to demultiplex the received multiplexed optical signal 20 transmitted from the transceiver 2 such as to extract the first to fifth input signals, and to provide the first to fifth input signals respectively to the first to fifth waveguides 161 to 165. The first to fifth waveguides 161 to 165 connect outputs of the demultiplexer 14 to inputs of the second multiplexer 20, and are configured to propagate respectively the first to fifth input 25 signals extracted by the demultiplexer 14 to the second multiplexer 20. The first to fifth waveguides 161 to 165 include respectively first to fifth phase controlling elements 181 to 185 configured to provide respectively first to fifth phase shifts to optical signals propagating in the waveguides 161 to 165. Each of the first to fifth phase shifts may be different from one another. Thus the first to fifth phase controlling elements 181 to 185 are configured to produce first to 30 fifth phase-shifted signals. In this example, the phase controlling elements I81 to 185 are configured to provide individual and controllable phase shifts to each of the input signals extracted by the demultiplexer 14 before these signals propagate to the second multiplexer 20. 35 The second multiplexer 20 is configured to perform wavelength-division multiplexing such as to multiplex the signals received from the phase controlling elements 181 to 185, and to provide the multiplexed output to the further phase controlling element 22. In other words, the second 09 05 24 multiplexer 20 is configured to join the first to fifth phase-shifted signals together such that they may be provided as input to the further phase controlling element 22. The further phase controlling element 22 is configured to receive the output of the second 5 multiplexer 20, and to provide a further phase shift to the received signal. The further phase controlling element 22 is configured to output a phase-shifted signal to an output port of the PIC 4 which is connected to an optical channel 24. The optical channel 24 may be an optical fibre. 10 The receiver 6 includes a second demultiplexer 26 with first to fifth output channels 28i to 28s, and is configured to receive the phase-shifted multiplexed optical signal transmitted by the PIC 4 over the optical channel 24. The second demultiplexer 26 is configured to demultiplex the received phase-shifted multiplexed optical signal such as to extract the first to fifth input signals, and to provide the first to fifth input signals respectively to the first to fifth output 15 channels 28i to 28s. In an embodiment, all components of the PIC 4 are integrated on a single photonic integrated circuit chip. This allows the PIC 4 to have a smaller footprint and to be more suitable for volume production than corresponding systems consisting of discrete off-chip optical components. 20 Furthermore, providing all components of the PIC 4 on a single photonic integrated circuit chip reduces assembly time and manufacturing costs. Another advantage over systems consisting of discrete off-chip optical components is that further photonic integrated devices can be monolithically integrated into the PIC 4 which is generally not possible for off-chip components. 25 PIC 4 may be formed of any appropriate semiconductor material system including InP, Si, SOI, SiN, SiO2, SiON, or GaAs. Alternatively, PIC 4 may be formed of glasses or polymers. Hybrid integration and heterogeneous integration techniques can be used to form PIC 4 using more than one of material system. 30 The choice of material for forming PIC 4 may be application specific. For example, forming the PIC 4 on SiN can provide a device with low optical loss, and thus may be used in applications employing weak optical signals. Further, forming the PIC 4 on Si may enable a device with a particularly small footprint, and thus may be used in applications that require small devices. Further, forming the PIC 4 on SOI or InP may enable to use both thermal phase shifters and 35 electro-optic phase modulators as phase controlling elements on the same chip without the need of integrating different material platforms. 09 05 24 In an example, an application may require polarization maintenance of the wavelength-division multiplexed optical signals sent from the transmitter 2 to the receiver 6. To this end, the structure of the waveguides and the (de)multiplexer may be designed such that PIC 4 is polarization maintaining, i.e. that the polarization of the optical signals which propagate 5 through the PIC 4 are well maintained. In this case PIC 4 overcomes a problem of polarization dispersion commonly present in conventional off-chip systems. The photonic waveguides 161 to 16s of PIC 4 may be formed using (optical and / or electron) lithography, (plasma and / or chemical) etching, direct laser writing, ion exchange, nanoimprint, 10 and the like. In general, the choice of wavelength and bandwidth of the photonic waveguides I61 to 165 of PIC 4 as well as the wavelength spacing between different photonic waveguides 161 to 165 depends on the specific requirements of the application in which the PIC 4 is to be used. 15 In an embodiment, the demultiplexer 14 and the second multiplexer 20 are arrayed waveguide gratings (AWGs). This is particularly advantageous for applications that require dense wavelength-division multiplexing (DWDM), because AWGs enable narrow waveguide bandwidth and narrow wavelength spacing between different waveguides. 20 In another embodiment, the demultiplexer 14 and the second multiplexer 20 are angled multimode interferometers (AMMIs). This is particularly advantageous for applications that require coarse wavelength-division multiplexing (CWDM), because AMMI are typically easier to design and less sensitive to fabrication tolerances. 25 In another embodiment, the demultiplexer 14 and the second multiplexer 20 are distributed wavelength multiplexers. In an embodiment, the phase controlling elements of the PIC 4 are thermal phase shifters. 30 Thermal phase shifters can provide large phase shifts, for example phase shifts of several pi. Thermal phase shifters can typically provide larger phase shifts and exhibit lower optical loss then electro-optic phase modulators. The thermal phase shifters may comprise a waveguide section, a filament, contact pads, and cladding material as a spacer between the waveguide and the filament. The phase shifting may be based on a thermo-optic effect and realised by 35 applying a driving current to the filament. 09 05 24 In another embodiment, the phase controlling elements of the PIC 4 are electro-optic phase modulators. This is advantageous because electro-optic phase modulators are typically fast phase modulators, i.e. they enable high speed phase control. Often electro-optic phase modulators can provide phase shifts faster than thermal phase shifters. 5 In a further embodiment, the PIC 4 includes thermal phase shifters in the individual waveguides 161 to 16s and an electro-optic phase modulator at the output of the multiplexer 20. This is advantageous because it allows the provision of large phase shifts at a low speed as well as small phase shifts at high speed. This is particularly advantageous for applications 10 in quantum communications, where phase noise that is introduced by optical fibres and that needs to be compensated exhibits fast fluctuations and slow drifts. In an embodiment, the first to fifth phase controlling elements I81 to 185 are thermal phase shifters. In another embodiment, the first to fifth phase controlling elements I81 to 185 are 15 electro-optic phase modulators. In an embodiment, the further phase controlling element 22 is a thermal phase shifter. In another embodiment, the further phase controlling element 22 is an electro-optic phase modulator. 20 In an embodiment, the first to fifth phase controlling elements I81 to 185 are thermal phase shifters and the further phase controlling element 22 is an electro-optic phase modulator. In an embodiment the PIC 4 includes a grating coupler or an edge coupler to receive 25 multiplexed optical signals from the optical channel 12. In an embodiment, the second to fifth phase controlling elements 182 to 185 are thermal phase shifters and the first element 181 is an electro-optic phase modulator. 30 In an embodiment, at least one of the first to fifth phase controlling elements 181 to 185 is a thermal phase shifter and at least one of the other first to fifth phase controlling elements 181 to 185 is an electro-optic phase modulator. For example, the first and second phase controlling elements 181 and 182 may be thermal phase shifters, and the third to fifth controlling elements I83 to I85 may be electro-optic phase modulators. 35 In an embodiment the PIC 4 includes a grating coupler or an edge coupler to provide multiplexed optical signals to the optical channel 24. 09 05 24 While the example photonic integrated circuit (PIC) 4 shown in Figure 1 has five photonic waveguides 16i to 16s, embodiments of the photonic integrated circuit are not limited to this example. For example, the photonic integrated circuit may have more than 5 photonic 5 waveguides. In other embodiments the photonic integrated circuit may have at least two photonic waveguides. In an embodiment the photonic integrated circuit has two photonic waveguides. In an embodiment, instead of being configured to receive the output of the multiplexer 20, the 10 further phase controlling element 22 may be configured to receive the multiplexed optical signals transmitted from the transceiver, and to provide a phase shift to the received multiplexed signals before transmitting the phase-shifted signals to the demultiplexer 14. While the example photonic integrated circuit (PIC) 4 shown in Figure 1 is described as being 15 configured to receive multiplexed optical signals from the transmitter 2 and transmit these signals (after providing appropriate phase shifts) to the receiver 6, those skilled in the art will appreciate that the photonic integrated circuit (PIC) 4 is bidirectional. For example, receiver 6 could provide multiplexed optical signals and send them to transmitter 2 via PIC 4 with the optical system 1 shown in Figure 1. 20 In an embodiment, the PIC 4 may be formed on a SOI platform. Photonic waveguides, (de)multiplexer, and any ports for receiving or providing optical signals may be formed together. To this end, patterns may be defined by photolithography and / or e-beam lithography, and corresponding structures may be formed by plasma dry etching methods. For example, 25 reactive-ion etching (RIE) methods, such as inductively coupled plasma (ICP) RIE and deep reactive ion etching (DRIE), may be used to form the structures. A thermal phase shifter may be included in the waveguides by forming a filament and providing cladding material as a spacer between the photonic waveguide and the filament, metal wires 30 and metal contact pads. The cladding material may be SiO2 or SiN, which may be deposited using (PE)CVD. The filament may be TiN, which can be deposited using a sputtering system. Those skilled in the art would understand that the filament may be formed of a wide range of materials. The metal wire and contact pads may be Au, which can be deposited using an evaporator. A further layer between the Au and the cladding may be included to act as a 35 seeding layer. The further layer may be formed of Cr / Ti / Ni. 09 05 24 Electro-optic phase modulator may be included in the photonic waveguides by incorporating a built-in PN junction into the waveguides, and by forming metal tracks and vias to connect the metal tracks and the PN junction. The metal tracks and the waveguide may be spaced using cladding materials such as SiO2 or SiN. The PN junction may be formed using ion 5 implantation. The metal tracks, vias, and contact pads may be a stack of different metals configured so as to carry RF signals. The metal tracks, vias, and contact pads may be formed using a sputtering system and / or an evaporation system. Those skilled in the art will appreciate that various modifications, for example using other 10 known micro / nano fabrication methods, may be made to the above described embodiment without departing from the scope of the present invention. During operation of the optical system 1 illustrated in Figure 1, the transmitter 2 prepares a multiplexed optical signal by providing first to fifth input signals to the first multiplexer 10 at the 15 first to fifth input channels 81 to 85. The first multiplexer 10 performs wavelength-division multiplexing such as to multiplex the first to fifth input signals onto the optical channel 12. The PIC 4 receives the multiplexed optical signal transmitted from the transceiver 2 over the optical channel 12, and provides controlled phase shifts to the first to fifth input signals before 20 transmitting the phase-shifted signals to the receiver 6. The demultiplexer 14 demultiplexes the received multiplexed optical signal transmitted from the transceiver 2 such as to extract the first to fifth input signals, and to provide the first to fifth input signals respectively to the first to fifth waveguides 161 to 165. The first to fifth phase controlling elements 181 to 185 provide respectively first to fifth phase shifts to the demultiplexed optical signals propagating in the 25 waveguides I61 to 165. Thus the first to fifth phase controlling elements I81 to 185 produce first to fifth phase-shifted signals. The second multiplexer 20 performs wavelength-division multiplexing such as to multiplex the signals received from the phase controlling elements 181 to 185, and provides the multiplexed 30 output to the further phase controlling element 22. The further phase controlling element 22 receives the output of the second multiplexer 20, and provides a further phase shift to the received signal. The further phase controlling element 22 outputs the phase-shifted signal to an output port of the PIC 4 which is connected to the optical channel 24. 35 The receiver 6 receives the phase-shifted multiplexed optical signal transmitted by the PIC 4 over the optical channel 24. The second demultiplexer 26 demultiplexes the received phase- 09 05 24 shifted multiplexed optical signal such as to extract the first to fifth input signals, and provides the first to fifth input signals respectively to the first to fifth output channels 28i to 28s. Phase noise introduced by the optical channels 12, 24 is compensated, i.e. corrected, by 5 controlling the first to fifth phase controlling elements 18i to 18s and the further phase controlling element accordingly. During operation in an example application, the receiver 6 determines an optical phase of each of the optical signals provided at the first to fifth output channels 28i to 285. The receiver 10 may use interferometers to determine the phases. Then for each of the five output channels 28i to 285, the receiver determines a difference between the determined phase and a respective predetermined phase, and controls the corresponding phase controlling element, i.e. one of the phase controlling element 181 to 185, such that the provided phase shift reduces the difference between the phase of the optical signal provided at the corresponding output 15 channel and the predetermined phase. In this example, the phase of each signal included in the multiplexed signal can be accurately controlled and corrected. In a further example, one of the first to fifth input signals, say the first input signal, is used as a reference signal to correct the phase of all input signals. In this example, the receiver 6 20 determines an optical phase of the optical signal provided at the first output channels 28i. The receiver may use an interferometer to determine the phase. Then the receiver 6 determines a difference between the determined phase and a respective predetermined phase, and controls the further phase controlling element 22 such that the provided phase shift reduces the difference between the phase of the optical signal provided at the first output channels 28i 25 and the predetermined phase. Because the further phase controlling element 22 provides a phase shift to the multiplexed signal received from the multiplexer 20, each of the optical signals included in this multiplexed signal is phase shifted. In this example, the phase of signals provided at the second to fifth output channels 282 to 285 is corrected based on the phase of the optical signal provided at the first output channels 28i. This is particularly 30 advantageous in applications, where the reference signal is a strong optical signal and the second to fifth input signals are weak optical signals, for example at the level of single photons. The further phase controlling element 22 may provide a phase shift which depends on the wavelength of the optical signal. This means when one input signal is used as reference signal, 35 as described above, the further phase controlling element 22 may not be able to correct the phase of all input signals accurately. In this case, as the wavelength dependency of the further phase controlling element 22 is generally known, additional phase shifts provided by the phase 09 05 24 controlling elements 181 to 18s can further improve the phase compensation such that all signals provided at the first to fifth output channels 28i to 285 are accurately compensated. In a further example application, where the first to fifth phase controlling elements 181 to 185 5 are thermal phase shifters and the further phase controlling element 22 is an electro-optic phase modulator, the receiver 6 determines, as described above, either the phase of each of the signals at the output channels 28i to 285 or only the phase of a reference signal, and controls the further phase controlling element 22 to provide compensation for fast phase fluctuations and the phase controlling elements I81 to 185 to provide compensation for slow 10 phase drifts. In this case the further phase controlling element 22 may provide phase correction at a high frequency, for example at a rate faster than 100 kHz. The phase controlling elements 181 to 185 may provide phase correction at a low frequency, for example at a rate slower than 100 kHz. 15 In a further example application, when one of the input signals, say the first input signal, is known to be more affected by fast phase noise than the other four input signals, the PIC 4 may comprise thermal phase shifters as second to fifth phase controlling elements 182 to 185 and an electro-optic phase modulator as first element 181. This enables fast feedback for the first input signal affected by fast noise while using thermal phase shifter with low optical loss 20 for signals that are mainly affected by slow phase drift. In a further example application where one input signal is used as reference signal, as described above, the reference signal may be provided by the receiver 6 instead of the transmitter 2. In this case the reference signal propagates in the opposite direction than the 25 optical input signals that are multiplexed and sent by transmitter 2. This is advantageous because cross-talk and interference between co-propagating signals is reduced. The aforementioned examples illustrate that the PIC 4 enables phase control of multiplexed signals for a variety of applications. 30 Figure 2 shows a variation on the optical system of Figure 1. Optical system 50 is an example of a system that uses the PIC 4 to stabilise an optical phase difference between two arms of a Mach-Zehnder interferometer. 35 The transmitter 2 is configured to provide a multiplexed optical signal as described with reference to Figure 1. In contrast to the optical system 1 shown in Figure 1, the transmitter 2 09 05 24 in the optical system 50 does not send the multiplexed signals directly to PIC 4 but into a first input port of the Mach-Zehnder interferometer. The Mach-Zehnder interferometer is formed by a first and second optical beam splitter 52, 56. 5 A first arm of the Mach-Zehnder interferometer is formed by the optical channel 54 that connects a first output of the first beam splitter 52 with a first input of the second beam splitter 56. A second arm of the Mach-Zehnder interferometer is formed by the optical channel 58, the PIC 4, and the optical channel 60. The optical channel 58 connects a second output of the first beam splitter 52 and an input port of the PIC 4. The optical channel 60 connects an output 10 port of the PIC 4 to a second input port of the second beam splitter 56. In an embodiment an overall optical path length of the first arm of the Mach-Zehnder interferometer is substantially the same as an overall optical path length of the second arm. In other words, the two arms form a symmetric Mach-Zehnder interferometer. Alternatively, the 15 overall optical path length of the first arm of the Mach-Zehnder interferometer is different from the overall optical path length of the second arm. In this case the two arms form an asymmetric Mach-Zehnder interferometer. A first output of the second beam splitter 56 is connected to the receiver 6 described with 20 reference to Figure 1. During operation of the optical system 50, the transmitter 2 sends multiplexed signals to the Mach-Zehnder interferometer. The receiver 6 may determine an optical power level at any of the first to fifth output channels 28i to 28s, and may determine from the detected power level 25 that the optical path length difference between the first and second arm has changed. In response, a feedback signal may cause the first to fifth phase controlling elements 181 to 18s and the further phase controlling element 22 to be controlled such that the change of the optical path length difference is corrected. Depending on the application, the feedback signal may be such that the path length difference is corrected for all optical signals in the multiplexed 30 signal. Alternatively, the feedback signal may be such that only the optical path length difference of the signal that power is determined is corrected. Figure 3 shows a further variation on the optical system of Figure 1. Optical system 70 is an example of an optical ring system that uses the PIC 4 to stabilise an optical path length of the 35 ring. 09 05 24 The transmitter 2 is configured to provide a multiplexed optical signal as described with reference to Figure 1. Like in the optical system 50 shown in Figure 3, the transmitter 2 sends the multiplexed signals to the first input port of the beam splitter 52. As shown in Figure 3 the first and second output port of the first beam splitter 52 are connected with each other via the 5 optical channels 72, 74 and the PIC 4. A second input port of beam splitter is connected to receiver 6 as described with reference to Figure 1. During operation of the optical system 70, the transmitter 2 sends multiplexed signals to the first input port of the beam splitter 52. The beam splitter 52 may be configured may be 10 configured to have a transmission to reflection of 99 to 1. Then the multiplexed signal circulates inside the optical ring, i.e. the beam path formed by the first beam splitter 52, the optical channels 72, 74 and the PIC 4. Because of interference of the circulating optical signals at the first beam splitter, the relative intensity of the optical signals exiting the ring via the first or second input port depends on the optical path length of the ring. It is therefore possible to 15 control the exiting port of the optical signal by controlling the phase shifts provided by the first to fifth phase controlling elements 181 to 18s and the further phase controlling element 22. The optical signals exiting the ring via the second input port of the beam splitter 52 are received and demultiplexed by receiver 6. 20 In an example application, an optical power level of the optical signals exiting from either the first or second input port of beam splitter 52 may be determined and a feedback signal may be generated based on the detected power level such that noise in the optical path length of the ring is corrected when the feedback signal is applied to the first to fifth phase controlling elements 181 to 185 and the further phase controlling element 22. 25 In an embodiment, the optical system 70 of Figure 3 is part of an optical gyroscope. In another embodiment, the optical channel 72 of the optical system 70 of Figure 3 comprises an optical gain medium configured to increase the intensity of at least one of the circulating 30 optical signals. For example, in this case the optical system 70 may be part of a fiber laser system. Figure 4 is a schematic of an example optical system. Figure 4 shows a transmitter 30 connected to a receiver 32 by a communication channel 34. 35 In this embodiment, the transmitter 30 comprises a light source 34 which emits an optical signal of a single wavelength Ai to an encoder 36. In this embodiment, the light source 34 is 09 05 24 provided within the transmitter 30. However, the light source 34 may be provided outside the transmitter 30 and a conduit provided within the transmitter to guide the light from light source 34 to the encoder 36. However, the encoder provides a phase varying modulation to the signal from the light source 34 to produce a first information signal having a single wavelength Ai. 5 The transmitter also comprises a second optical source 38 which emits an optical signal of a single wavelength Aref, the first reference signal. Ai and Aref are different to each other. However, in an embodiment, they will be similar in wavelength. 10 The first information signal and the first reference signal are then multiplexed together by wavelength division multiplexer 40 to produce first multiplexed signal. First multiplexed signal is then outputted to communication channel 34 which carries the first multiplexed signal to the receiver 32. 15 The receiver 32 comprises a photonic integrated circuit (PIC) 4’ and a wavelength division demultiplexer 36. PIC 4’ is identical to the PIC 4 described with reference to Figure 1 but PIC 4’ includes only first and second photonic waveguide photonic waveguides 16i, 162 which comprise first and second phase controlling elements I81, 182. Like PIC 4 described with reference to Figure 1, PIC 4’ enables phase corrections to be applied i) individually to the 20 signals comprised in the multiplexed signal via the first and second phase controlling elements 181, I82, and ii) collectively to all signals comprised in the multiplexed signal via the further phase controlling element 22. PIC 4’ receives the first multiplexed signal, and demultiplexer 14 demultiplexes them to provide the first reference signal to the first photonic waveguide photonic waveguides 161, and the first information signal to the second photonic waveguide 25 photonic waveguides 162. The first and second phase controlling elements I81, 182 provide respectively phase shifts to the first reference signal and the first information signal before multiplexer 20 multiplexes these signals and provides them to the further phase controlling element 22. The further phase controlling element 22 provides a further phase shift to the multiplexed signal and transmits a phase-shifted first multiplexed signal to the wavelength 30 division demultiplexer 36. The wavelength division demultiplexer 36 receives the phase-shifted first multiplexed signal and demultiplexes them to recover the first information signal and the first reference signal. The recovered first information signal is then directed into decoder 38 which decodes the 35 phase information from the first information signal. The recovered first reference signal is then directed into phase controller 40. 09 05 24 The information carried by the first information signal is encoded in phase. As the information signal passes from the transmitter 30 to the receiver 32, phase noise will occur. Thus, if it is desired to decode the phase at the decoder 38, a correction needs to be made for the phase drift. 5 This is achieved by processing the first reference signal via phase controller 40 such as to determine the phase drift experienced by the first reference signal. Phase controller 40 is then used to provide feedback to the PIC 4’ such that the first and second phase controlling elements 18i, and 182, and the further phase controlling element 22 cooperate to provide 10 phase shifts to the first information signal and the first reference signal such that the phase noise is corrected. In some communication networks, such as quantum communication networks, it is necessary to keep the intensity of the first information signal very low. Also, to preserve the security of 15 the network, is it not possible to amplify the first information signal and to therefore use this to correct for the phase drift. A specific example of the arrangement of Figure 4 applied to a twin field quantum key distribution “TF-QKD” system will now be described with reference to Figure 5. 20 Figure 5 shows a TF-QKD scheme where Alice and Bob each encode information in the phase of an optical pulse and where phase noise reduction is applied. In Figure 5, the expression ‘WDM’ is used to refer to wavelength division multiplexing and 25 demultiplexing. Alice 82 and Bob 84 generate light at a wavelength Ai, indicated by solid lines (-) in the figure, using local continuous wave (CW) lasers LS1. Alice’s LS1 86 acts as a phase reference. Its light is split in two at a beam splitter BS 88. One part is sent to Bob 84 through a service fibre 30 90, depicted by a dash-dot-dash line (-.-) and is used to lock Bob’s LS1 92 via a heterodyne optical phase-locked loop (OPLL). Both Alice 82 and Bob 84 send some of the light from their LS1 to an encoder 94. The encoder 94 performs phase and intensity modulation to output phase encoded pulses, in order enables 35 different TF-QKD protocols to be run. 09 05 24 Alice 82 further comprises a second laser LS2 that generates the bright reference signal Aref. Light from LS2 at a wavelength of Aref (denoted by dot-dot (■ ■) lines in the figure) is directed to a beam splitter where it is divided in two parts. Part of the bright reference is directed to a multiplexer where it is multiplexed with pulses from LS1 at a wavelength of Ai. The multiplexed 5 light, denoted by dash-dot-dot (-■■-) in the figure, is then sent to Charlie 98 via an optical channel (also referred to as a quantum channel). The other part of the bright reference is directed to a multiplexer that combines it with the light from LS1 from beam splitter 88. The multiplexed light from LS1 and LS2 is directed to Bob via service fibre 90. 10 At Bob 84, light from the service fibre 90 is directed to demultiplexer where the light from LS1 at A1 is separated from the bright reference at Aref. Light at A1 from Alice is directed to the OPLL which locks Bob’s LS1 to Alice’s LS1. Bright reference light at Aref is directed to another multiplexer, where it is multiplexed with pulses from Bob’s encoder 94, and then sent to Charlie via a quantum channel. 15 At Charlie 98, light received from Bob is sent to beam splitter 100. At Charlie 98, light received from Alice is passed through PIC 4’. PIC 4’ is described above with reference to Figure 4. Demultiplexer 14 separates the quantum signal from the bright 20 reference signal. The bright reference is passed through the waveguide comprising the phase controlling element 181, while the quantum signal is passed through the waveguide comprising the phase controlling element 182 which may adjust the phase of the quantum signal. The phase controlling element 182 implements a slow feedback system that will be described below. 25 The quantum signal is then recombined with the reference signal in the multiplexer 20, passed through the further phase controlling element 22 and then directed to beam splitter 100. The further phase controlling element 22 acts on both the quantum signal and the bright reference signal and may adjust the phase of the signals. The further phase controlling element 22 30 implements a fast feedback system which will be described. At beam splitter 100, light from Alice and Bob interfere. The output of the interference is directed to two demultiplexers where the interference outcome of the quantum signal (at Ai) is separated from the interference outcome of the reference signal (at Aref). At one 35 demultiplexer, the interference outcome at Ai is monitored by detector DO, and the interference outcome at Aref is monitored of detector D2. The output of D2 is directed to the further phase controlling element 22 to implement fast feedback. DO provides the output of Charlie. At 09 05 24 another demultiplexer, the interference outcome at Ai is separated and then directed to detector D1 for monitoring. The output of D1 is directed to phase controlling element 182 to implement slow feedback. 5 The output at detector DO may be output by Charlie according to a TF-QKD scheme. In a Twin field QKD (TF-QKD) scheme information is encoded in the electromagnetic phase of a photon. It is desired for Alice and Bob to exchange a key. Alice 82 and Bob 84 transmit to trusted quantum receiver 98. TF-QKD has been developed for the situation where the security of the measurement devices owned by Bob might be in doubt. In TF-QKD, the user Bob 84 is 10 configured as an optical transmitter, similarly to the other user Alice 807. The two optical transmitters Alice 82 and Bob 84 send light pulses to a relay station, usually called “Charlie”, which optically couples and measure them. Alice and Bob can distil a secret key from the publicly announced results of Charlie’s counts. In TF-QKD the users Alice and Bob are both configured as optical transmitters, therefore the security is not threatened by the vulnerabilities 15 of the optical receiver. Protecting optical transmitters is far easier than protecting optical receivers. In the former case, the optical pulses are prepared locally by a trusted user, whereas in the latter they are received from the outside, prepared by someone who is untrusted and possibly interested in breaking the security of the system. It is worth noticing that if Charlie is evil and does not comply with the correct execution of the TF-QKD protocol, 20 the two honest users Alice and Bob can always detect his attempt at cheating with very high probability by the laws of quantum mechanics. In this simplified scenario, a common fixed phase reference <pR is available to all the users all the time. As the phase reference is common to everybody and constant, it can be assumed 25 without loss of generality that <pR = 0. Alice 82 has a phase locked light source 86 and a phase modulator 94. The phase locked light source 86 generates optical pulses with constant phase, and outputs them to the phase modulator 94. The output of the phase modulator 94 is then multiplexed with the reference signal by WDM as explained above. 30 Bob’s transmitter 84 is configured in the same manner as Alice’s 82 and to avoid any unnecessary repetition like reference numerals will be used to denote like features. Alice prepares a first light pulse using her light source 86 to produce a pulse and then encodes her secret information in the electromagnetic phase difference between the light pulse and the 35 phase reference <pR using phase modulator 94. In this particular example, the encoding of the BB84 protocol [C. / 7. Bennett and G. Brassard, Proc, of IEEE Int. Conf, on Comp. Sys. Sign. Process. (IEEE, New York, 1984), pp. 175-179] is considered, where Alice encodes a random 09 05 24 “basis”, either Z or X, by selecting a phase value aA = 0 ora^ = n / 2, respectively, and a random “bit”, either 0 orl, by selecting a phase value ^ = 0 or^ = n, respectively. The optical pulse prepared by Alice will then carry a total electromagnetic phase aA + 0A. Then 5 Alice moves to the next pulse and repeats the procedure. Bob performs similar steps with phases aB and^B. The total electromagnetic phases of the pulses exiting Alice and Bob’s modules are indicated by <pA and <pB respectively: Alice: (pA = aA + pA (1) Bob: <pB = aB+(3B (2) Since all the phases are stable, Alice 82 and Bob’s 84 phase values remain constant during the propagation through the communication channels. When the optical pulses reach Charlie’s 10 non-polarising beam splitter, they undergo a so-called “1st-order interference”, which is of the same kind as the one seen in a double-slit interference experiment and in standard QKD. This means that in order to interfere deterministically, the phases of Alice and Bob’s pulses should satisfy the following interference condition: <pB — <pA = 0 mod rr, (3) where “modir” means “addition modulon”. As the phase values associated to the bits are 15 either 0 or nr, Eq.(3) reduces in this case to the following condition about the matching condition of the bases: aB-aA = 0. (4) If this condition is satisfied, then when ^-^ = 0, (5) the light emerges from the port connected to detector 0, whereas when PB~PA=^: (6) the light emerges from the port connected to detector. Therefore, after Charlie announces his 20 counts and after Alice and Bob announce their bases, Alice and Bob can reconstruct the bit value prepared by the other user in all cases where the bases match. In case of non-matching bases, the users discard the data, as in the standard BB84 protocol. In an embodiment, Charlie announces all the instances where exactly 1 of his detectors clicked. For these instances he also announces which detector clicked. 25 Another possibility is that Charlie announces also when both his detectors clicked. These double clicks are useless for the final key and can be treated in two ways: 1) Alice and Bob discard the runs where Charlie announced the double clicks; 2) Alice and Bob transform a double click into a single click by deciding at random which of 30 Charlie’s detectors clicked. The security is the same in both cases. In an example DO and D1 are single photon detectors. According to an example, DO and D1 are superconducting nanowire single photon detectors 5 (SNSPDs). According to a further example, the SNSPDs are Single Quantum EOS 410 CS cooled at 2.9 K. In an example, detector D2 is the same as DO or D1. Alternatively, D2 is a photodiode detector. 10 Note that to compensate for the different expansion / contraction rates and changes in length of the quantum channels connecting Charlie to the two users, Alice and Bob, the pattern encoding of one user with respect to the other may be delayed by an experimentally determined amount, such that the pulses arriving at Charlie's BS 821 are time-aligned. The amount of delay may be tuned at regular intervals to maintain time alignment. In an example, 15 the amount of delay is tuned between once every 4 minutes, up to once every of 30 minutes. Phase stabilisation may use the count rate of detector D1 of dim references or D2 of bright references as feedback signals. 20 With short integration intervals, the counts detected by D1 or D2 can be written as: 09 05 24 C = Co + Ci(1 - cos AS), (7) where Co represents the count floor, while Ci is the amplitude of the interference between the reference pulses, and AS is the phase difference which drifts rapidly due to phase noise from 25 the channel. For the feedback provided by D2, Co = 0, while Ci is the count rate of the reference signals (at Aref) prepared by Alice 82 and Bob 84. For the feedback provided by D1, Co is the count rate associated with the phase encoded pulses sent over Ai, while Ci is the count rate of the dim reference signals (phase unmodulated) sent by Alice 82 and Bob 84 over Ai. For the feedback provided by D1, Co = Ci when the intensity and probability (of occurrence) 30 of the phase encoded pulses and reference pulses are equal. The locking point for the stabilisation may be chosen to be at the quadrature point (A3 = tt / 2). Near the locking point of A3 = tt / 2, the count rate is approximately a linear function of the phase drift and therefore allows improved phase compensation. 35 The locking point is obtained by varying the phase shifts provided by phase controlling element I82 or the further phase controlling element 22 such that the phase drift can be counteracted 09 05 24 and a constant count rate of Co + Ci can be maintained. The phase compensation is provided by PID controllers that receive as input the number of photons collected by the photon counter connected to detectors D2 or D1 and that modify control voltages configured to control the phase provided by phase controlling element 182 or respectively the further phase controlling 5 element 22 through its amplified 12-bit DAC. In an example, the phase provided by the further phase controlling element 22 is corrected every 5 us, while the phase provided by phase controlling element 182 is corrected every 10-100 ms. The fast feedback strategy described above enables stabilising the quantum channel without 10 affecting the encoding in the wavelength reserved for the quantum signal (Ai) or the clock rate at which quantum signal pulses may be encoded. In an example, the clock rate is set at 500 MHz. The fast feedback system is configured to stabilise fast drift introduced by the optical channel. 15 The fast feedback system is configured to provide a fast error signal that can in turn be used to stabilise the fast drift. The stabilisation method comprises monitoring the interference of the bright reference at a wavelength Aref. From the interference of the bright pulse, a signal (i.e. an error signal) is derived and used to adjust the phase shift applied by the further phase controlling element 22. The fast feedback system may provide near-instantaneous correction 20 of the phase drift. For example, in an embodiment, fast feedback is fast enough to be capable of correcting phase drifts arising over the communication channels. In the case of optical fibres for example, already over short distances (tens of kilometres) the phase drifts are in the order of tens of radians per millisecond. In this example, the further phase controlling element 22 may be an electro-optic phase modulator. 25 The bright reference signal provides a bright optical signal that enables the feedback system to operate at a fast rate. The bright reference signal (Aref) may be brighter than the signals of interest (Ai) because they are transmitted at a different wavelengths. 30 The fast feedback system comprises a closed loop cycle that locks the interference between Alice's and Bob's bright reference beams to a given intensity level. This, in turn, locks the phase offset between these signals to a fixed value. The bright reference interference is monitored by the detector D2. 35 In an example, the bright reference comprises single photons, which are detected by D2 are integrated over a time period. In an example, the time period is 5 ps. The intensity value of the bright signal can take value within a range defined by a maximum and a minimum level. The 09 05 24 minimum level is the minimum intensity that the reference has to have in order to provide an error signal fast enough to correct for the channel phase drift. The maximum level is the maximum intensity that the reference signal can have before it introduces too much error on the signal of interest (or quantum signal). If the intensity of the reference signal is too high, 5 effects such as inelastic scattering along the communication channel (Raman scattering), or limited optical isolation of the WDMs, will leak noise photons into wavelength of the signal of interest, which is a problem. As noted above, Detector D2 doesn’t have to be a single photon detector. It could well be a 10 photodiode. In that case, since photodiodes requires more light to provide useful information, the operator will have to test that the reference signal does not introduce too much noise on the signal wavelength Also in the case of the integration time, that will depend on the application. Mostly on the 15 magnitude of phase noise that has to be corrected. But since the phase drifts introduced by a long communication channels are in the orders of tens of rad per second, the integration time for the information used by the quantum feedback will have to be in the range of few (tens of) microseconds, to hundreds of nanoseconds. 20 The difference between the integrated number of counts and a set value, constitutes the error signal of a PID controller. In an example, the PID controller is an FPGA clocked at 200 kHz. By tuning the control voltages for the further phase controlling element 22 that acts on the light coming from Alice, the FPGA controls the interference between the bright references. 25 Note that the phase shift applied by the further phase controlling element 22 affects both the wavelengths Aref and Ai. According to an example, the feedback based on Aref fully stabilises the bright reference light while it only partially stabilises the quantum signal at Ai. The remaining (slow) phase drift on Ai is related to two factors: (i) Ai and Aref may travel 30 separately in certain sections of the network, and (ii) the fast feedback introduces a phase drift over Ai when the path length difference seen by Ai and Aref varies over time. The former component of the slow phase drift can be seen as the phase noise picked up by an asymmetric Mach-Zehnder interferometer having the dimensions of those sections of the network where the two wavelengths travel separately. The latter component can be explained as a 35 consequence of the finite range of the further phase controlling element 22, and of the phase locking of the fast feedback over Aref, rather than Ai. 09 05 24 The further phase controlling element 22 in the fast feedback actively compensates the fast phase drift. However, its finite adjustment range may not compensate the entirety of the phase drift caused by fibre length variation. It relies on multiple (M) resets in order to maintain the Aref phase difference to $ = 2ttM + <|)t, where <|)t is the target phase. Due to the Aref - Ai 5 wavelength difference, this compensation will introduce a residual phase drift (△<)>) over Ai equal to: △<]) = 2ttM x (AREf - Ai) / Ai. The residual drift introduced by the AREF-stabilisation over Ai is estimated to be <t> / A<|) = Ai / (Aref - Ai) » 1000 times smaller than the original fibre phase drift, if assuming unidirectional fibre 10 length drift. In practice, the fibre length drift direction is random. With cancellation of positive and negative 2tt resets, a substantially higher reduction factor of may be obtained. The phase compensation scheme acting on the residual phase drift on Ai is also referred to as the slow feedback system. The error signal is obtained from the interference of the quantum 15 signals at Ai and obtained from the detector D1. The difference between this value and a set value provides the error signal for a PID controller implemented with a micro-controller. The micro-controller corrects the phase offset by controlling the phase shift provided by phase controlling element 182 acting on the quantum signal coming from Alice. Unlike the stabilisation based on the further phase controlling element 22 which acts on both Aref and Ai, 20 the slow feedback acts solely on the quantum signals (Ai) and can therefore correct its residual phase drift. In this example, a thermal phase shifter can be used. Different from the fast feedback, the slow feedback does not require high operation bandwidth. Thus, thermal phase shifters that vary the optical path length are suitable the slow feedback. However, other types of phase modulator could be used for this task for example, electro-optic modulators. 25 Note that for the slow feedback system, the quantum signal at Ai comprises pulses encoded with information as well as dim reference pulses. The dim reference pulses may be interleaved with the information encoded pulses. In an example, the dim reference pulses have the same intensity as the brightest information carrying pulse. The presence of the dim reference pulses 30 provides an interference output that is related to the residual phase offset in Ai. The interference output is retrieved by integrating the single photons detected by D1 over a time period. In an example, the time period is 50 ms or 100 ms, depending on the distance of the optical channel. The difference between this value and a set value provides the error signal for a PID controller implemented with a micro-controller operating at the frequency of 20 Hz 35 or 10 Hz, depending on the distance of the optical channel. 09 05 24 According to an example, the further phase controlling element 22 is an electro-optic modulator, wherein the refractive index of the material is a function of applied electric field. Changes in refractive index result in changes in optical path length and results in changes in the phase shift applied by phase modulator. Different voltages are applied to the phase 5 modulator so as to impart a different phase shift. The phase modulator such as described can comprise a crystal, such as a lithium niobate (LiNbO3) crystal, in which the refractive index is a function of electric field strength, and an electric field may be applied by applying a voltage to electrodes positioned around the LiNbO3 crystal. In an embodiment, the further phase controlling element 22 has a high operation bandwidth. For example, the further phase 10 controlling element 22 uses the electro-optic effect (maximum bandwidth in the order of GHz). The relative phase shift imparted by the further phase controlling element is set by the PID controller described above which may be configured to apply a control voltage signal to the further phase controlling element 22. 15 Note that a phase modulator provided in the encoder 94 may be similar to the PIC described above. Alternatively, the adjustment of the phase shift can be achieved through either tuning a phase modulator in the encoder 94, or adding a control voltage signal to the driving signal applied to the phase modulator in encoder 94. 20 By using a reference signal which does not carry the sensitive information of the first information signal, it is possible to amplify the reference signal and use this to provide feedback to correct for the phase drift in the first information signal. Also, by using a reference signal which has a different wavelength to the information signal, it is possible to wavelength 25 multiplex the information signal and reference signal and thus the rate of information transmitted to the receiver is not reduced by additionally sending the reference signal. Figure 6 is a schematic of an example optical system 200 comprising a transmitter 202 configured to transmit optical signals to receiver 206 via the optical channel, a photonic 30 integrated circuit 204, and optical channel 214. The photonic integrated circuit 204 comprises a photonic waveguide 208 comprising a first and a second phase controlling element 210, 212. The photonic integrated circuit 204 is configured to receive the optical signal transmitted by transmitter 202 and to provide the optical signal to waveguide 208. The first phase controlling element 210 is configured to provide a first phase shift to the received optical signal. 35 The second controlling element 212 is configured to provide a second phase shift to the received optical signal. The first phase controlling element 210 is a thermal phase shifter and the second phase controlling element 212 is an electro-optic phase modulator, or vice versa. 05 24 The photonic integrated circuit 204 may be manufactured similar to the photonic integrated circuit 4 described with reference to Figure 1. In an embodiment, the transmitter 202 of the optical system 200 may encode information in the phase of the transmitted optical signal such that the receiver 206 can decode the optical signal to retrieve the information. 5 During operation, photonic integrated circuit 204 may be used to compensate phase noise introduced by the optical channel. In general terms, the transmitter 202 sends alternatingly signals carrying information and reference signals. To compensate phase noise, the receiver 206 detects the reference signals, determines a feedback signal so as to compensate the 10 phase noise, and uses the first phase controlling element 210 to compensate for slow phase drifts and the second phase controlling element 212 to compensate the fast phase fluctuations. Whilst certain embodiments have been described, these embodiments have been presented by way of example only, and are not intended to limit the scope of the inventions. Indeed, the 15 novel devices, and methods described herein may be embodied in a variety of other forms; furthermore, various omissions, substitutions and changes in the form of the devices, methods and products described herein may be made without departing from the spirit of the inventions. The accompanying claims and their equivalents are intended to cover such forms or modifications as would fall within the scope and spirit of the inventions. 20 07 11 25

Claims

1. A photonic integrated circuit formed on an integrated chip comprising:a demultiplexer;a first, a second and a third phase controlling element; and5 a multiplexer;the demultiplexer configured to:receive a multiplexed optical signal comprising a first signal having a first wavelength and a second signal having a second wavelength different to the first wavelength, and10 demultiplex the multiplexed optical signal to extract the first signal andthe second signal;the first phase controlling element configured to provide a first phase shift to the first signal to produce a first modified signal;the second phase controlling element configured to provide a second phase15 shift to the second signal to produce a second modified signal;the multiplexer configured to multiplex the first modified signal and the second modified signal to produce a modified multiplexed signal, andthe third phase controlling element configured to provide a third phase shift tothe modified multiplexed signal to produce a further modified multiplexed signal, and to output 20 the further modified multiplexed signal,wherein the first and the second phase controlling element are thermal phase shifters and the third phase controlling element is an electro-optic phase modulator.

2. The photonic integrated circuit according to claim 1, wherein the demultiplexer and / or 25 the multiplexer are arrayed waveguide gratings.

3. The photonic integrated circuit according to claim 1 or 2, wherein the demultiplexer and / or the multiplexer are angled multimode interferometer.30 4. The photonic integrated circuit according to any preceding claim, wherein thedemultiplexer and / or the multiplexer are distributed wavelength multiplexer.

5. A method of operating a photonic integrated circuit formed on an integrated chip, the method comprising:35 receiving a multiplexed optical signal comprising a first signal having a first wavelengthand a second signal having a second wavelength different to the first wavelength;07 11 25demultiplexing, using a demultiplexer, the optical signal to extract the first signal and the second signal;providing, using a first phase controlling element, a first phase shift to the first signal to produce a first modified signal;5 providing, using a second phase controlling element, a second phase shift to thesecond signal to produce a second modified signal;multiplexing, using a multiplexer, the first modified signal and the second modified signal to produce a modified multiplexed signal;providing, using a third phase controlling element, a third phase shift to the modified10 multiplexed signal to produce a further modified multiplexed signal, andto outputting the further modified multiplexed signal,wherein the demultiplexer, the first phase controlling element, the second phase controlling element, the third phase controlling element and the multiplexer are comprised in the photonic integrated circuit,15 wherein the first and the second phase controlling element are thermal phase shifters and the third phase controlling element is an electro-optic phase modulator.

6. The method according to claims 5, wherein the demultiplexer and / or the multiplexer are arrayed waveguide gratings.

207. The method according to claim 5 or 6, wherein the demultiplexer and / or the multiplexerare angled multimode interferometer.

8. The method according to any one of claims 5-7, wherein the demultiplexer and / or the 25 multiplexer are distributed wavelength multiplexer.

9. An optical Mach-Zehnder interferometer with a first arm and a second arm, wherein the first arm comprises a photonic integrated circuit according any one of claims 1 to 4.3010. An optical ring interferometer comprising a photonic integrated circuit according any one of claims 1 to 4.

11. The optical ring interferometer according to claim 10 further comprising an optical gain 35 medium connected to the photonic integrated circuit via an optical channel and configured to increase the intensity of at least one optical signal circulating in the optical ring.07 11 2512. An optical system comprising a transmitter and a receiver, said transmitter comprising an encoding unit configured to encode information using phase on a first optical information signal, said first information signal having a single first wavelength, said transmitter being configured to output a reference signal, said reference5 signal having a reference wavelength which is different to the first wavelength, the transmitter further comprising a first multiplexer configured to multiplex the first information signal and the reference signal to produce a multiplexed signal and output the multiplexed signal to a communication channel,said receiver comprising:10 a photonic integrated circuit formed on an integrated chip comprising:a first demultiplexer;a first, a second and a third phase controlling element, anda second multiplexer;the first demultiplexer configured to:15 receive the multiplexed first signal, anddemultiplex the multiplexed signal to extract the reference signal and the first information signal;the first phase controlling element configured to provide a first phase shift to the reference signal to produce a modified reference signal;20 the second phase controlling element configured to provide a second phaseshift to the first information signal to produce a modified information signal;the second multiplexer configured to multiplex the modified reference signal and the modified information signal to produce a modified multiplexed signal, andthe third phase controlling element configured to provide a third phase shift25 to the modified multiplexed signal to produce a further modified multiplexed signal, and to output the further modified multiplexed signal;a second demultiplexer configured to demultiplex the further modified multiplexed signal received from the transmitter to extract the first information signal and the reference signal;30 a decoder configured to decode the phase information in the first information signal;anda phase compensation unit configured to estimate the phase change of the first information signal caused by the communication channel from the first reference signal and to compensate, using the photonic integrated circuit, the phase change of the first information 35 signal caused by the communication channel.07 11 2513. The optical system according to claim 12, wherein the first and the second phase controlling element are thermal phase shifters and the third phase controlling element is an electro-optic phase modulator.5 14. The optical system according to claim 12 or 13, further comprising:a second transmitter, wherein said second transmitter comprises an encoding unit configured to encode information using phase on a second optical information signal, said second information signal having said first wavelength, said transmitter being configured to output a second reference signal having said reference wavelength, the transmitter further comprising 10 a multiplexer configured to multiplex the second information signal and the second reference signal to produce a multiplexed second signal and output the multiplexed second signal to a communication channel,and wherein the decoder is configured to decode the phase information in the firstand second information signals; and15 the phase compensation unit is configured toestimate the phase change of the first and second information signals caused by the communication channel from an interference of the first and second reference signals;control the third phase shift such that the phase change of the first and second information signals caused by the communication channel is reduced;20 estimate a remaining phase change of the first and second information signals causedby the communication channel from an interference of the first and second information signals, andcontrol the second phase shift such that the phase change of the first and second information signals caused by the communication channel is further reduced.2515. A quantum communication system comprising the optical system according to claim 14, wherein the quantum communication system is configured to distribute a key between the first sending unit and the second sending unit.30

Images