NO318004B1 - Method and apparatus for a variable optical attenuator - Google Patents

Description

The present invention relates to a device and a method as well as applications of a variable optical attenuator (VOA), and in particular to control light in an optical fiber by an adjustable dynamic grating in a system for optical communication, according to the appended , independent claims 1 and S, and in the embodiments specified in the attached non-independent claims.

The demand for bandwidth for communication and information exchange has grown exponentially. The transmission capacity has been able to accelerate particularly thanks to the introduction of optical wavelength multiplexing technology to be able to send several independent optical signals in the same optical fiber. Active components are required in addition to the passive optical fiber to generate, amplify, route and filter signals. This has led to the development of a wide range of technologies for manipulating light in an optical fiber. Such components include filters, switches, amplifiers and attenuators. Unfortunately, the high cost of more advanced components slows down the development of optical communication systems and the introduction of all-optical networks. It is therefore necessary to develop cost-effective components that have the necessary specifications, but still allow low-cost production and assembly. Optical communication systems are in use within telecommunications, local networks, regional networks, distribution of television and video, instrumentation networks, etc. together with all other types of communication system where optical methods are best suited for communicating symbols, messages and signals.

A component of particular demand in fiber optic communication systems is the variable optical attenuator or attenuator. Attenuators are used as independent components, e.g. to compensate for aging effects in other components and to avoid saturation of optical detectors. However, there are more dynamic network structures such as in a fully optical network, where the signal strength from different sources or transmission paths varies significantly. This makes it necessary to introduce reconfigurable or dynamically variable optical attenuators. Variable optical attenuators are also central parts of modules such as signal equalizers and optical multiplexers that can switch wavelengths both in and out of an optical fiber. It is in particular the ability to scale up the technology in the number of optical channels that will determine the final price of the module.

The following list of publications represents previous innovation in the field: GB 2,265,024 - Geoffry Martland Proudly -15.09.1993 - A spatial light modulator assembly, US 3,835,346 - Fred Mast et. al -10.09.1974 - Cathode Ray Tube, US 5,867,301 - Graig D. Engle - 02.02.1999 - Phase Modulating Device, US 4,879,602 - 07.11.1989 - William E. Glenn et. al - Electrode Patterns for Solid State Light Modulator, US 5,116,674 - 05/26/1992 - Beat Schmidhalter et. al - Composite Structure, US 5,221,747 - 22.06.1993 - Gerald R. Goe et. al - improved Process and Catalyst for the Preparation of 2,2'Bipyrdilys, US 4,529,620 - 16.07.1985 - William E. Glenn - Method of Making Deformable Light Modulator Structure, US 4,857,978 - 15.08.89 - Efim Goldburt et. al - Solid State Light Modulator Incorporating Metallized Gel and Method of Metallization, US 4,900,136 -13.02.1990 - Efim Goldburt et. al - Method of Metallizing Silica-Containing Gel and Solid State Light Modulator incorporating the Metallized Gel, WO 99/09440 - 25.02.1999 - Foster Miller inc. - Switch able Optical Components, WO 01/48531 - 05.07.01 - Yury Guscho - Optical Systems.

Several realizations have been proposed for tunable diffraction gratings for use in fiber optic components. One of the known methods is diffractive microelectromechanical systems (D-MEMS). This technology is available from e.g. LightConnect and Silicon Light Machines. These devices are based on a movable diffraction grating consisting of at least two separate parts. A stationary reflective bottom surface and a moving set of thin blades (the grating) are both etched out of silicon. The leaves can be moved up and down using an appropriate electric field. This results in a diffraction grating where the effective phase shift of the light is given by the relative position of the blades and the reflecting surface below. This allows the grid to be turned on and off (activated and de-activated) with a response time of a few milliseconds. However, a voltage of the order of ten to a few hundred volts is needed to move the blades. This device can be used to implement efficient variable attenuators, but the blades must be etched out of silicon. This is an expensive process where the yield drops dramatically as the system size increases. Components realized as D-MEMS are therefore efficient, but expensive.

The present invention aims to have the same performance as D-MEMS, but with as simple and uncomplicated manufacturing process as Liquid Crystal (LC) or Liquid Crystal On Silicon (LCOS). The present invention is based on adjustable surface diffraction gratings. Such gratings have been published in the technical literature and in patents. E.g. then our preferred realization is based on the arrangement described in articles and books published by Guscho in Russia (Guscho: Physics of Reliefography, 1992 Nauka Moscow) and in the international patent application WO 01/48531 by Yuri Guscho. These examples of optical systems are mainly focused on projector technology. However, embodiments in accordance with the present invention can also be based on modulators with surface coating as described by Engle (US 5,867,301). The basic principles of these modulators are well known and have been used for projector purposes since the introduction of the Eidophor projector almost 50 years ago. However, the contrast of the light on the screen is very important for projector applications. These applications have therefore been based on utilizing the light in the first and second order diffraction maxima. For use in fiber optic components, the light in the zero-order diffraction maximum is used.

British patent application GB 2 265 024 describes a spatial light modulator composed of a deformable material layer which can be deformed in response to variation in an electrostatic field or potential difference. The technical solution described in this patent application is based on the physical phenomenon called frustrated total internal reflection (FTIR) from a surface that is always unmodulated and therefore always perfectly plane. See page 11 (bottom) and page 12 (top): 'The advantages of the assemblies in figures 1 and 2 are that the reading beam only sees a flat surface which is the boundary surface 7 and does not pass through or be reflected by a physically deformed surface such as in a conventional modulator using a deformable layer<*>.

The incoming reading trawl has such an angle of incidence to the flat surface 7 that total internal reflection (TIR) is achieved. TIR is characterized by the existence of a so-called evanescent field on the side of the reflective surface which is opposite to the incoming beam. This evanescent field does not transport any energy, and the field strength is proportional to the expression e(' 2lt'^ where z represents the distance to the reflecting surface 4 and A. is the optical wavelength in the gas medium (see page 50 of the '7th Edition of'Principles of Optics ' by Max Born and Emil Wolf, Cambridge University Press').The effective penetration depth into the gas medium 6 is of the order of l/2n which is smaller than the optical wavelength.

A modulated surface is located near the reflective surface, but on the opposite side relative to the incoming reading beam. Medium 6 is a gas and lies between 7 and 10. This means that the deformable material 5 and/or layer 10 is in contact only with the evanescent field from the reading beam. It is therefore only the evanescent field from the read beam that is affected by the local spatial modulation of the refractive index. See last sentence on page 8 and first on page 9: 'Amplitude modulation of the incoming read beam 2 is created by local disturbance of the total internal reflection (TIR) of the beam undergoing TIR at the surface 7 between the first layer 1 and the gas in the space 6 ,....This perturbation of the total internal reflection leads to frustrated TIR and a local change in the reflection coefficient of the laser beam at that point.'.

The part of the incoming light beam that is not totally reflected due to the frustrated TIR, is transmitted through layer 7 and absorbed by subsequent layers and/or designations (see middle of page 3).

One aspect of the present invention is based on the incoming light being reflected from a surface which is itself spatially modulated and therefore perfectly flat only when full TIR is desired.

Another aspect of the present invention is that the spatially modulated surface and/or its surface treatment (if present) is therefore always in contact with the non-evanescent field of the incoming light.

A further aspect of the present invention is that no part of the incoming light is absorbed in the component: the entire light intensity propagates out of the component, only the direction is controlled by the surface modulation of the gel or membrane.

A further aspect of the present invention is that it allows zero diffraction order light to be adjusted continuously from full intensity to an attenuation of 20 dB,

or more. Together with a combination of driver electronics and optical solutions for sending the light in and out of the modulator, the invention constitutes a method and device for an adjustable surface diffraction grating for applications such as variable optical attenuator in optical communication systems. Figure 1 shows a device in accordance with an example of an embodiment of the present invention. Figure 2 shows an example of a layout of the silicon chip in accordance with an example of the embodiment of the present invention.

Figure 3 shows theoretical calculations of the intensity in different diffraction maxima.

Figure 4 shows a side perspective of the modulator in accordance with the design shown in Figure 1. Figure 5 shows the intensity in various diffraction maxima as a function of the amplitude of the gel modulation for conical diffraction. Figure 6 shows the intensity in different diffraction maxima as a function of the amplitude of the gel modulation for diffraction with a plane of incidence normal to the electrode direction. Figure 7 shows the intensity of each diffraction maxima in conical diffraction as a function of the thickness of the sawtooth grating.

Figure 8 shows an example of the embodiment of the present invention.

Figure 9 shows a block diagram of an example embodiment of the present invention. Figure 10 shows a double-pass configuration operating in the Fraunhofer regime, a fiber is used to link the collimator pair; Figure 11 shows a multipass configuration with mounted collimators operating in the Fresnel regime, the light beam is 0.4 mm, the collimators are 2.8 mm in diameter and the thickness of the glass where multipass occurs is 0.7 mm. Figure 12 shows the layout of a ten-channel device in accordance with the present invention. Figure 13 shows feedback in an example of an embodiment according to the present invention. Figure 14 shows a typical operation sequence in a feedback system in accordance with an embodiment of the present invention. Figure 15 shows two 1x2 switches side by side and the distribution of optical intensity between the two outputs of the lx2 switch as a function of the gel amplitude. Figure 16 shows an example of the DGE element with integrated VOA element;

Figure 17 shows the VOA and switching function.

Figure 18 shows the formation of a sawtooth profile in accordance with the present invention.

The present invention is based on diffraction of light using modulation of the surface in a thin gel layer or a membrane with corresponding optical and mechanical properties. The basic modulator design and principles are shown in Figure 1.

The modulator consists of a thin gel layer (or membrane) attached to a transparent prism. The refractive index of the gel (membrane) is adapted to that of the prism, and the gel (membrane) has a low absorption of light in both the visible and the infrared range (less than 2% for a typical system). The gel layer is typically 15-30 micrometers thick. The electrodes are processed on a flat substrate (fig. 2) separated from the gel surface by a thin air gap (5-10 micrometers thick). The distance can be arranged in various ways that will be known to a person skilled in the art.

A bias voltage is set up across the gel and the air gap. The result of the electric field is a force that affects the gel surface. In addition, it is possible to individually address each individual electrode. By applying a local signal voltage, the surface is affected by forces that result in a local surface modulation. The response time of the elastic surface is fast and of the order of a few tens of microseconds. However, there are also various charging and relaxation processes over longer time scales in the gel. This results in a slow operation of the surface modulation with a time scale from microseconds to minutes. The present invention has arrangements for feedback to achieve stable operation (described later in the text).

The typical dimension of the active area (electrodes) on the example of a modulator embodiment shown in Figure 2 is 3mm x 6mm. The electrodes are placed as opposing chambers with 125 line pairs/mm as maximum resolution in this example of an embodiment of the present invention. The resolution is limited by the high precision required for a small gel thickness and narrow air gap required for high density of the surface modulation.

A periodic variation of the drive voltages on the individual electrodes creates an approximately sinusoidal surface modulation with the same period as the voltages. Typical values for bias and signal voltage are 100 V.

The modulator acts as a diffraction grating when the light is reflected by the modulated surface. For high line density, conical diffraction should be used to reduce polarization-dependent effects and transmission loss. The incoming light beam then has the same direction as the electrodes and the associated depressions and elevations in the surface that occur due to the surface modulation. The result is that the light is diffracted into higher order diffraction maxima.

Figure 3 shows the intensity in zero, first and second order diffraction maxima based on theoretical calculations (apart from zero order, all other maxima occur in pairs). The intensity is wavelength dependent, and the associated phase shift is to first order proportional to a/X where a is the amplitude of the surface modulation and Å is the optical wavelength of the light. The variation of wavelength over the C-band (approx. 1530-1570 nm) typically produces a variation in the attenuation of 1%. By varying the amplitude of the surface modulation, the zero-order damping can be varied continuously. At an optical wavelength of 1550 nm, a modulation amplitude of 300 nm is required to achieve full zero-order attenuation. Figure 3 shows that the maximum intensity in the first order diffraction maximum at full attenuation of zero order is approximately 30% (for one of the two first order maxima).

If the prism is 5 mm high, 10 mm wide and has a refractive index of 1.45, then the following distances can be measured on the surface of the prism if light of optical wavelength 1554 nm is used: Distance between zero and first order diffraction maxima is 0.67 mm.

The distance between the first and second order diffraction maxima is 0.70 mm.

These distances define the positions where optical fibers can be connected to the surface of the prism as is known to a person skilled in the art.

The gel arranged as a modulator is a programmable diffraction grating. The grid can e.g. realized as a sinusoidal gel surface with adjustable amplitude. An electric field set up by electrodes controls the gel amplitude, while the distance between the electrodes gives the grating period. The modulator is a reflection grating with an incidence angle of approx. 45 degrees (see figure 4).

The diffraction can be conical or diffraction with a plane of incidence normal to the electrode direction, depending on the orientation of the incoming beam relative to the electrode pattern.

Configuration 1: Conical diffraction

Conical diffraction occurs when the prism top is normal to the electrode lines. Figure 5 shows the intensity of different diffraction maxima as a function of the gel amplitude. The grid period in this example is 8 micrometers which corresponds to 125 line pairs/mm.

Advantages of conical diffraction:

- All incoming light is reflected by the grating, no light is transmitted through the gel.

- Less polarization dependence

- The zero-order diffraction maximum can be attenuated to very close to zero.

Configuration 2: Diffraction with a plane of incidence normal to the electrode direction Diffraction with a plane of incidence normal to the electrode direction occurs when the prism top is parallel to the electrode lines. The incident plane is spanned by the incoming light beam and an imaginary normal to the grating plane at the point where the light beam hits the grating. Figure 6 shows the light intensity in different diffraction maxima as a function of the gel amplitude. The grid period in this example is 8 micrometers which corresponds to 125 line pairs/mm.

Advantages of diffraction with a plane of incidence normal to the electrode direction:

All diffraction maxima lie on a straight line. This makes design ag

assembly easier.

Positive and negative diffraction maxima are asymmetric. This gives higher

light intensity in certain maxima, which is interesting for switching applications.

Asymmetric diffraction grating

It is desirable in some applications of adjustable surface diffraction gratings within optical communication to be able to steer an incoming light beam in a single direction.

However, there is a symmetrical light distribution between diffraction maxima of negative and positive order in both conical and planar diffraction: For example, the maximum intensity of light in ±1. diffraction maxima equal to 33.9% in +1. maximum and 33.9% in -1. maximum (this is valid for monochromatic light and a sinusoidal gel modulation). An asymmetric distribution of light between different diffraction maxima can be achieved by an asymmetric surface modulation of the gel. One of several possible solutions is a sawtooth profile. The present invention shows a simple method and device for creating a sawtooth-shaped surface modulation. Such a sawtooth shape will distribute almost all light between zero and a specified diffraction maximum. This is suitable for switching applications. Figure 7 shows the intensity in each diffraction maximum as a function of the gel amplitude in an example of such a design of the present invention.

• Mathematical background

The simplest sawtooth profile can be obtained by a sum of two harmonic components, as can be seen from equations (l)-(3) and from figure 18.

If the amplitude of the second term F2 (x) is less than 20 or greater than 30, then the result will be a less close approximation to a sawtooth profile compared to Figure 18. The same applies to the phase term of the second term: The value n/2 gives the best approach to a sawtooth profile.

Practical implementation

A sawtooth-shaped potential distribution in the electrode plane can be obtained by the electrode structure shown in figure 18. Bold dashed lines symbolize linear electrodes, the value above each individual line is an example of an electric voltage applied to this electrode. The top line corresponds to the term Fj (x), the middle line corresponds to the term F2 (x) and the bottom line corresponds to an electrode structure which alone gives the sum F£ (x). The fields Fj (x) and F2 (x) will modulate the surface with respective components Aj (x) and A2(x) of the surface profile.

It should be taken into account that the sawtooth-shaped potential profile F in the electrode plane shown in figure 18 will not give a surface modulation A in the surface plane with exactly the same shape for two reasons: Firstly, the amplitude of the spatial harmonics in a potential distribution will decrease when you move away from the electrode plane to the gel surface. That reduction can take place approx. a factor of two faster for the harmonic term F2(x) than for the term Fj (x) because F2 (x) has twice the spatial frequency (half the spatial period of Fj (x)). Secondly, the amplitude of the harmonic component A2(x) of the surface modulation given by the field component F2(x) will be reduced in relation to A] (x) even if the fields Fj (x) and F2 (x) had been of the same strength at the gel surface as it is the case for a so-called ponderomotive force (force due to the inhomogeneous field). The sum A, (x) + A2 (x) will therefore be a less exact approximation to a sawtooth profile than the sum F, (x) + F2 (x).

To compensate for this, a relatively higher amplitude should be used for F2 (x) relative to the amplitude of Fj (x). Overall, this will give a good approximation of a sawtooth profile.

Driving techniques

A memory effect creates unfortunate consequences when examples of devices based on the present invention are driven by electrical signals. The gel is affected by the surface modulation if the same electrical signals are applied over a long period of time. Various techniques can be used to solve this:

• Alternating driving

The electrodes can be configured so that every second is applied with a voltage and every other is the ground. An electrode pair consisting of an earth electrode and an electrode applied voltage gives a sinusoidal period. By periodically inverting the voltage configuration, memory effects can be avoided. However, light passing during the reconfiguration time would not be controllable and such a drive arrangement for communication purposes should include a light blocking mechanism as is known to a person skilled in the art.

• Rolling alternating driving

In rolling alternating driving, the voltage configuration only changes in a few electrode pairs at a time. The communication signal can be sent continuously through the modulator. Any problems caused by local phase shifts in the diffraction grating can be compensated for with devices elsewhere in the light path.

• Static drive with feedback

A memory effect will occur during static operation. A feedback system can be used to allow the electrical drive signals to compensate for this.

A surface pattern is set up on the gel when voltage signals are applied to the electrodes. This surface pattern deflects or attenuates optical signals. When the voltage signal is removed, the gel surface does not immediately return to its original state.

The result of that memory effect is that in a VOA application there is not a predictable relationship between the attenuation achieved and the applied signal voltage

Therefore, the measured attenuation level of the VOA is used as a feedback signal to control the signal voltage of the modulator. The feedback system is shown in Figure 13.

The main controller provides the desired attenuation level to the feedback controller. The optical input and output levels are measured, and the feedback controller measures the attenuation as

the modulator performs. This damping level is compared with the desired value and results in a correction signal. The drive voltage is adjusted based on this correction signal. A typical operation sequence is shown in Figure 14.

The left vertical axis in Figure 14 shows the voltage difference E = VI - V2 between the electrodes in the electrode pair. The right vertical axis shows the damping level A. At the start time, the damping level is Aj with a stable signal voltage Ei. At time ten, the main controller receives a new desired damping level A2. The feedback controller adjusts the signal voltage to E2 to achieve the new attenuation level. Because of. the memory effect in the gel, the signal voltage will over time be adjusted down to a stable level E3. At time t2, the main controller switches the desired value back to A]. Because of. the memory effect reduces the feedback signal voltage to E4. This can be a negative voltage which indicates that the signal voltage V2 applied to the electrode has a positive voltage relative to the signal voltage VI applied to the electrode. After some time the signal voltage will approach a stable level E5.

The present invention establishes gel technology as an advanced material platform that has the potential to create more electrically adjustable products for optical communication. To satisfy requirements in such communication applications, conical diffraction (low optical loss, low polarization dependence) and static driving with feedback (high precision, no loss of information) should be preferred designs.

Three core functions can be implemented by these methods in accordance with the present invention: Light intensity modulation, spectral filtering and optical switching. In the following paragraphs, examples of embodiments are given that demonstrate these functions.

Two basic examples of VOA are shown. The two applications are described below.

VOA based on substrate with 125 line pairs/mm

The following functions are implemented in this exemplary embodiment of the present invention:

- variable damping

- variable coupler/pin (switch)

- spectral selectivity/filter

- monitoring (feedback)

The example of execution consists of an array of 2 attenuators with monitoring on the optical chip. It includes two incoming optical fibers and 4 outgoing optical fibers (see Figure 8). ±1. order diffraction maxima can be used as taps to deflect a limited part of the optical effect in order to monitor the optical effect in each individual incoming optical signal. Spectral selectivity can also be achieved by using the angular selectivity in collimating optics.

Feedback from ±1. order diffraction maxima can be used to optimize accuracy.

Example of execution based on a substrate with 33 line pairs/mm

Substrates with 33 line pairs/mm should also be considered as they are simpler and less expensive to produce than substrates with a higher line density. With a line density of 33 line pairs/mm, the angular separation between different diffraction maxima will be small. This complicates the practical manufacture of devices with several optical outputs which are possibly to be connected to optical fibres. The example of the embodiment of the present invention with a substrate with 33 line pairs/mm based on only two optical input signals and two optical output signals. Higher order diffraction maxima are not used. The facility consists of:

• A modulator with the possibility to control the gel for two optical channels

A programmable electronic board for controlling the attenuation level in both optics

channels via feedback

A functional diagram of the design is shown in figure 9, the layout of the substrate is shown in figure 12.

Two photo detectors per optical channel provides information about the optical power level at the input and output of each optical channel. The control electronics use this information to optimize the voltage levels on the gel to attenuate the optical channels with great precision to achieve a nominal value.

Two main functions are required:

1. The user defines a power level at the output of each optical channel. In addition, own remedies transfer this value to the control electronics which then optimize the signal voltages to attenuate the input signal until the desired optical power levels on the outputs are achieved.

Example: Pin (channel 1) = 12mW, Pin (channel 2) = lOmW

Set by user: Put (channel 1) = 6mW, Put (channel 2) = 0.1 mW

The modulator and the control electronics attenuate both signals until the desired levels are reached (6 mW at the output of the first channel and 0.1 mW at the output of the second optical channel). The measured values from the two photodetectors on the outputs are used for optimization in the feedback. 2. The user defines an attenuation level for each of the two optical channels. The control electronics optimize the signal voltages to attenuate the two optical input signals until the desired attenuation level is reached.

Example: Piim (channel 1) =12mW, Pil)n (channel 2) =10mW

Set by user: attenuation level (channel 1) =3dB, attenuation level (channel 1) =20dB

The modulator and the control electronics dampen both optical input signals until the user's request is reached, i.e. 6 mW from the first optical channel and 0.1 mW from the second optical channel. The ratio between the measurement values given by the photodetectors at the input and output of each optical channel is used for optimization in the feedback.

This exemplary embodiment of the present invention provides dynamic balancing of optical signals.

Two optical channels with time-varying optical power levels are sent to the two inputs of the device. The device dynamically identifies and attenuates the channel with the highest optical power until both channels have the same power level at the output. Balancing can be achieved in less than 1 ms.

Finally, in another implementation, it may be interesting to connect two optical channels in series to achieve higher attenuation levels.

Different multipass configurations can be used to counteract polarization-dependent effects such as e.g. polarization dependent losses.

Conceptually, there are two ways to implement a multipass configuration of the modulator. The first method is to capture the zero-order diffraction maximum after the first diffraction and then allow this light to be diffracted again. This allows for very high damping levels. A collimator or lens must be used to capture only zero diffraction maxima by operating in the Fraunhofer regime as shown in Figure 11.

The second method is to reflect back the zero diffraction maximum of the diffracted beam and allow it to be diffracted again. In this case, the distance between two diffractions is very short and we operate in the Fresnel regime. With this arrangement, the total phase shift will be the sum of all phase shifts that occur when the light beam hits the diffraction grating. If the light beam hits the diffraction grating N times, the gel amplitude can be reduced by a factor N without the resulting attenuation level changing.

1x2 switch

A simple 1x2 switch can be implemented by attaching collimators at the zero and first order diffraction maxima. By turning the surface modulation on and off, the light can be switched from one starting position to the other (by continuously adjusting the grating amplitude, this can also be used as a continuously adjustable pin). A sketch of a possible design is shown in Figure 15.

However, a modulator based on a symmetric grating (e.g. with sinusoidal surface modulation) will transmit only 30% of the light to each of the two first-order diffraction maxima, and the intensity of the zero-order diffraction maximum is also not minimal for this grating amplitude.

A sawtooth-shaped grating (as described earlier) offers the possibility of improving this dramatically: Up to 90% of the incoming optical power can be switched to a first-order diffraction maximum (see figure 15).

Dynamic balancing of optical gain

A channel balancing device or a gain equalizing device can be implemented as a combination of a free-space spectral demultiplexer and an array of VOAs. The channel balancing device consists of the VOA element in combination with a spectral demultiplexing element, e.g. a diffraction grating or a thin film filter. The demultiplexing element is dispersive so that the optical wavelengths in the input signal are spatially separated. An example of a device for dynamic balancing of optical amplification is shown in Figure 16.

Reconfigurable optical add-drop multiplexer

The modulator basically has a free-space type optical interface. It is therefore possible to do a hybrid integration of several optical functions (spectral multiplexing/demultiplexing, array of switches and VOAs) to build a reconfigurable optical add-drop multiplexer. This is a central building block in all-optical dynamic communication systems.

An example of an embodiment of a hybrid-integrated reconfigurable optical add-drop multiplexer in accordance with the present invention is shown in figure 17.

Devices for monitoring optical power level

The diffraction mechanism means that the invention can be used for monitoring optical power levels where monitoring is desired without disturbing the main part of the optical beam. This can be done by using e.g. one of the two first order diffraction maxima to tap off part of the optical effect.

Claims (14)

1. Method for modulating light in an optical communication system, characterized by the steps of: attaching a gel layer or membrane adjacent to a transparent prism that transmits and/or receives light to/from the optical communication system, providing a substrate with a plurality of individually addressable electrodes suitably spaced from a surface of the gel layer or membrane facing away from the prism, to apply electrical voltages to each of the electrodes to thereby create a pattern on the surface of the gel layer or membrane superimposed on an initial state of the gel layer or membrane, and to regulate the electrical the voltages applied to the electrodes so that the gel layer or membrane returns to its initial state when such voltages are no longer present. 2. Method according to claim 1, characterized in that at least one optical fiber is attached to the prism so that light can be passed to/from the optical communication system. 3. Method according to claim 1 or 2, characterized in that the prism is oriented over the substrate so that the direction of the prism is normal to dividing lines between the electrodes on the substrate. 4. Method according to claims 1 and 2, characterized in that the prism is oriented over the substrate so that the direction of the prism is parallel to dividing lines between the electrodes on the substrate 5. Device for modulating light in an optical communication system, wherein the device includes: a gel layer or membrane, a prism that conducts light to/from the optical communication system, a substrate with a plurality of electrodes, and a drive device for applying voltage to each of the electrodes , whereby a surface pattern in the gel layer or membrane is superimposed on an initial state of the gel layer or membrane, characterized by: - that the gel layer or membrane is attached adjacent to a surface of the prism, - that said electrodes are individually addressable and located at a distance from a surface of the gel layer or the membrane facing away from the prism, - that the drive device is arranged to apply regulated voltages to each of said electrodes, the drive device including means to compensate the voltages so that the gel layer or the membrane returns to its initial state when said voltages are not present. 6. Device according to claim 5, characterized in that at least one optical fiber is attached to the prism to conduct light to/from the optical communication system. 7. Device according to claim 5, characterized in that the prism is attached to the substrate so that the direction of the prism is normal to dividing lines between the electrodes on the substrate. 8. Device according to claim 5, characterized in that the prism is attached to the substrate so that the direction of the prism is parallel to dividing lines between the electrodes on the substrate. 9. Device according to claim 5, characterized in that the drive device is designed to apply voltage to each of the electrodes. 10. Device according to claim 5, characterized in that the drive device is designed to apply individual voltage to each of the electrodes. 11. Device according to claim 5, characterized in that the electrodes are arranged in at least three adjacent parallel rows, where each row has an equal number of addressable lines, thereby providing at least three different patterns on the surface of the gel layer or the membrane which results in optical addition of reflected or attenuated light from these three patterns. 12. Device according to claim 5, characterized in that the drive device applies a ground potential to every other electrode while the other remaining electrodes are applied a voltage, and in that the drive device periodically changes the voltages so that those that are grounded receive voltages while those that have voltages are grounded. 13. Device according to claim 5, characterized in that the drive device includes a photodiode or phototransistor which is designed to measure the emitted light level from the prism, the drive device being designed to emit an output signal which is communicated as a correction signal to the means which compensates the voltages. 14. Device according to claim 5, characterized in that the drive device includes a memory with target values for the voltages applied to the electrodes, where the memory is in communication with the means that compensates the voltages.