• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    The invention relates to a method for designing a device for multiplanar conversion of a light radiation E (p, q), the device implementing a plurality M of phase masks intercepting the light radiation in order to phase this radiation in order to to apply a certain transformation to the light radiation. According to the invention, a first and a second family of modes (u, v) with separable variables (p, q) are defined. We choose a number N of pairs of indices {i, j} k to form a first and, respectively, a second family of exploited modes by selecting the pair modes of index {i, j} k in the first family and , respectively, in the second mode family. Then we establish the phase shift magnitudes φ1 (p, q), M phase masks for transforming each torque mode of index {i, j} k of the first family of modes operated in the same pair mode. index {i, j} k of the second family of exploited modes. The invention also relates to a phase plate obtained via the design method and to a multiplan conversion device.
    公开号:FR3078412A1
    申请号:FR1851664
    申请日:2018-02-26
    公开日:2019-08-30
    发明作者:Olivier Pinel
    申请人:Cailabs SAS；
    IPC主号:
    专利说明:

    METHOD FOR DESIGNING A MULTIPLANE CONVERSION DEVICE, PHASE BLADE OBTAINED THROUGH THIS METHOD, AND MULTIPLANE CONVERSION DEVICE
    FIELD OF THE INVENTION
    The present invention relates to a method for designing a multiplane conversion device. It also relates to a phase plate and a multiplane conversion device obtained by means of this process.
    TECHNOLOGICAL BACKGROUND OF THE INVENTION
    We know from US9250454 and US2017010463 such optical devices, as well as methods for designing them. These devices (often designated by the acronym MPLC, of “Multi Plane Light Conversion” according to the English expression) make it possible to carry out any unitary spatial transformation of a light radiation.
    From a theoretical point of view, and as has been established in "Programmable unitary spatial mode manipulation", Morizur et Al, J. Opt. Soc. Am. A / Vol. 27, No. 11 / November 2010, a unitary spatial transformation can actually be broken down into a succession of primary transformations, each primary transformation affecting the transverse phase profile of light radiation.
    The document Design of High Order Mode-Multiplexers using Multiplane Light Conversion by N. Fontaine et Al, presented at the 43rd conference on optical communications, (ECOC, 2017) recalls that the design of an MPLC device consists in determining the characteristics in a plane (x, y) of a set of phase masks, each mask i introducing a phase shift of φι (χ, γ). It also recalls in detail a method of designing these masks which is based on an optimization algorithm by matching wave fronts. It evokes the possibility, as an alternative, of basing the design on a method of stochastic optimization of solutions generated by calculation.
    This document also proposes a method of designing a spatial optical multiplexer making it possible to limit the number of phase masks necessary to implement the transformation taking place on the incident light radiation. More specifically, this document discloses an MPLC device which converts an input composed of a linear beam of 45 fibers emitting Gaussian light radiation, in the first 9 groups of modes of a multimode optical fiber with an index gradient. The conversion is carried out using 14 phase masks. The output modes of the device must be represented in a base of Hermite-Gauss modes and, simultaneously, the input light beams must be rearranged in the transverse plane to arrange them in the shape of a triangle.
    This document does not give any explanation to understand how the chosen combination (and in particular the choice of the output base and the rearrangement in triangle of the input light beams) makes it possible to carry out the conversion into 14 phase masks, if this 'is to affirm that the symmetry of the modes and the arrangement in triangle could be the cause.
    The present invention aims to establish a method of designing an MPLC optical device which is not limited to the very specific example presented in this document. It aims in particular to provide a design method making it possible to implement a transformation using a relatively small number of phase masks, while giving great freedom of choice in the nature of the mode bases and in the arrangement of the beams.
    BRIEF DESCRIPTION OF THE INVENTION
    With a view to achieving this aim, the object of the invention provides a method of designing a device for multiplan converting light radiation, the device implementing a plurality of phase masks intercepting light radiation for respectively apply to it a spatio-frequency phase shift leading to a determined transformation, the design process comprising the following steps:
    - Define a first family of modes, each mode of the first family being designated by a pair of indices {i, j} l <= i <= n, l <= j <= m, the first family being with spatial variables separable;
    - Define a second family of modes, and each mode of the second family being designated by a pair of indices {i, j}, l <= i <= n, l <= j <= m, the second family
    (.Vij (p, q)) being with separable spatial variables
    - Form a first and, respectively, a second family of modes used by selecting N torque modes of index {i, j} kl <= k <= N, N being less than or equal to nm and strictly greater than n and m, in the first family and, respectively, in the second family, the N modes of the second family not being Hermite-Gauss modes and, simultaneously, the N modes of the first family not being arranged in the triangle-shaped plane;
    - Establish the spatio-frequency phase shift of the phase masks making it possible to transform each mode of torque of index {i, j} k of the first family of modes used in the mode of same torque of index {i, j} k of the second family of operating modes.
    By imposing in such a way that the first and the second family of exploited modes present the characteristic of separability of the spatial variables, and that each mode of couple of index {i, j} k of the first family of exploited modes is transformed by the phase masks in the mode with the same pair of indices {i, j} k of the second family of exploited modes, a determined transformation can be established by means of a reduced number of phase masks compared to a method design that does not impose these constraints.
    According to other advantageous and non-limiting characteristics of the invention, taken alone or in any technically feasible combination:
    - the first family of modes and the second family of modes are bases;
    - the bases of the first family of modes and of the second family of modes are orthonormal;
    - the separable variables are Cartesian coordinates;
    - the separable variables are polar coordinates;
    the establishment of the spatio-frequency phase shift of the M phase masks is carried out by optimizing a degree of correspondence between the transformation of the modes of the first family of exploited modes and the modes of the second family of exploited modes;
    - the degree of correspondence is optimized by a stochastic optimization method;
    - the degree of correspondence is optimized by matching wave fronts;
    the step aiming to establish the spatio-frequency phase shift of the phase masks comprises:
    i. Establish respectively a one-dimensional phase shift according to a first spatial variable and according to a second spatial variable;
    ii. Combine the one-dimensional phase shifts according to the first spatial variable and according to
    the second spatial variable for establish the phase space-frequency of the masks of phase. The invention relates also on a program
    computer comprising instructions adapted to the implementation of at least one of the steps of the design process, when the program is executed on a computer.
    It also relates to a phase plate manufactured from the spatio-frequency phase shift of at least one phase mask obtained at the end of the design process.
    According to another aspect, the invention provides a device for multiplan converting a first light radiation, the device comprising a conversion block comprising a plurality of optical parts and implementing a plurality of phase masks carried by at least one of the parts. optics, the phase masks applying a spatio-frequency phase shift aiming to develop a second radiation, the determined transformation respectively placing N index modes {i, j} k, l <= k <= N, of a first family of modes used with separable spatial variables (p, q) describing the first light radiation at a first transverse plane with N modes of the same index {i, j} k, l <= k <= N, of a second family of modes used with separable spatial variables describing the second light radiation at a second transverse plane, the N modes of the second family not being Hermite-Gauss modes and, simultaneously nt, the N modes of the first family not being arranged in the first transverse plane in the shape of a triangle
    According to other advantageous and non-limiting characteristics of the invention, taken alone or in any technically feasible combination:
    - The multiplane conversion device comprises at least one stage for injecting the first and / or the second radiation into the conversion block and / or extracting the first and / or the second radiation from the conversion block;
    the stage comprises a bundle of optical fibers arranged in a transverse stage plane, the first transverse plane being the transverse stage plane;
    the spatial variables of the first family of exploited modes can be separated in a Cartesian coordinate system and the optical fiber bundle of the stage is arranged in the transverse plane of stage in the form of a rectangle or square;
    the spatial variables of the first family of exploited modes can be separated in a polar coordinate system and the bundle of optical fibers of the stage is arranged in the transverse plane of stage along concentric circles and along radial lines;
    the stage comprises at least one multimode fiber;
    - at least one of the optical parts carries at least one phase mask with separable spatial variables;
    - The optical parts comprise two reflecting optical parts arranged opposite one another, the phase masks being carried by one and / or the other of the two reflecting optical parts;
    - the two optical parts consist of a microstructured phase plate and a mirror, and the phase masks are carried by the microstructured phase plate.
    BRIEF DESCRIPTION OF THE DRAWINGS
    Other characteristics and advantages of the invention will emerge from the detailed description of the invention which will follow with reference to the appended figures in which:
    - Figures 1 to 3 show by way of illustration families of modes with separable variables;
    - Figure 4 shows the steps of a method according to the invention;
    - Figure 5 shows a first example of a multiplane conversion device according to the invention;
    - Figures 6a to 6c show another example of a multiplane conversion device according to the invention.
    DETAILED DESCRIPTION OF THE INVENTION
    Light radiation is defined in the present application as a family composed of at least one mode of the electromagnetic field, each mode forming a spatio-frequency distribution of amplitude, phase, polarization. Consequently, the modification or transformation of the light radiation designates the spatio-frequency modification or transformation of at least one of the modes of the radiation.
    It will be considered in the present description that the radiation is polarized in a single direction and has a single frequency. However, the principles set out are entirely applicable to radiation having more than one direction of polarization or more than a single frequency. Thus, the spatial phase shift applied to radiation by a phase mask in the description framework can be expressed more generally as a spatio-frequency phase shift, that is to say modifying the one the phase shift according to the space variables in a plane. transverse and according to the optical frequency.
    For the sake of completeness, we recall the operating principles of a multi-plane light conversion device, more commonly designated by the Anglo-Saxon terms "Multi-Plane Light Converter" (MPLC). In such a device, an incident light beam undergoes a succession of reflections and / or transmissions, each reflection and / or transmission being followed by propagation of the beam in free space. At least some of the optical parts on which the reflections and / or transmissions operate, and which guide the propagation of the incident beam, have a microstructured surface which modifies the incident light beam.
    By “microstructured surface” it is meant that the surface of the optical part can have “pixels” whose dimensions are between a few microns to a few hundred microns. Each pixel has an elevation, with respect to an average plane defining the surface in question, of at most a few microns or at most a few hundred microns. An optical part having such a microstructured surface forms a phase mask introducing a local phase shift within the cross section of the radiation which is reflected or transmitted there.
    Thus, a light beam which propagates within the component undergoes a succession of local phase shifts separated by propagations. The light radiation thus undergoes a complex transformation. In other words, the succession of elementary transformations (for example ten or more successive transformations) establishes an overall transformation of the spatial profile of the radiation. It is thus possible to configure the microstructured surfaces to transform a first light radiation, which in particular has a specific shape, into a second radiation whose shape is different.
    It will be considered in the present description that the shape of a light radiation corresponds to the spatio-frequency distribution of the amplitude and of the transverse phase of a mode or of the family of modes composing the light radiation. It is recalled in this connection that a mode at any point in space (x, y, z) can be expressed as a function of the mode in a single plane, here z = 0, according to the Fresnel equation in the Fresnel approximation:
    P P OO
    E (x, y, z) = J ECX, y ', 0) e ^ l (' x ~ x ^ + ^ y ~ y '^ dx'dy'
    These definitions being thus posed, the design of an MPLC device aiming to transform the first radiation into the second radiation amounts to establishing a configuration of phase masks of the device such that each mode of a light radiation composed of a first family of modes ui (x, y) be transformed as closely as possible into a mode Vj (x, y) of a second light radiation composed of a second family of modes. This transformation can be theoretically perfect when the families of modes Ui and Vj form orthonormal bases.
    If the families of modes do not form orthonormal bases, one can sometimes also find a perfect transformation, or seek to approximate this transformation as well as possible. Indeed, an MPLC conversion device is not limited to the transformation of orthonormal mode bases. In a general way, that the spatial variables are separable or not, one can also conceive such a device starting from families of modes which are not necessarily orthogonal.
    The mathematical and physical demonstration of the existence of a configuration of phase masks for any unit transformation was exposed in the article by Jean-François Morizur cited in the introduction to this request. The feasibility of such a component has been demonstrated, as well as its universal character, namely the ability to carry out any unit transformation of the shape of an incident light radiation.
    A practical realization of such a component was for example proposed in the article by Guillaume Labroille and others, Efficient and mode-selective spatial mode multiplexer based on multi-plane light conversion, in Optics Express June 30, 2014 vol 22 N ° 13 p 15599 which explains the constitution of this component. More specifically, this component uses two mirrors arranged opposite one another, at least one of these mirrors carrying a plurality of microstructured zones forming the phase masks. This architecture forms a multi-passage cavity in which a radiation is propagated which is reflected several times, and in particular at the level of the various microstructured zones, on the two mirrors.
    The design of the phase masks of an MPLC device is done in practice by simulation in a computer capable of modeling the behavior of light radiation during its propagation guided by the various optical parts, and in particular those carrying the microstructured surfaces forming the masgues. phase. The computer simulates the propagation, in this succession of optical elements, of incident light radiation having an input shape and it calculates the transformed radiation which results therefrom.
    In general, to design the MPLC device, we seek to determine each phase mask leading to coinciding the transformed radiation composed of the modes Ui 1 , that is to say the modes u ± having been transformed by the device , with a desired radiation. This desired radiation is made up of all the modes of a second family Vj, sometimes designated as the output family.
    The degree of correspondence between these two radiations a can be expressed, for example, as a numerical value a = Σί | ΪΙu- (x, y) v / (x, y) dxzïy | which we therefore seek to maximize during the design process.
    This research can be posed as an optimization problem aiming to determine for each phase mask φι, the phase shift φι (Xk, yk) of each pixel arranged at a point (Xk, yk) of the microstructured surface forming the mask. By way of example, when the optical device comprises more than 10 phase masks and the pixel resolution of these masks is important, this optimization step seeks to determine a very large number of parameters, which can exceed one million.
    The document WO 2012/085046 presented in the introduction to this request proposes design methods aimed at simplifying this particularly complex optimization problem by matching wave fronts at the level of each phase mask. These methods, which are also recalled by the article by N. Fontaine, make it possible to simplify the optimization problem and converge asymptotically towards a solution. To improve the quality of the transformation obtained, a typical approach consists in increasing the number of phase masks used. However, a high number of phase masks makes the manufacture of the MPLC device particularly complex.
    The present invention takes a completely different approach. It imposes characteristics on the families of modes that the transformation associates. These constraints lead to pose a new optimization problem whose solutions form a class of transformations which is particularly simple to implement, i.e. requiring only a small number of phase masks and which are simple to manufacture.
    More specifically, the invention is based on the choice of the first and second families of modes allowing respectively the decomposition of the incident radiation and the desired radiation so that these families are with separable spatial variables.
    Recall that a field with separable variables is a field whose whose transverse profile in a plane z = 0 is written in the form E (x, y, 0) = fo (x) .go (y), where fo and go are called "generating functions". We can show that when this radiation propagates under the Fresnel approximation, it remains separable in any plane z. Indeed, the Fresnel integral can be factorized in the form:
    in other words E (x, y, z) - f z (x) .g z (y), which proves its character "with separable spatial variables" in any plane z. We note that the functions f 2 (x) and g z (y) are respectively and only dependent on fo (x) and go (y) · In other words, in any plane z, f z (x) is only determined by fo (x) · The same is true for g z (y).
    The influence of a spatial phase (p (x, y), for example during a reflection or a transmission on a phase mask, is done on the product of the field by an exponential term e t <p ( x, y The phase shift introduced by a phase mask "with separable spatial variables" is naturally defined as a sum of phase φ (x, y) = ψ (x) + Q (y)
    Thus a radiation E (x, y) with separable variables undergoing the phase shift φ (x, y) is expressed as:
    = (/ (x) e Wl >) and also remains with separable spatial variables. The phase along x only influences the component along x of this radiation, and the same goes for the phase along y. In other words, under the effect of phase masks with separable variables, the two components along x or y of the radiation remain independent of each other and "with separable variables". A mode with separable spatial variables therefore remains separable after its propagation in free space and under the influence of a phase mask with separable spatial variables.
    In the context of the present application, a family with separable spatial variables (and more concisely "with separable variables") is therefore defined, a family u composed of the modes Uij of indices {i, j}, l <= i <= n and l <= j <= m, such that for all i and for all j we can express:
    uij (x, y) = fi (x) .gj (y)
    It is specified that this characteristic of separability of the spatial variables relates well to the family of modes as a whole and that it is not sufficient that one can separate the spatial variables from each of the modes of a family for this characteristic is actually present for the family in question.
    Of course, it is not always possible, in reality, to always verify very precisely the equalities defining a field or a mask with separable variables. And we will therefore be able to tolerate deviations in respect of these equality, and nevertheless consider that we are dealing with a family or a phase mask with separable spatial variables.
    The invention takes advantage of these observations to provide a method of designing a device for the multiplan conversion of light radiation. It proposes in particular to define a first and a second family of modes, both with separable variables, and to establish the phase masks making it possible to transform each mode of the first family into a corresponding mode of the second family (this is to say the same index (i, j}).
    It is noted that this principle is very far from the hypotheses of symmetry of the modes and arrangement in triangle presented by the document of N. Fontaine of the state of the art. These hypotheses are indeed based on the results of a very specific design process in which, simultaneously, the output modes are chosen to be Hermite-Gauss modes and the input modes rearranged in the transverse plane for the arrange in a triangle shape. This document, in the absence of detailed explanations justifying the performance of the MPLC designed, cannot therefore provide education allowing to generalize this experience beyond the configuration which has been chosen.
    In no case is it also a question of positioning the modes of the first family in an arrangement of the plane which takes up the arrangement of the lobes of maximum intensity of the modes of the second family, as has also been suggested. And in addition, the modes forming the families do not have to have any axis of symmetry for a family to have separable variables.
    A design method in accordance with the invention works in particular with any second family of modes with separable variables which would not be Hermite-Gauss. It also works with any layout of input modes that is not triangular.
    To illustrate the fundamental differences between the principle of the present invention and a simple effect of symmetry proposed by the document by N. Fontaine, FIGS. 1 to 3 show several examples of families of modes with separable variables. To simplify these figures, a limited number of modes has been shown, but of course the invention finds all its advantage when the number of modes processed by the MPLC device is large, greater than 10, 20, or even 50 or 100 modes.
    Thus, in FIG. 1, a mode family uij (x, y) has been represented in a plane provided with a Cartesian coordinate system (x, y). This family of modes is with separable variables because all the modes Uij (x, y) can be expressed as the product uij (x, y) = f ± (x) .gj (y) for any couple (x, y) plan. The generating functions are expressed here in the form of Gaussians centered on the points of the plane (ί.Δχ, j. Δγ) where the steps Δχ and Δγ are chosen to be constant. Another arrangement could be envisaged that does not fix the constant Δχ, Δγ steps while retaining the separable character of the family.
    FIG. 2 shows, also by way of illustration, a second example of a family of modes with separable variables in a Cartesian coordinate system. In this figure 2, we observe that the Uij modes are not arranged uniformly on a regular grid of the plane (x, y). The generating functions gi, g2, g3 are centered on points respectively separated by a step Δγΐ, Δγ2 distinct from each other. It can also be noted that the generating functions fi, Î2, Î3 are of very different natures. The generating functions need not have any symmetry.
    In practice, to design an MPLC device, it will be possible to choose any family of modes with separable variables which is suitable. Some families, however, are particularly helpful. These include:
    - of the family composed of modes whose generating functions (fi, gi) are Gaussian, as shown in FIG. 1 and as obtained at the output of a bundle of optical fibers;
    - of the family of Hermite-Gauss modes which form a mode base of an optical fiber with a parabolic index gradient;
    - from the family of elliptical Hermite-Gauss modes.
    The principle of separability of variables is not necessarily limited to a plane provided with a Cartesian coordinate system. The same theoretical principles presented above are verified in another coordinate system, such as a polar coordinate system. One can then in this case pose the principle of separability of the variables of a family of modes by the expression:
    Uij (r, ô) = fi (r) .gj (0) and a mask with separable variables like:
    φ (r, θ) = ψ (τ) + θ (0).
    As an example of this choice of coordinate system, there is shown in FIG. 3 a family of modes with separable variables in a polar coordinate system as well as the generating functions fi and gj.
    In order not to be dependent on the chosen coordinate system, we will generically designate in the rest of this presentation by the letters (p, q) the spatial variables of the chosen coordinate system, knowing that these variables can indifferently correspond to (x, y) in a Cartesian coordinate system or at (r, 0) in a polar coordinate system.
    These general principles having been laid down, the invention therefore proposes a method for designing a multi-plane MPLC conversion device, implementing a plurality of phase masks to apply a determined transformation. The process is schematically represented by the steps of FIG. 4.
    The design process therefore aims to numerically determine the phase masks allowing this transformation to be carried out. As we have seen, these masks are defined by a plurality of pixels. However, for the sake of simplification of the description which follows, the phase shift imposed in the plane by the mask 1 will be expressed as a function <pi (p, q), by masking its discrete character.
    A step of a method according to the invention aims to define a first family of modes u, each mode (uij) of the first family being designated by a pair of indices {i, j} l <= i <= n , l <= j <= m. This first family is with separable variables so that we can express uij (p, q) = fi (p) .gj (q);
    Another step of a method in accordance with the invention aims to define a second family of modes v, each mode of the second family (v ± j) being designated by a pair of indices {i, j}, l <= i <= n, l <= j <= m. The second family is also with separable variables so that one can express Vij (p, q) = f '± (p). g 'j (q);
    These first stages of the design method therefore make it possible to define the families of modes with separable variables which will be transformed towards each other by the MPLC device. These two families have the same cardinality n.m. However, it is not necessary that all the modes {i, j}, l <= i <= n, l <= j <= m making up these families are actually used by the device.
    Thus, in the illustrative example of FIG. 1, among the nine modes of the family represented, it is possible to choose to use only a part of them in the MPLC device. One could for example choose not to exploit the modes one and U33. In this case, modes of the second family must also be eliminated in order to preserve the same cardinality of the families. To also preserve the property of separability of the variables during the propagation, the reflections and / or transmission of the radiation, it is important that each mode of the first family of index {i, j} is transformed into the mode of the second family of the same index {i, j}. This therefore leads to eliminating the modes seen and V33 from the second family of modes in the example taken in illustration.
    More generally, a design method according to the invention comprises a step aimed at choosing a number N, less than or equal to nm and strictly greater than n and m, of pairs of indices {i, j} k, l < = k <= N and form a first and, respectively, a second family of modes exploited by selecting the torque modes of index {i, j} kl <= k <= N in the first family and, respectively, in the second family.
    It is noted that the separability character of the spatial variables is preserved in the first and the second family of exploited modes.
    To exploit the full potential of the design process, we choose this number N of modes used so that it is strictly greater than n and m. That is to say that the number of modes used does not consist in simply choosing a row, a column or a diagonal of the family of modes, when these are for example arranged in a plane as shown in the figure 1.
    As we have seen, the mode families can advantageously form an orthonormal base, but this characteristic does not constitute a necessary limitation of the invention.
    According to the invention, the transformation sought associates each mode of torque of index {i, j} k of the first family of modes used in the mode of the same couple of index {i, j} k of the second family of modes exploited. Thus, each mode of torque of index {i, jh of the first family of modes used is mapped after it has been propagated in the MPLC device with the mode of the same torque of index {i, j} k of the second family of operating modes. This preserves the "separable" character of the quantities <pi (p, q) defining the phase masks.
    When the N modes of the second family of exploited modes are all Hermite-Gauss modes and that, simultaneously, the chosen modes {i, j} k of the first family are arranged in a plane in the shape of a triangle, we find again in a way fortuitous the solution proposed in the article by N. Fontaine of the state of the art.
    Once these constraints relating to the families of imposed modes and their associations, one can deploy in a following stage of a method according to the invention, an optimization stage aiming to establish the quantities of phase shift cpi (p, q ), 1 <= 1 <= M, M phase masks allowing the first family of exploited modes to be transformed as closely as possible into the second family of exploited modes.
    This optimization step, as such, is not necessarily different from that of the prior art. It therefore aims to establish the phase shift magnitudes <pi (p, q), 1 <= 1 <= M, M phase masks maximizing a degree of correspondence a between the transformation of the modes of the first family of exploited modes and the modes of the second family of exploited modes.
    And the invention is not limited to a particular optimization method. It can thus be a method of stochastic optimization or by approximation of wave fronts as described in the documents of the state of the art.
    At the end of this optimization step, there is therefore a digital model çi (p, q) of each of the M phase masks, making it possible to apply a determined transformation to an incident radiation. When this optimization is well carried out, each of the M phase masks has the property of being with separable variables or is close to having this property.
    The first and second families of operating modes need not form the input and output families of the MPLC device. In general, it can be expressed that the determined transformation respectively relates the N modes of the first family of exploited modes describing the first light radiation at the level of a first transverse plane, with the N modes of the same index of the second family of operating modes describing the desired light radiation at a second transverse plane. The M phase masks make it possible to carry out the transformation between the first and the second transverse plane.
    We can provide some additional phase masks arranged upstream and / or downstream of the M phase masks necessary for transforming the inputs to the first mode family at the first transverse plane or the second mode family at the second level. transverse plane towards the exits.
    According to a first approach, these complementary masks are directly added to the digital model of the MPLC device during the design process which has just been described. We then seek to determine, during this process, the characteristics of all the phase masks, including the complementary masks. During the optimization stage, it can be sought to impose that the families of modes used at the level of the first transverse plane and at the level of the second transverse plane are well in accordance, respectively, with the first and with the second family, and therefore with separable variables.
    According to another approach, the complementary masks are determined during a second iteration of the optimization step in which they are added to the digital model of the MPLC device provided with the M phase masks established at the end of the first iteration .
    In this regard, we note that a design process in accordance with
    The invention can provide a plurality of iterations of the optimization step.
    These successive iterations can seek to adjust the phase masks obtained at the end of a previous iteration by the exposed design method. It may be a question of adding complementary phase masks, as we have just seen, or of making the masks more manufacturable, for example by imposing constraints on the spatial variations in phase shift of a mask. The adjustment of the phase masks which is carried out by these successive iterations can lead to these masks (or some of them) no longer verifying, finally, the separability relation of the spatial variables.
    However, in certain cases one can find a mask or a plurality of masks imposing a phase shift <pi (p, q) with separable variables, that is to say that çu (p, q) satisfies the relation <pi (p, q ) = ψ 1 (ρ) + θ 1 (ς).
    According to a particularly advantageous variant of the invention, the optimization step aimed at developing the quantities pi (p, q), 1 <= 1 <= M, of M masks can directly take advantage of their "separable" nature.
    Indeed, using the same notations for the family of modes Uij (p, q) = f ± (p). gj (q) and the second family of modes Vij (p, q) = f 'i (p). g'j (q) we can simplify the digital model of the MPLC device to reduce it to a one-dimensional optimization problem. In this case, the M phase masks with indices 1, ψι (ρ) are sought, such that the generating functions fi (p) for i = 1 to n are respectively transformed into the generating functions f'i (p). Simultaneously, the M one-dimensional phase and index masks 1 0i (q) are determined such that the generating functions gj (q) for j = 1 to m are respectively transformed into the generating functions g'j (q). In other words, a one-dimensional phase shift is established respectively according to a first spatial variable (p) and according to a second spatial variable (q).
    One can then combine the one-dimensional phase shifts according to the first spatial variable (p) and according to the second spatial variable (q) to establish the spatio-frequency phase shift of the phase masks. More precisely, the phase masks <pi (p, q) 1 <= 1 <= M are determined by the expression <pi (p, q) = ψι (ρ) + 0i (q). The algorithmic complexity of optimization is reduced by this approach, and greatly the power or the computation time necessary to achieve this optimization.
    In general, a design method in accordance with the invention is advantageous in that it is possible to work, directly or indirectly during optimization, on the n + m relationships between the generating functions fi, fi ', gi, gi' rather than on the relationships between nm modes. The complexity is reduced from one degree n.m to one degree max (n, m), so we understand the very particular interest of this approach.
    Whatever the optimization method chosen to establish the numerical model <pi (p, q) of the M phase masks, it can be used to manufacture by any suitable technique (for example by machining, molding and / or photolithography and etching of a raw optical part) and which aims to microstructure one or more reflection-transmission surfaces of an optical part, to make it a phase plate. In other words, at least one phase mask cpi (p, q) obtained at the end of the conforming design process can be used to manufacture a phase plate. When the optical part, such as a phase plate, is microstructured according to a pattern resulting from the implementation of a method according to the invention, the microstructured surface does not have the shape of a triangle.
    The model can also be implemented by other means, for example by “programming” the spatial phase shift imparted by a deformable mirror or a spatial light modulator (generally designated by the acronym SLM according to the Anglo-Saxon denomination “Spatial Light Modulator ”).
    With reference to FIG. 5, an MPLC device 1 designed and manufactured using a design method according to the invention comprises a conversion block 2 comprising a plurality of optical parts 2a, 2b and implementing a plurality of masks phase 5 carried by at least one of the optical parts 2a, 2b, the phase masks (5) applying a spatial phase shift <pi (p, q) aimed at developing a second radiation.
    Thus, in the example of FIG. 5, the architecture of the conversion block 2 has been chosen so that it comprises two reflecting optical parts 2a, 2b arranged opposite one another, the phase masks 5 being worn by one and / or the other of the two reflecting optical parts and forming a multi-passage cavity. Here, the phase masks 5 are all carried by a microstructured phase plate 2a, the second optical part 2b being formed of a simple mirror.
    In this MPLC device 1, in accordance with the present description, there is a first transverse plane PI to the light radiation in which this radiation can be described by the first family of modes with separable variables (p, q). And there is a second transverse plane P2 in which the desired radiation can be described by a second family of modes with variable (p, q) separable.
    As we have seen, the phase shifts cpi (p, q) of the M phase masks have been established so that the determined transformation puts N modes of index {i, j} k, l <= k <= in relation, respectively. N, of the first family with N modes of the same index {i, j} k, l <= k <= N of the second family. The modes of the second family are not Hermite-Gauss modes and, simultaneously, the modes of the first family are not arranged in the first transverse plane in the shape of a triangle.
    In a particular case of design of such a device 1, at least one of the optical parts of the conversion block 2 can wear a phase mask with separable variables. This part can correspond to a phase plate, to a deformable mirror or to a spatial light modulator.
    Another example of an MPLC device according to the invention is shown in FIG. 6a. We find the conversion block 2, comprising the multi-passage cavity as described in the previous example. At least one stage is also provided for injecting the first or second radiation into the conversion block 2 and / or extracting the first and / or second radiation from the conversion block 2.
    In this new example, the MPLC device more specifically comprises a first stage 3 for in particular injecting into the conversion block 2 the incident light radiation and a second stage 4 for in particular extracting the transformed radiation from the conversion block 2.
    However, it should be noted that, in general, these stages 3, 4 could have the opposite function as that described here, or that the same stage 3, 4 can both inject and extract incident radiation from the conversion block 2 and transformed. The first stage 3 may comprise an optical fiber or a bundle of optical fibers 3a arranged linearly or not in a first transverse stage plane. Similarly, the output stage can comprise an optical fiber 4a or a bundle of optical fibers freely arranged in a second transverse stage plane. There could also be less than 2 floors or more than 2 floors.
    In the advantageous configuration of FIG. 6a, the first stage 3 consists of a bundle of optical fibers 3a arranged in a first transverse plane of stage Pe. The first transverse stage plane constitutes the first transverse plane of the MPLC device, that is to say that in the transverse stage plane Pe, the incident radiation is expressed as a mode family with separable spatial variables, ie the first family of modes.
    Thus, the optical fibers 3a of the bundle can be arranged in the form of a regular rectangular or square matrix as illustrated on the transverse floor plan of FIG. 6b so that the spatial variables of the first family of modes are separable in a Cartesian coordinate system.
    The first stage 3 can group into a bundle 81 single-mode optical fibers 3a arranged and assembled in a square form composed of 9 columns and 9 lines. The free ends of the fibers 3a are provided with a collimation optic, as is well known per se, in order to gaussian shape the light radiation coming from each fiber. The incident light radiation which is provided by stage 3 does indeed constitute a family of modes with separable variables. Each mode is formed by the Gaussian radiation extracted from one of the optical fibers of the bundle, which can be designated by a pair of indices {i, j} or i and j respectively denote the row and the column of the fiber in question in the beam.
    The second stage 4 of the example reproduced in FIG. 6a comprises a multimode fiber 4a whose guided modes are Hermite-Gauss modes. These modes, as we have seen, form a family of modes with separable spatial variables when they are designated by their usual pairs of indices {i, j}. The second transverse stage plane defined by this second stage 4 constitutes the second transverse plane of the MPLC device 1 in which the desired radiation is expressed by the second family of modes with separable spatial variables.
    The multiplanar device 1 of this example aims to transform the radiation coming from the 81 fibers 3a of the first stage 3 in order to couple them to the multimode fiber 4a of the second stage 4 and form a spatial multiplexer. We note that this device 1 can completely be used in the opposite direction, and spatially demultiplex an incident radiation coming from the multimode fiber 4a and couple the modes of this fiber to the optical fibers 3a of the first stage
    3.
    In the particular example of FIG. 6a, the phase plate 2a carries 11 phase masks 5 designed using a method according to the invention. That is to say that a digital model of the device 1 which has just been presented is used, by simulation, to determine the characteristics of the 11 phase 5 masks worn by the phase blade 2a in order to best adjust the transformation of each mode of index {i, j} of the first stage 3 to the corresponding mode of the same index {i, j} of the family of modes Hermite-Gauss of the second stage.
    In a variant of the example taken in illustration, the first stage 3 may consist of a bundle of optical fibers 3a arranged, in the first transverse plane of stage Pe, along concentric circles and along radial lines, as shown in Figure 6c. This distribution injects light radiation which can be approximated by a first family of modes with separable spatial variables in a system of polar coordinates.
    To gain precision, one or a plurality of complementary phase masks 5 can be provided in the transformation block 2 making it possible to form, on a first transverse plane Pi of the device 1, radiation which is as perfectly as possible with separable variables.
    Alternatively, one can tolerate imprecision of implementation, that is to say design the phase masks 5 from a design method according to the invention for which, in the first floor plan Pe , the incident light radiation will be described from the first family of perfectly separable variable modes. The MPLC device will be produced from the digital models of the phase masks 5 obtained at the end of this design process, and from a first stage in which the arrangement of the optical fibers in the floor plan is only approximated. by a family of modes with separable polar variables.
    Finally, we have shown that a transformation carried out by an MPLC device matches the modes of a first family of modes with those of a second family of modes, the two families being with separable spatial variables and the modes used correspondence with the same index, could be implemented by a reduced number of phase masks.
    The constraints posed on the two families of modes and on the correspondence between the modes of these families, lead to posing an optimization problem whose solutions form a class of transformations simple to carry out.
    Of course, the invention is not limited to the modes of implementation described and it is possible to add variants without departing from the scope of the invention as defined by the claims.
    权利要求:
    Claims (20)
    [1" id="c-fr-0001]
    1. Method for designing a device for multiplan converting light radiation, the device implementing a plurality of phase masks intercepting the light radiation in order to respectively apply to it a spatio-frequency phase shift leading to a determined transformation, the design process including the following steps:
    - Define a first family of modes (u), each mode (uij) of the first family being designated by a pair of indices {i, j} l <= i <= n, l <= j <= m, the first family (uij (p, q)) being with separable spatial variables (p, q);
    - Define a second family of modes (v), and each mode of the second family (vij) being designated by a pair of indices {i, j}, l <= i <= n, l <= j <= m , the second family (vij (p, q)) being with separable spatial variables (p, q);
    - Form a first and, respectively, a second family of modes used by selecting N torque modes of index {i, j} kl <= k <= N, N being less than or equal to nm and strictly greater than n and m, in the first family and, respectively, in the second family, the N modes of the second family not being Hermite-Gauss modes and, simultaneously, the N modes of the first family not being arranged in a triangle-shaped plane;
    - Establish the spatio-frequency phase shift of the phase masks making it possible to transform each mode of torque of index {i, j} k of the first family of modes used in the mode of same torque of index {i, j} k of the second family of operating modes.
    [2" id="c-fr-0002]
    2. Design method according to the preceding claim wherein the first family of modes (u) and the second family of modes (v) are bases.
    [3" id="c-fr-0003]
    3. Design method according to the preceding claim wherein the bases of the first family of modes (u) and the second family of modes (v) are orthonormal.
    [4" id="c-fr-0004]
    4. Design method according to one of the preceding claims, in which the separable variables (p, q) are Cartesian coordinates (x, y).
    [5" id="c-fr-0005]
    5. Design method according to one of claims 1 to 3 wherein the separable variables (p, q) are polar coordinates (r, theta).
    [6" id="c-fr-0006]
    6. Design method according to one of the preceding claims, in which the establishment of the spatio-frequency phase shift of the M phase masks is carried out by optimizing a degree of correspondence between the transformation of the modes of the first family of exploited modes and the modes of the second family of operating modes.
    [7" id="c-fr-0007]
    7. Design method according to the preceding claim wherein the degree of correspondence is optimized by a stochastic optimization method.
    [8" id="c-fr-0008]
    8. Design method according to claim 6, in which the degree of correspondence is optimized by matching of wave fronts.
    [9" id="c-fr-0009]
    9. Design method according to one of the preceding claims, in which the step aimed at establishing the spatio-frequency phase shift of comprises:
    - Establish respectively one-dimensional according to a phase mask a phase shift first spatial variable (p) and according to a second spatial variable (q);
    - Combine the one-dimensional phase shifts according to the first spatial variable (p) and according to the second spatial variable (q) to establish the spatio-frequency phase shift of the phase masks.
    [10" id="c-fr-0010]
    10. Computer program comprising instructions adapted to the implementation of at least one of the steps of the design method according to one of claims 1 to 9, when the program is executed on a computer.
    [11" id="c-fr-0011]
    11. Phase blade manufactured from the space-frequency phase shift of at least one phase mask obtained at the end of a design process according to one of claims 1 to 9.
    [12" id="c-fr-0012]
    12. Multiplan conversion device (1) of a first light radiation, the device comprising a conversion block (2) comprising a plurality of optical parts (2a, 2b) and implementing a plurality of phase masks (5) carried by at least one of the optical parts (2a, 2b), the phase masks (5) applying a spatio-frequency phase shift aiming to develop a second radiation, the determined transformation respectively relating N index modes {i, j } k, l <= k <= N, of a first family of operating modes with separable spatial variables (p, q) describing the first light radiation at a first transverse plane with N modes of the same index {i, j} κ, l <= k <= N, of a second family of modes used with separable spatial variables describing the second light radiation at a second transverse plane, the N modes of the second family not being Hermite-Gauss modes and simultaneously the s N modes of the first family not being arranged in the first transverse plane in the shape of a triangle.
    [13" id="c-fr-0013]
    13. Multiplan conversion device (1) according to the preceding claim comprising at least one stage (3,4) for injecting the first and / or the second radiation into the conversion block (2) and / or extracting the first and / or the second radiation from the conversion block (2).
    [14" id="c-fr-0014]
    14. Multiplan conversion device (1) according to the preceding claim in which the stage (3,4) comprises a bundle of optical fibers (3a) arranged in a transverse stage plane (Pe), the first transverse plane being the transverse floor plan.
    [15" id="c-fr-0015]
    15. Multiplan conversion device (1) according to the preceding claim wherein the spatial variables of the first family of exploited modes are separable in a Cartesian coordinate system and the optical fiber bundle (3a) of the stage (3,4 ) is arranged in the transverse floor plan (Pe) in the shape of a rectangle or square.
    [16" id="c-fr-0016]
    16. Multiplan conversion device (1) according to claim 14 in which the spatial variables (p, q) of the first family of exploited modes can be separated in a polar coordinate system (r, theta) and the optical fiber bundle ( 3a) of the stage (3) is arranged in the transverse stage plane (Pe) along concentric circles and along radial lines.
    [17" id="c-fr-0017]
    17. Multiplan conversion device (1) according to one of claims 13 to 16 in which the stage (4) comprises at least one multimode fiber (4a).
    [18" id="c-fr-0018]
    18. Multiplan conversion device (1) according to one of claims 12 to 17 in which at least one of the optical parts (2a, 2b) carries at least one phase mask (5) with separable spatial variables (p, q) .
    [19" id="c-fr-0019]
    19. Multiplan conversion device (1) according to one of claims 12 to 18 wherein the optical parts (2a, 2b) comprise two reflective optical parts (2a, 2b) arranged opposite one of the 'other, the phase masks (5) being carried by one and / or the other of the two reflecting optical parts (2a, 2b).
    [20" id="c-fr-0020]
    20. Multiplan conversion device (1) according to the preceding claim wherein the two optical parts (2a, 2b) consist of a microstructured phase plate (2a) and a mirror (2b), and the phase masks (5) are carried by the microstructured phase plate (2a).
    类似技术:
    公开号 | 公开日 | 专利标题
    EP3100095B1|2021-03-10|Device for processing light/optical radiation, method and system for designing such a device
    FR3078412A1|2019-08-30|METHOD FOR DESIGNING A MULTIPLAN CONVERSION DEVICE, PHASE BLADE OBTAINED THROUGH THIS METHOD, AND MULTIPLAN CONVERSION DEVICE
    EP3430736B1|2020-01-01|Multimode optical fiber communication device comprising a component for modal dispersion compensation
    FR3049719B1|2019-09-13|DEVICES AND METHODS FOR TRANSPORTING AND CONTROLLING LUMINOUS BEAMS FOR ENDO-MICROSCOPIC IMAGING WITHOUT LENS
    FR2676825A1|1992-11-27|OPTIMIZED WAVEGUIDE FOR LOW-LOSS COUPLING WITH OPTICAL FIBERS AND METHOD FOR DETERMINING THE SAME.
    Ruffato et al.2019|Non-paraxial design and fabrication of a compact OAM sorter in the telecom infrared
    EP0877266A1|1998-11-11|Selective optical device and wavelenght multiplexer-demultiplexer
    EP3149813A1|2017-04-05|System for phasing a large number of laser sources
    FR3074587A1|2019-06-07|PHOTONIC CHIP WITH OPTICAL PATH FOLDING AND INTEGRATED COLLIMATION STRUCTURE
    EP2614561B1|2017-11-01|Method and device for amplifying an optical signal
    WO2018134532A1|2018-07-26|Device for controlling the position of a focussing point of a light beam
    EP3491438B1|2020-10-14|Multi-spectral optical coupler with low receive losses
    EP3732531B1|2021-11-03|Optical device possessing means for the precise assembly thereof, assembly or test method for said device
    FR2976663A1|2012-12-21|DEVICE AND METHOD FOR CHARACTERIZING A LIGHT BEAM.
    FR3077889A1|2019-08-16|OPTICAL DEVICE FOR COMPENSATING IMPERFECTIONS OF OPTICAL COMPONENTS COMPRISING THE DEVICE AND METHOD FOR MANUFACTURING SUCH A DEVICE
    CA2160217C|2000-08-22|Wavelength selective optical coupler
    FR2860077A1|2005-03-25|SPATIAL PHASE FILTER FOR OPTICAL BEAM, SYSTEM AND METHOD THEREOF
    WO2017158262A1|2017-09-21|Multimode optical fiber communication device featuring interversion of the propagation modes
    Hu et al.2020|Quantitative phase imaging based on wavefront correction of a digital micromirror device
    Burnham et al.2011|Effects of aliasing on the fidelity of a two dimensional array of foci generated with a kinoform
    WO2020161126A1|2020-08-13|Optical device combining a plurality of beams and system using same
    FR3067457A1|2018-12-14|METHOD AND SYSTEM FOR CONTROLLING THE SPEED OF A LASER PULSE
    Rothe et al.2021|Benchmarking analysis of computer generated holograms for complex wavefront shaping using pixelated phase modulators
    WO2020254774A1|2020-12-24|Light source comprising at least one semiconductor chip bearing at least one diode
    FR3084480A1|2020-01-31|Light radiation processing device comprising a multiplanar light conversion device
    同族专利:
    公开号 | 公开日
    FR3078412B1|2022-01-14|
    US20200409141A1|2020-12-31|
    CN111771147A|2020-10-13|
    WO2019162637A1|2019-08-29|
    EP3759533A1|2021-01-06|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    EP2469221A1|2010-12-21|2012-06-27|Universite Pierre Et Marie Curie - Paris 6|Method and system for configuring a device for correcting the effect of a medium on a light signal, method, device and system for correcting said effect.|
    FR3016973B1|2014-01-30|2017-08-11|Cailabs|DEVICE FOR PROCESSING LIGHT / OPTICAL RADIATION, METHOD AND SYSTEM FOR DESIGNING SUCH A DEVICE|US11108469B2|2019-06-19|2021-08-31|University Of Central Florida Research Foundation, Inc.|Mode demultiplexing hybrid apparatus, methods, and applications|
    FR3104271B1|2019-12-04|2021-12-17|Alpao|ADAPTIVE OPTICAL SYSTEM WITH IMPROVED RESPONSE TIME, RELATED USE AND PROCESS|
    法律状态:
    2019-02-25| PLFP| Fee payment|Year of fee payment: 2 |
    2019-08-30| PLSC| Publication of the preliminary search report|Effective date: 20190830 |
    2020-02-17| PLFP| Fee payment|Year of fee payment: 3 |
    2021-02-11| PLFP| Fee payment|Year of fee payment: 4 |
    2022-02-15| PLFP| Fee payment|Year of fee payment: 5 |
    优先权:
    申请号 | 申请日 | 专利标题
    FR1851664A|FR3078412B1|2018-02-26|2018-02-26|METHOD FOR DESIGNING A MULTIPLANE CONVERSION DEVICE, PHASE PLATE OBTAINED THROUGH THIS METHOD, AND MULTIPLANE CONVERSION DEVICE|
    FR1851664|2018-02-26|FR1851664A| FR3078412B1|2018-02-26|2018-02-26|METHOD FOR DESIGNING A MULTIPLANE CONVERSION DEVICE, PHASE PLATE OBTAINED THROUGH THIS METHOD, AND MULTIPLANE CONVERSION DEVICE|
    PCT/FR2019/050433| WO2019162637A1|2018-02-26|2019-02-26|Method for designing a multi-plane conversion device, phase plate obtained by this method, and multi-plane conversion device|
    CN201980014723.8A| CN111771147A|2018-02-26|2019-02-26|Method for designing a multiplanar transformator, phase plate obtained by this method and multiplanar transformator|
    EP19711994.4A| EP3759533A1|2018-02-26|2019-02-26|Method for designing a multi-plane conversion device, phase plate obtained by this method, and multi-plane conversion device|
    US16/975,673| US20200409141A1|2018-02-26|2019-02-26|Method for designing a multi-plane conversion device, phase plate obtained by this method, and multi-plane conversion device|
    [返回顶部]