Precision alignment between optical arrays and steering arrays

Abstract

A system for aligning optical signals for optical networking includes a light station, a collimator element corresponding to the light station, a steering array including a steering element corresponding to the light station, and a set of optical transfer elements. The set of optical transfer elements is positioned between the steering array and the light station. The set of optical transfer elements forms a light image from a first plane positioned at a first location onto a second plane positioned at a second location.

Description

PRECISION ALIGNMENT BETWEEN OPTICAL ARRAYS AND STEERING ARRAYSCROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of U.S. Provisional Application No. 63 / 643,896 filed May 7, 2024 (Attorney Docket No. 52753-19) and U.S. Provisional Application No. 63 / 711,672 filed October 24, 2024 (Attorney Docket No. 52753-21). The entire disclosures of the above applications are incorporated by reference.FIELD

[0002] The present disclosure relates to optical connections and more particularly to optical circuit switches.BACKGROUND

[0003] Modern data centers rely on circuit switching to create a network between network nodes via a dedicated communications channel (a circuit). The reliability and speed of such networks are highly dependent on the circuit switches which route and form the circuit. In recent years, traditional copper cabling has given way to fiber optic cables which generally provide greater capacity with less signal loss and interference. However, fiber optic circuit switches require specialized and complex systems to maintain accuracy and reliability.

[0004] Insertion loss, which measures optical power transfer efficiency of the switch (optical power out divided by optical power in) is a primary parameter of an OCS. Any misalignment that results in lost light increases the insertion loss and degrades the performance of the OCS.

[0005] A common optical circuit switch (OCS) architecture uses an array of input ports — that is, light sources, such as fiber, lasers, and / or light emitting diodes (LEDs) — aligned via a collimator array to steer light to an input array of steering elements, then to a second output array of steering elements, and finally to an output collimator array and output port array — that is, light detectors, such as fiber. Some architectures include over 100 inputs and over 100 outputs. This design allows for flexibility in switching light from input to output ports. However, OCS architecture poses significant stability and reliability challenges with regards to maintaining alignment of the steered light because the light must be precisely aligned across free space gaps on the scale of 0.1-1 pm to focus into the fiber (or other light detector), which has a core of around 10pm. This alignment must be maintained while the OCS architecture is distorting due to humidity, air pressure, vibration, creep, hysteresis, and / or thermal variation in the surrounding environment. Variations on the scale of a single micron (1pm) are sufficient to increase insertion loss throughout the switch. In some implementations, other non-fiber optical components like chiplets are used as light sources and detector ports instead of fiber ports. In some implementations, chiplets include microLEDs, micro-lasers, and / or micro-photodetectors. These chiplets require similar precision with a single micron of misalignment resulting in loss.

[0006] The OCS must efficiently route the light from an input port to an output port with minimal loss, minimal cross-talk, and be able to change the connection in minimal time. A typical OCS uses an array of light sources to launch signals through an array of small lenses, where each lens is located at the end of each fiber (or light source). The output from a steering array in an OCS is generally desired to be a parallel set of beamlets aligned to the primary optical axis of the architecture, where each beamlet may be collimated or diverging depending on the optical setup. The lenses are intended to collimate the light from each fiber (or other light source) and focus it onto a specific steering element.

[0007] A major challenge in OCS design is ensuring the collimator array and light source array fabrication are tightly controlled both during manufacturing and during operation to ensure proper emission angle on the collimated beam. Any error in the beam angle results in error displacement of the beam on the surface of the steering element. Error displacement causes an increase of insertion loss and can cause cross-talk with other channels (i.e., data from an input is detected at an output other than the intended output).

[0008] The high precision required for efficient OCS operation stresses fabrication tolerances which results in decreased yield, increased costs, and acts as a point of sensitivity to small structural distortions (e.g., thermal and / or mechanical) during operation. Even if perfectly calibrated after initial assembly, the alignment will change during the varying mechanical and thermal loads on the OCS, resulting in cross talk and insertion loss. Sensitivity during operation poses a major challenge to OCS designs as compensation generally requires either extreme materials, extreme manufacturing tolerances or active elements in the design that add cost, complexity, and power draw.

[0009] There is increasing demand for OCS performance improvements in the domain of insertion loss. Any misalignment that results in lost light increases insertion loss and increased insertion loss requires additional power to ensure the signal reaches its endpoint. Reduced insertion loss allows a larger quantity of OCSs to be chained together, creating a more broadly linked data center.

[0010] The background description provided here is for the purpose of generally presenting the context of the disclosure. Work of the presently named inventors, to the extent it is described in this background section, as well as aspects of the description that may not otherwise qualify as prior art at the time of filing, are neither expressly nor impliedly admitted as prior art against the present disclosure.SUMMARY

[0011] A system for directing optical signals for optical networking includes a light source, a collimator element corresponding to the light source, a steering array including a steering element corresponding to the light source, and a set of optical transfer elements. The set of optical transfer elements is positioned between the steering array and the light source. The set of optical transfer elements forms a light image from a first plane positioned at a first location onto a second plane positioned at a second location. The first location coincides with the collimatorelement and the second location coincides with the steering array. The set of optical transfer elements is defined by at least one of a first characteristic or a second characteristic. The first characteristic is that an optical signal from the light source passes through each element of the set of optical transfer elements at most once. The second characteristic is that the set of optical transfer elements is a single element that forms the light image from the first plane onto the second plane.

[0012] In other features, the set of optical transfer elements directs light from the light source to the steering element. In other features, a distance between the single element and the light source is based on a focal distance of the single element. In other features, a focal distance of the collimator element and focal distances of the set of optical transfer elements are equal. In other features, the set of optical transfer elements includes a single element that forms the light image from the first plane onto the second plane.

[0013] In other features, the set of optical transfer elements includes at least two elements that form the light image from the first plane onto the second plane. In other features, a focal distance of the at least two elements is equal. In other features, a distance between the at least two elements and the light source is based on one or more focal distances associated with the at least two elements, and a distance between the at least two elements is based on one or more focal distances associated with the at least two elements.

[0014] In other features, a center axis of the collimator element is offset from a center axis of the light source. In other features, an offset distance between the center axis of the collimator element and the center axis of the light source is based on a distance of the light source from a center axis of a light source array that includes the light source.

[0015] In other features, a center axis of the light source is angularly offset from parallel with respect to a center axis of the collimator element. In other features, an angle offset between the center axis of the light source and a center axis of the collimator element is based on a distance of the light source from a center axis of a light source array that includes the light source.

[0016] In other features, each element of the set of optical transfer elements has a fixed position. In other features, the system includes an optical element that is physically distinct from the set of optical transfer elements. In other features, the optical element is a Fourier lens element. In other features, the system includes a second steering array.

[0017] In other features, the system includes a light detector, a second collimator element corresponding to the light detector, and a second set of optical transfer elements. In other features, the second set of optical transfer elements is positioned between the second steering array and the light detector. In other features, the second set of optical transfer elements forms a second light image from a third plane positioned at a third location onto a fourth plane positioned at a fourth location. In other features, the third location coincides with the second steering array. In other features, the fourth location coincides with the second collimator element. In other features, the second set of optical transfer elements is defined by at least one of a third characteristic or a fourth characteristic. The third characteristic is that the optical signal from the second steering array passes through each element of the second set of optical transferelements at most once. The fourth characteristic is that the second set of optical transfer elements is a single element that forms the second light image from the third plane onto the fourth plane.

[0018] In other features, the second set of optical transfer elements is defined by the third characteristic. In other features, the second set of optical transfer elements is defined by the fourth characteristic. In other features, the second set of optical transfer elements directs light from a second steering element of the second steering array to the light detector. In other features, the set of optical transfer elements is defined by the first characteristic. The set of optical transfer elements is defined by the second characteristic. In other features, the single element is a lens. In other features, the single element is a curved reflector.

[0019] An optical networking device includes a system for directing optical signals for optical networking. The system includes a light source, a collimator element corresponding to the light source, a steering array including a steering element corresponding to the light source, and a set of optical transfer elements. The set of optical transfer elements is positioned between the steering array and the light source. The set of optical transfer elements forms a light image from a first plane positioned at a first location onto a second plane positioned at a second location. The first location coincides with the collimator element and the second location coincides with the steering array. The set of optical transfer elements is defined by at least one of a first characteristic or a second characteristic. The first characteristic is that an optical signal from the light source passes through each element of the set of optical transfer elements at most once. The second characteristic is that the set of optical transfer elements is a single element that forms the light image from the first plane onto the second plane.

[0020] A method for directing optical signals for optical networking includes shining an optical signal onto a set of optical transfer elements. A collimator element directs the optical signal from a light source. The set of optical transfer elements is positioned between a steering array and the light source. The set of optical transfer elements forms a light image from a first plane positioned at a first location onto a second plane positioned at a second location. The first location coincides with the collimator element. The second location coincides with the steering array. The set of optical transfer elements is defined by at least one of a first characteristic or a second characteristic. The first characteristic is that an optical signal from the light source passes through each element of the set of optical transfer elements at most once. The second characteristic is that the set of optical transfer elements is a single element that forms the light image from the first plane onto the second plane.

[0021] In other features, the set of optical transfer elements is defined by the first characteristic. In other features, the set of optical transfer elements is defined by the second characteristic.

[0022] Further areas of applicability of the present disclosure will become apparent from the detailed description, the claims, and the drawings. The detailed description and specific examples are intended for purposes of illustration only and are not intended to limit the scope of the disclosure.BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The present disclosure will become more fully understood from the detailed description and the accompanying drawings.

[0024] FIG. 1 is an example of a common optical circuit switch (OCS) architecture with two steering arrays.

[0025] FIG. 2 is an example of positional changes in OCS architecture.

[0026] FIG. 3 is an example of parameter changes in OCS architecture.

[0027] FIGS. 4A-4B are an example of updating alignment in OCS steering arrays with out- of-channel sensing elements.

[0028] FIG. 5 is an example steering array with out-of-channel sensing elements.

[0029] FIG. 6 is a block diagram of an example system for precision out-of-channel sensing.

[0030] FIGS. 7A-7B are flowcharts of an example method for aligning out-of-channel sensing elements.

[0031] FIG. 8 is a flowchart of an example method for updating calibration of steering elements.

[0032] FIG. 9 is a flowchart of an example method for aligning an optical input port with an optical output port.

[0033] FIG. 10 is an example of fiber misalignment in an OCS architecture.

[0034] FIGS. 11A-11B are examples of OCS light sources and collimator arrays.

[0035] FIG. 12 is a block diagram of an example system for precision alignment in an optical circuit switch.

[0036] FIG. 13 is an example of an OCS with image transfer elements.

[0037] FIG. 14A is an example of a light source collimator for use in an OCS with image transfer elements.

[0038] FIG. 14B is a close-up view of a region of the example light source collimator of FIG. 14A.

[0039] FIG. 15 is an example of an OCS with an image transfer element.

[0040] FIG. 16 is an example of an OCS with image transfer elements.

[0041] FIG. 17 is an example of an OCS with image transfer elements.

[0042] FIG. 18 is an example of an OCS with image transfer elements.

[0043] In the drawings, reference numbers may be reused to identify similar and / or identical elements.DETAILED DESCRIPTIONOUT-OF-CHANNEL ALIGNMENT FEEDBACK INTRODUCTION

[0044] The present disclosure describes a system for achieving high stability, low optical loss, and actively monitored alignment between ports in an optical circuit switch (OCS). The present disclosure avoids drawbacks of active alignment schemes including the cost, complexity, andpower losses associated with sensing on each port. While the disclosure focuses on a particular input and output port architecture, the same approach can be applied to other OCS architectures. Similarly, while optical fibers (including fiber arrays used as input and output sources) are frequently described throughout the disclosure, the approaches, structures, methods, and systems described below are equally applicable to systems that use alternatives to fiber, including various other light sources, light detectors, and light transmission media (such as chiplets, microLEDs, micro-lasers, and / or micro-photodetectors).

[0045] In various embodiments, an input and / or output port includes a single optical data source. In various embodiments, an input and / or output port includes two or more optical data sources. For example, an input and / or output port may include a bundle of optical sources (such as fiber, lasers, etc.) pointed at a single steering element, which results in more optical input and / or output sources than steering array elements. In various embodiments, each source of the bundle of optical sources corresponds to a collimator unique to the source.

[0046] In various embodiments, there are an equal number of input ports and output ports. In various embodiments, the number of input ports and output ports and not equal. In various embodiments, data ports act as both optical sources and detectors. For example, a data port can transmit an optical signal and receive another optical signal simultaneously.

[0047] FIG. 1 is an example of a common OCS architecture including input port array 104, input steering array 108, output steering array 112, and output port array 116. Input port array 104 includes input port 104-1, input port 104-2, input port 104-3, and input port 104-4. In some implementations, input port array 104 includes more than four input ports (e.g., 10, 100, 500, or 1000 ports). Output port array 116 includes output port 116-1, output port 116-2, output port 116-3, and output port 116-4. In some implementations, output port array 116 includes a more than four output ports (e.g., 10, 100, 500, or 1000 ports). In this example, a data signal (such as an optical signal) is transmitted (via laser) from an input port (such as input port 104-2) on input port array 104. The signal is then reflected by a steering element in input steering array 108 into a steering element on output steering array 112. Finally, the signal is reflected into an output port (such as output port 116-1) on output port array 116.SHORTCOMINGS OF EXISTING OCS ARCHITECTURES

[0048] Electrostatic micro-electro-mechanical systems (MEMS) are a common solution for steering elements in OCS architectures. A voltage command (or multiple voltage command vector) is supplied to an electrostatic MEMS, which generates specific motion responses. However, the response of a MEMS may change over time as the performance parameters change. For example, an electrostatically actuated MEMS may be sensitive to: i) electrostatic charge buildup producing non-zero actuation forces, ii) Young’s Modulus variation with temperature changing the force-to-position mapping, iii) equilibrium orientation variation due to stress balance variation with temperature, iv) equilibrium orientation variation due to hysteresis, v) equilibrium orientation variation due to creep, and vi) optical coupling effects includinglocalized heating. All of these effects alter the electrical-to-mechanical transformation for a steering element and create issues for port-to-port alignment repeatability in OCS architectures.

[0049] A common 2D array OCS architecture (as seen in FIG. 1) requires a four degrees of freedom (4DOF) command vector to be specified for each port-to-port mapping. The 4DOF command includes two specific 2D commands (such as tip and tilt) for both the input and output steering arrays. Finding the port-to-port command vector relies on the stability of the OCS and architecture level calibration done on the whole OCS after assembly, when the input and output steering arrays can each be scanned over their range of motion to find the specific ports. A unique 4D command vector can then be extracted as the vector which maximizes the output port signal for a given input port (m) and output port (n). A lookup table can then be created for all possible m and n combinations, which is used during regular operation. The created 4D command vector will always be the optimal solution if the OCS architecture is completely stable (i.e., without performance variation over time). However, OCS architectures are not completely stable and variation is commonplace in OCS systems. Therefore, the true 4D command vector for any given input and output port combination will change over time.

[0050] Other approaches rely on in-operation feedback to adjust (in real-time during operation) the 4D command vector to achieve a desired port-to-port alignment via an in-channel alignment signal (i.e., within the channel of a data signal beam). An in-channel signal shares the same physical optical path as data signals rather than a spectral channel. In some implementations, an in-channel signal differs in frequency from the data signal. In-channel alignment requires in- channel sensing for the alignment signal to be detected. In-channel signal sensing includes two primary approaches: i) tapping, wherein a fraction of the data signal is read from the output fiber for feedback purposes, and ii) as a separate signal injected into the signal early in the OCS architecture, then extracted from the signal at a later part of the architecture.

[0051] In-channel sensing via tapping into the output fiber signal provides full end-to-end measurement, allowing all possible errors in the system to be captured by the alignment measurement. A full measurement approach such as tapping provides high precision, but often results in power loss for the primary data signal due to the tap on the output fiber signal. It also requires additional equipment for each tap that can significantly raise the cost and complexity of the OCS.

[0052] In-channel sensing via injecting a separate frequency beam and extracting that within the OCS architecture provides partial measurement of the architecture errors. Errors in the architecture before the alignment signal injection location and after the extraction location are not captured by the alignment measurement. The partial measurement approach results in uncaptured error, (which may cause larger port-to-port alignment error than full measurement approaches) but will not significantly attenuate the primary data signal.

[0053] Both in-channel approaches result in unwanted optical losses through the OCS. In- channel active feedback approaches, whether partial or full measurement, resolve the issue of tracking the time-changing optimal 4D command vector. This lowers the scale and variability of link insertion loss. In-channel active feedback brings significant feedback hardware complexity,cost, and power use, as the feedback must be implemented in-channel within each channel in the array, thus the hardware requirements scale with the array size.

[0054] Out-of-channel approaches have been proposed, however these designs also have significant drawbacks. For example, in some implementations, a separate control beam is reflected off the steering elements along a different path from the data signal beam such that it can be steered to a position sensitive detector (PSD). The reading on the PSD can then be used to monitor changes in the steering elements. This approach provides only partial measurement of the alignment errors in the OCS architecture thus resulting in significant potential alignment error.

[0055] Another approach includes a combination of in-channel and out-of-channel sensing. However, existing combination methods fail to provide the accuracy that would enable high precision measurement, are unable to account for significant unmeasured alignment error added by the control feedback structures, and / or are unable to account for the time variation of the steering elements inherent in the design - resulting in improper alignment. Thus, there is a need for an OCS port-to-port alignment approach which provides the benefit of the in-channel high accuracy active alignment approach without the drawback of other in- and out-of-channel approaches.PRECISION OUT-OF-CHANNEL CONTINUOUS ALIGNMENT PROTOCOL

[0056] Precision Out-of-Channel Continuous Alignment Protocol (POCCAP), as various implementations of the present disclosure are named, achieves active feedback alignment for each port without using in-channel sensing while overcoming the drawbacks of current out-of- channel solutions. POCCAP provides the benefit of the active alignment in-channel sensing approach without the drawbacks of typical in-channel sensing solutions including: i) additional optical losses, ii) injection of unmeasured alignment errors, and iii) additional hardware scaling at the per-element level. POCCAP uses out-of-channel sensing and thus avoids additional optical loss and additional hardware scaling at the per-element level, unlike prior in-channel sensing designs. The use of continuously calibrated and time stable elements (specifically the steering elements) resolves the issue of time stability found in prior out-of-channel sensing approaches and thus allows for commands on un-sensed elements to be inferred from neighboring sensed elements. This is a key component for achieving the benefit of in-channel sensing without the drawbacks of in-channel sensing. The use of alignment sensing ports physically linked to primary optical elements avoids the issues with prior out-of-channel calibration and feedback sensing approaches which inject additional alignment error to the measurements.

[0057] POCCAP provides precision full measurement feedback alignment, which ensures low- loss port-to-port links. POCCAP generally relies on four features in the OCS architecture: i) continuously calibrated elements, ii) time stable element performance for all primary elements, iii) dedicated out-of-channel alignment sensing ports physically linked to primary optical elements to measure array level changes in the OCS architecture, and iv) a way todetermine the relative 4D command vector for a given desired data port-to-port alignment. In operation, the POCCAP controller combines the relative 4D command vectors with the corrections generated by the out-of-channel alignment measurements. POCCAP enables a system whose port-to-port alignment for data channels can be tracked and actively maintained in operation by dedicated sense channels without impeding the operation of the data channels.CONTINUOUSLY CALIBRATED ARRAY

[0058] The first feature of the POCCAP approach is a continuously calibrated array. The standard steering array used in an OCS is not a continuously calibrated array. In standard OCS architecture, an uncalibrated array is controlled with a command electrical signal (such as a voltage) which is translated through the steering element electromechanical response to produce a steering position such as a tip / tilt location. Uncalibrated operation can be used when treating the system as a black box for finding the commands to achieve a specific orientation, however each achieved steering position must have been previously mapped. There is no algorithm capturing the transformation from electrical command to achieved steering position, so adjustments of the steering position into unmapped regions cannot be achieved with high accuracy. Therefore, the system cannot accurately implement small changes in steering position around the previously mapped points, instead the system can only accurately shift between premapped locations. When the system does not change with time, the ideal alignments can be premapped and uncalibrated operation is possible. When the system changes with time, uncalibrated operation will result in errors as the required alignment will deviate from the premapped alignment.

[0059] The proposed solution employs a continuously calibrated steering array which includes an element calibration transform between the user supplied calibrated input and the electrical command into the steering element. The steering element is functionally a transformation from the electrical domain (the command) to the mechanical domain (the physical position), and the calibration is the mathematical inverse transformation. The calibration modifies user input so that the supplied user input maps to the desired steering position (preferably via a 1 : 1 mapping where both input and output are of the same units). The calibrated input is in the form and domain of a steering position command, while the uncalibrated input is in a different form and domain of electrical commands. Calibrated inputs directly represent steering position outputs, meaning that command vectors for calibrated elements are position command vectors.

[0060] Calibrated operation enables the system to accurately implement small changes in steering position around the previously mapped points. Properly calibrated elements show an equivalence in relative response such that a given position command vector change produces the same steering element positional change for all the calibrated elements. This equivalence provides the performance effect relied on by POCCAP, enabling position command changes observed on one channel to be accurately transferred to other unobserved channels. The change in the position command vector to a calibrated element can accurately be read as the change inthe actual element position, and accurately transferred to other elements. This can only be achieved with systems that are continuously calibrated.

[0061] Steering elements are considered calibrated when the mapping of the given position command vector to an equivalent elemental position change is accurate at the time of consideration. One time calibration states create a temporarily calibrated steering element, however steering element electrical-to-mechanical performance parameters change over time so the steering element will gradually become uncalibrated, as described above. The combination of charge buildup, material property variation due to optical heating, environmental change, hysteresis, and creep alter the electrical-to-mechanical transformation for a steering element. If these effects are not tracked and compensated, then the steering elements will quickly lose their calibration.

[0062] One method of mitigating calibration loss uses devices with extensive design symmetries to cancel at least most out the changes, but this will not resolve all changes. A second, more robust way of compensating for the effects is with active feedback. Capacitivebased positional feedback, as one example, provides a solution to ensuring continuously calibrated elements as capacitance measurements operate by observing charge flow on / off the capacitors, producing a measurement which is independent of the static charge on the device, the material Young’s modulus, or the equilibrium stress distribution. The feedback thus ensures the steering element retains the same positional response to a given position command vector, definitionally remaining continuously calibrated. In some implementations, each steering element is controlled via a group of adjustable paddles whose position is measured by capacitance sensing.TIME STABLE PERFORMANCE

[0063] The second feature of POCCAP is the time stable performance of the OCS elements, which is easily achieved once all elements are continuously calibrated. There are two main types of changes in the OCS architecture: position and parameter. Position type changes are easily observed and compensated for. However, parameter type changes are more difficult to observe and compensate for.

[0064] Compensation changes in the position command vectors for a steering array can be grouped by their spatial distribution into two categories: array and sub-array. Array level changes are defined as changes to the position command vectors for any given element which are a function of the spatial location of the element in the steering array. The array level changes follow discernable patterns over the array, such as a constant uniform changes for all elements (owing to the reorientation of the whole array) or lens-like distortion increasing with distance from the center (owing to bow on the array). FIG. 2 depicts an example of array level change (input steering array 108 has changed orientation, and output steering array 112 has changed position).

[0065] The order of the array level changes can be thought of as the order of the exponent on the position vector required to define the change as a function of location on the array. Forexample, a uniform change over the array would be a Oth order array level change as it is independent of location on the array and as such the array level change d0(r) can be defined as d0(r) = C * r° (where C is a constant coefficient and r is the radius of the lens). A lens-like distortion increasing with distance from the center (owing to bow on the array) would be a 1st order array-level change as it would be defined with a function of the form d0(r) = Co* r° + * r1.

[0066] Sub-array level changes do not occur at the array level but instead can vary on an element-by-element basis. Array level changes are easier to measure and mitigate compared to changes at the sub-array level which show no spatial dependence and require unique position command changes for each element in the array. Array level changes can be observed by a limited number of elements then inferred for the other non-measured elements. Sub-array level changes cannot be inferred from other elements and thus pose a much larger problem to precision operation.

[0067] Parameter changes are changes in the defining parameters of the components, such as the electrical-to-mechanical mapping of the steering elements. The compensation of the position command vector for positional changes primarily occurs at the array level. The compensation of the position command vector for the parameter changes (as seen in FIG. 3) often occurs at the sub-array level. In FIG. 3, sub-array level changes have occurred on each of the steering elements of steering array 304, resulting in a sub-array level change in the 2D command corresponding to each steering element. The inference possible with array level changes is relied on by POCCAP, as inference enables indirect prediction of the changes occurring at a given element without requiring in-channel sensing for that element. Effective POCCAP requires that the OCS architecture be designed to suppress sub-array level changes in favor of array level changes. Therefore, parameter changes should generally be mitigated in a POCCAP approach, while positional changes can be tolerated.

[0068] The first positional type of change in OCS architecture are in the six degrees of freedom (6DOF) position (location and orientation) of the optical elements. As noted previously, the most common errors in the OCS architecture are generated by location and orientation changes of the fiber, collimator, reflector, lens and / or steering elements. These changes can be compensated for by adjustments of the 4D position command vector (that is, the steering positions of the steering array(s)). In various implementations, some changes (for example, caused by non-linear cases such as aperture clipping) cannot be compensated for by adjusting the 4D position command vector.

[0069] This can be understood by examining the primary optical axis of the design. The primary optical axis sets the nominal beam path, with all off-axis behavior effectively modeled as changes relative to the nominal path. Changes in the nominal 4D position command vector can provide the necessary adjustments to re-align the nominal beam path to the changed primary optical axis caused by positional changes of the optical elements. Updates to the command vector required to create a beam parallel to the nominal beam ensure that the optical system always has a consistent response to relative changes around the nominal beam path.

[0070] Positional changes of optical elements can be compensated by altering the nominal 4D position command vector (the process for which is described in greater detail below with respect to FIGS. 4A-4B and FIGS. 7A-7B). The nominal 4D position command vector provides a defined 2D nominal position command vector for each of the 2D steering elements. The nominal position command captures the position command for each steering element required to send the beam parallel to the primary optical axis. The relative 4D position command vector for any given port-to-port mapping is defined relative to the nominal position command vector, so the complete port-to-port command combines the nominal and relative position commands. In some implementations, the complete command is determined by adding the nominal command to the relative command. In some implementations, the complete command is determined by tracking the change in the nominal position command vector and adding that change to the relative position command vector. In both cases the nominal position command tracks changes in the nominal beam path and the complete position command combines this change with the constant relative position command. The active compensation adjusts the nominal 4D position command vector that is the baseline term on which all the port-to-port relative 4D position command vectors are defined.

[0071] The ability to compensate for all positional changes with changes in the steering element orientation is reviewed for each case (location and orientation). Small scale location changes ( dx) of downstream (by dz) optical elements can be mitigated by small scale orientation changes (approximately dx / dz depending on element type and location) of the beam leading to the downstream elements. Small scale orientation changes (d0) for optical elements produce downstream angle errors which can be mitigated by small scale orientation changes of the beam (approximately — dQ depending on element type and location). Both types of positional changes (location and orientation) can thus be largely compensated by proper adjustment of the steering element position commands. As noted previously, the compensation of the position command vector for optical element positional change occurs at the array level, meaning that it is spatially defined for the elements in the steering array with a uniform value being the most common form. Optical element positional changes alter the primary optical axis for the whole array and thus alter the nominal beam angle (and thus the nominal position command vector) equally for all elements in the steering array.

[0072] The second type of change in OCS architecture is in the parameters of the optical elements. The parameter changes cause changes to the position command vectors that generally do not show array-level variation but rather vary on an element-by-element or even port-to-port mapping level. Each of the primary optical elements in the OCS (fiber, collimator, reflector, lens, and steering elements) have performance parameters, such as the fiber core diameter, the collimator lens focal lens, reflector / lens focal length and steering element position command- to- output mapping. Variation in these parameters can alter the 4D position command vector for each input and output port combination in a way that will vary on a sub- array element-by- element level. For example, if a reflector placed between the input and output steering element warps (changing from infinite focal length to a finite value), then the 4D position commandvectors for any given port-to-port mapping will change in a way that is not a function of the input steering element location on the array.

[0073] Additionally, changes in the steering element parameters including their electrical-to- mechanical behavior will also cause sub-array level changes. Sub-array level changes are not acceptable for POCCAP operation. It is therefore important to use optical elements in the OCS which will be highly stable and not show significant parameter changes. Performance stability is generally easily achieved with passive elements like lens focal lengths, but this can be more difficult with active elements like steering elements. Specifically, steering elements often drift in their electrical-to-mechanical performance but should have a consistent behavior for the POCCAP, such that the electrical-to-mechanical performance is stable over time. POCCAP steering elements must remain continuously calibrated over time. Positional feedback at the element level will result in POCCAP compatible, continuously calibrated, and time stable steering elements.

[0074] The requirement for time stability in the elements can be quantitatively defined via the mapping of the performance effect to the beam location on the output fiber. Error propagation analysis of sub-array level performance variations in time should result in negligible insertion loss (much less than IdB, preferably less than O.ldB) in order to support overall low insertion losses for a high-performance OCS architecture. Time stable optical elements can be defined as elements whose time variation (in a specific environment) that when propagated to link insertion losses show less than O.ldB scale effect. This definition is specific to the environment as well as the optical element. Performance changes of concern will appear as changes beam location or orientation change at the output fiber face. If these grow too large, then the variations will cause the beam to lose alignment to the fiber core and lose the link. An OCS design using non-time stable optical elements may function but will have sacrificed much of the potential performance benefit of the POCCAP approach. An OCS designed with time stable optical elements will more effectively exploit the performance benefit of the POCCAP approach.OUT-OF-CHANNEL SENSING

[0075] The third feature of POCCAP is the use of out-of-channel sensing (OoCS) - sensing that does not share the optical beam path of any data signal beams through the OCS. This prevents the sensing from interfering with the data signal. In-channel sensing approaches used in other OCS designs impose some alignment penalty, whether through partial measurement failing to observe certain errors or through full measurement attenuating the data signal. The OoCS approach uses dedicated optical sources and sensors located throughout the OCS architecture that can measure changes in the 4D position command vectors but do not interfere with the primary data signal channels.

[0076] Optical feedback sensing brings significant cost, complexity, and power requirements to the OCS design. Therefore, it is better to use the minimum required sensing capability for a given OCS architecture. Some in-channel approaches use feedback sensing for every optical channel, leaving the OCS design burdened with significant cost, complexity, power, and opticalloss challenges. When the first two features of POCCAP are met (continuous calibration and time stability) for an OCS architecture, then a cascade of benefits accrue. Parameter changes in the OCS architecture are suppressed and therefore the primary architectural changes become position changes (which are array level changes that can be measured at one location and inferred at other locations). The continuous calibration of the elements ensures that relative changes in position command vectors can be transferred, so measured changes in the position command vector for certain sensing elements can be used to infer the necessary changes in the position command vector for the other un- sensed elements.

[0077] In some implementations, alignment sensing ports can be located throughout the OCS architecture (as shown in FIGS. 4A^IB) not merely at the port arrays. For example, in some implementations, alignment sensing ports located at the planes of the steering arrays are also useful, as these provide an intermediate measurement of alignment for establishing the nominal 4D position command vector for the steering arrays. Alignment sensing ports at the planes of the steering arrays can be created simply with the end of a fiber mounted onto the steering array frame, such that they capture any light incident on the fiber core. The alignment sensing ports are physically linked to a primary optical element and thus do not inject further alignment error to the measurement. For example, in FIGS. 4A and 4B, input steering array 416 is physically linked to alignment port 420- 1 and alignment port 420-2 and output steering array 424 is physically linked to alignment port 432-1 and alignment port 432-2.

[0078] Alignment sensing ports should be physically linked to primary optical elements such as the fiber arrays and steering arrays. In other words, sensing ports must have a physical connection to the primary optical element in such a manner that the alignment sensing process, whether from a single port or a combination of measurements from several ports in the same general area of the architecture, measures the position of the physically linked primary optical element. This physical linkage provides a metrology link between the alignment sensing and the important optical element location. As an example, alignment ports linked to the fiber array provide a measure of the positional change of the fiber array, likewise for alignment ports physically linked to the input / output steering arrays or any of the other optical elements. Alignment sensing that is not physically linked to an optical element, as shown in previous approaches to OCS architecture designs, provides a measurement of an unlinked reference feature which is unassociated with any important parts of the architecture. The resulting alignment measurement reads in changes in the un-linked reference feature in addition to changes caused by all other optical elements along the measurement path. The changes in the unlinked reference feature are of no value to ensuring port-to-port alignment and thus can contaminate the measurement process, degrading alignment.

[0079] In some implementations, the alignment sensing ports can be wired up to act as either a source or sensor in a switchable manner using a minimal number of sources and sensors such that the system can reallocate the alignment optical source to any set of alignment ports and the alignment sensor to any other set of alignment ports through fiber switches. A well designed POCCAP OCS architecture may only need a single alignment optical source and a single opticalsensor. The system can sequentially determine the mapping from source to sensor for each pair, building up an updated map of the nominal 4D position command vector in a matter of seconds, while the positional changes will be occurring at thermal time scales of minutes to hours.

[0080] FIGS. 4A-4B show an example of the in-process measurement of the nominal position command vector using intermediate measurements. Each alignment port of input fiber array 404 is sequentially chosen to act as the alignment source (for example, alignment source 412-1 is activated and then alignment source 412-2 is activated). The light (such as alignment light 450- 1) from the active fiber array alignment port (alignment source 412-1) shines on a dedicated alignment element (alignment steering element 422-1) in input steering array 416. Alignment steering element 422-1 in input steering array 416 uses a search algorithm to find each of the output steering array alignment sensing ports, one after another. For example, alignment port 432-1 is found and then alignment port 432-2. In some implementations, the search algorithm is an Archimedean spiral which starts at the last known location of the port, then begins an expanding search around that region. When alignment light 450-1 passes over the specially targeted output steering array alignment sensing port (alignment port 432-1), the alignment sensor tracks that a signal has been received and restarts the spiral process around the location corresponding to the alignment element’s orientation when the alignment sensor received the signal. This iterative process will converge the beam into alignment and after sufficient iteration that the target is continuously illuminated, can be smoothly adapted to a circular nutation measurement which provides continuous alignment measurement by extracting phase and amplitude as observed by the receiver.

[0081] Once the alignment measurement is complete, the beam is shifted to the next alignment sensing port (alignment port 432-2) of output steering array 424. The alignment measurements corresponding to each of the output steering array alignment sensing ports can then be combined to generate the nominal position command vector for the input steering array as seen by the active input fiber array alignment port (alignment source 412-1). In some implementations, output steering array 424 alignment port 432-1 and alignment port 432-2 are symmetrically distributed around output steering array 424, and the average of the alignment measurements can be considered the nominal input steering array position command vector measurement for the given active input fiber array alignment port (alignment source 412-1). This process is repeated for each of the input fiber array alignment ports (for example, repeated with alignment source 412-2, alignment light 450-2, alignment steering element 422-2, alignment port 432-1, and alignment port 432-2), generating redundant measurements of the nominal input steering array position command vector, which can be averaged together to create a single high accuracy nominal input steering array position command vector. The nominal position command vector measures and compensates for positional changes in all the optical elements leading up to the output steering array, including even the location of the output steering array.

[0082] In FIG. 4B, the process is repeated for the final leg of the OCS architecture to complete the in-process measurement of all changes. The nominal input steering array position command vector is updated as described above and alignment steering element 422- 1 and alignmentsteering element 422-2 on input steering array 416 are now steered to point at their corresponding alignment steering elements (alignment steering element 428-1 and alignment steering element 428-2) on output steering array 424 through the combination of the nominal with the relative commands. Once again, each alignment port of input fiber array 404 is sequentially chosen to act as the alignment source. This time, the light passes from the input fiber array alignment port to the corresponding input alignment steering element and now to the corresponding output alignment steering element (for example, from alignment source 412-2 to alignment steering element 422-2 to alignment steering element 428-2 on output steering array 424). Alignment steering element 428-2 on output steering array 424 is actuated using a search algorithm to find each alignment sensing port of output fiber array 436, one after another. Once each alignment measurement for a given output fiber array alignment sensing port (for example, alignment port 444-1) is complete, the beam is shifted to the next output fiber array alignment sensing port (for example, alignment port 444-2). As noted previously, the alignment measurements corresponding to each of the output fiber array alignment sensing ports can then be combined to generate the nominal position command vector for the output steering array as seen by the active input fiber array alignment port.

[0083] This process can be repeated for each of the input fiber array alignment ports, generating redundant measurements of the nominal output steering array position command vector, which can be averaged together to create a single high accuracy nominal output steering array position command vector. The combination of the nominal input steering array position command vector and the nominal output steering array position command vector provide a full measure of the nominal 4D position command vector.

[0084] Alternately, the measurement can be made directly from the input fiber array alignment ports to the output fiber array alignment ports with no intermediate measurements. This process would generate the full 4D position command vector without intermediate measurements, however the approach requires the OCS architecture to be well aligned at the start so that the signal can immediately be observed at the output port. Therefore, this direct end-to-end approach is best used in operation when the scale of change in the OCS architecture is small relative to the frequency of alignment measurements. Each of the input fiber array alignment ports is sequentially chosen to act as the alignment source. The corresponding alignment steering element on the input array is now steered to point at the corresponding alignment steering elements on the output array using the combination of the nominal with the relative commands. The light passes from the input fiber array alignment port to the corresponding input alignment steering element and to the corresponding output alignment steering element. Both the input and output alignment steering elements are now actuated using the previously defined search algorithm to optimize alignment to the target output fiber array alignment port.

[0085] This search can be done many ways including simultaneously in a 4D search or sequentially in two steps with a first nutation search of one of the steering elements to find optimal alignment, then a nutation search of the other steering element to find optimal alignment. The end result of the search is the 4D position command vector required for optimalalignment of the corresponding alignment ports. This process can be repeated for each of the output fiber array alignment sensing ports. As noted previously, the alignment measurements corresponding to each of the output fiber array alignment sensing ports can then be combined to generate the 4D nominal position command vector. This process can be repeated for each of the input fiber array alignment ports, generating redundant measurements of the nominal output steering array position command vector, which can be averaged together to create a single high accuracy nominal 4D position command vector.

[0086] The entire alignment measurement process can be repeated in the reverse direction, using the output fiber array alignment ports sequentially as the alignment source and working backwards using either of the two approaches described above (intermediate or no intermediate measurements). This reversed direction measurement process produces a redundant measurement of the nominal 4D position command vector that can be used to further improve the accuracy of the overall measurement. POCCAP uses a limited number of alignment sensing channels (for example, less than the number of data ports in the OCS). The number of required alignment sensing channels increases with the order of the array level changes, as more complex array level change must be sensed at more points. It is expected that the primary array level changes will be Oth order uniform changes, meaning that only a very limited number of alignment sensing channels will be required.

[0087] In the simplest case, alignment sensing channels can be implemented at the four corners of the port arrays. The use of the four corners provides redundancy to the measurement, as well as a measure of symmetry by averaging the measurement to mitigate secondary thermal effects. For example, in FIG. 5, steering array 500 includes four alignment ports (alignment port 508-1, alignment port 508-2, alignment port 508-3, and alignment port 508-4) physically linked to the corners of steering array 500. Steering array also includes alignment steering elements (element 504-1, element 504-2, element 504-3, and element 504-4) for directing alignment light. In some implementations, alignment steering elements are located at the corners, middle, and / or sides of a steering array. Steering array also includes data steering elements (512-1 - 512-12) for directing data signals from input ports to additional steering arrays and / or output port arrays. In some implementations, alignment ports 508-1, 508-2, 508-3, and 508-4 are not located at the corners of the steering array. In some implementations, alignment ports 508-1, 508-2, 508-3, and 508-4 are located symmetrically along the exterior the steering array.

[0088] Alignment sensing ports can also be reflectors such as retrocube reflectors to provide a return signal for alignment sensing, this will return the light back down the alignment sensing port it was originally emitted from. Whether the alignment sensing ports are fibers, reflectors, or other optical components such as sources / detectors / chiplets, in all cases they play the same role of creating a reference point upon which the system can align.

[0089] In some implementations, the alignment process (updating of the nominal command vector) is carried out while the OCS is operation (i.e., while data is being transmitted). For example, data signal 454-2 from data port 408-2 is received at output port 440-1 via steering elements in input steering array 416 such as data steering element 418 and data steeringelements in output steering array 424. In some implementations, data port 408-1 transmits data signal 454-1 and output port 440-2 receives data signals while the alignment occurs.DETERMINING THE RELATIVE 4D POSITION COMMAND VECTOR

[0090] The fourth feature of POCCAP is a way to determine the relative 4D position command vector for a given desired data port-to-port alignment. As noted previously, the relative 4D position command vector is combined with the nominal 4D position command vector to produce the complete 4D position command vector for a given port-to-port mapping for data channels. The nominal 4D position command vector can be continuously measured and updated during the operation of the device including during the calibration process.

[0091] The relative 4D position command vectors are not as easily discerned. The relative 4D position command vectors must be determined by calibration of the unit prior to regular operation. This calibration process can be carried out by supplying an optical signal to each input port sequentially then using initial estimates combined with the nominal position command vector and a search algorithm to find the best fit relative 4D position command vector for each output data port. This process produces a specific measured relative 4D position command vector for any given input and output port combination. The results of the process can be compared to the nominal position command vector, and the comparison (generally the difference) stored in a number of different ways including as a full lookup table or turned into a function with fit parameters. In all cases, the result is a defined relative 4D position command vector for any given input and output port combination.

[0092] The calibration process involves a full search through 4D space which can be time consuming. This process can be accelerated by using the intermediate reference features, the input / output steering array alignment sensing ports. The input steering element for a given input data channel can be scanned over the output steering array to find alignment to the output steering array alignment ports. Once this alignment is found, the desired output steering element is more easily located as a relative change. Now all that remains is the output steering element adjustment to align the beam into its corresponding port. This process of using the input / output steering array alignment sensing ports breaks the full 4D space search into a more efficient process of two sequential 2D space searches. The net result is the same relative 4D position command vector identification for the given input and output port combination.

[0093] In some implementations, the alignment process can be performed with data port 408-1 and data port 408-2 in addition to alignment source 412-1 and alignment source 412-2. For example, data port 408-2 emits light which is reflected from data steering element 418 on input steering array 416 into alignment port 432-1 and then alignment port 432-2. These measurements are recorded and averaged and used to update the relative command vector for the specific data port. This process can be repeated with each data port and its corresponding data steering element and each alignment port.

[0094] The input / output steering array alignment sensing ports provide another benefit of providing a mechanism for in-situ measurement of fiber-collimator parameter change andrecalibration. If the fiber moves relative to the collimator element, this changes the collimated beam direction and appears as a sub-array level element-by-element change. The OCS architecture can track these changes by the same approach used to find the nominal position command vector for the input steering array. A signal of any kind (including data) can be sent into the data channel of concern (m), then the corresponding input steering element can be steered to align to the output steering array alignment sensing ports. The net measurement from all these ports provides a measure of the nominal position command vector for that steering element (m). The OoCS measurement of the input steering array nominal position command vector should fully track all changes in the newly measured nominal position command vector. Any change not accounted for by the nominal position command vector measurement can be considered a change to the stable optical parameters of that mth data channel and that change can be applied to the lookup table for all the position commands which use the mth input steering element. In this way, the OCS architecture can be recalibrated in-situ by sequentially rerouting the data channels to the reference ports (taking them out of operation for a few fractions of a second), recalibrating and bringing them back online. This can be done when channels are not in use during regular operation. The same way works in reverse for recalibrating the output steering array by sending signals into the nth data channel and aligning these to the input steering array alignment sensing ports. For both directions this is an out-of-channel sensing approach as the data channel is not passing any usable data while this recalibration is being carried out, so the alignment signal is not sharing the path with any data signals.BLOCK DIAGRAM

[0095] FIG. 6 is a block diagram of an example system for precision out-of-channel-sensing. Control module 604 receives input / output connection commands at base angle determination module 608 from switch control module 656. Base angle determination module references the relative command vector LUT stored in relative position mapping database 644 and sends the initial data steering element position command (i.e., the relative command vector) for the steering elements that correspond to the request input / output combination. Relative command vector LUT stores position commands for connecting any input port with any output port (by adjusting data steering elements 640 on one or more steering arrays). In some implementations, relative command vector LUT is configured as a matrix and the command to connect input m with output n is stored at position [m, n] in the matrix.

[0096] The initial relative command vector stored in relative position mapping database 644 is provided by initial position module 652, which determines the initial relative command vector (e.g., at the time of assembly). Relative position mapping database 644 can be updated with data from relative position calibration module 648. Relative position calibration module 648 communicates with data steering elements 640 and / or alignment receivers 636 to update the relative command vector by performing a process similar to the alignment process described in FIGS. 7A-7B. Relative position calibration module 648 sends a data disable signal to switch control module 656 so that light source 660 (i.e., input and / or output ports used to transmit datasignals) transmits a light beam instead of a data signal. In some implementations, the light from light source 660 is a steady light source, which enables detection of the light source more easily than a light source that might turn off and off as with data signals. Light source 660 (e.g., an input or output port) is shone onto the corresponding data steering element (e.g., a steering element of data steering elements 640), and directed to alignment receivers 636 on a second steering array and / or data port array. Data steering elements 640 are actuated until the light is directed into each receiver of alignment receivers 636. The position of the steering element when aligned with each receiver of alignment receivers 636 is recorded and averaged and used to update the relative command vector.

[0097] The initial data steering element position command is adjusted at alignment module 616. Alignment module 616 adjusts the initial data steering element position command by summing the initial command with the measured positional offsets (i.e., the nominal command vector) stored in offset storage 620. Offset storage 620 communicates with offset determination module 624, which uses data from alignment steering elements 628, alignment sources 632, and alignment receivers 636 to detect array-level alignment changes in the OCS architecture. Alignment module 616 also manages element-level changes by measuring steering element positional changes via steering element calibration components 612 and adjusting the calibration and position of steering elements as necessary. Steering element calibration feedback is provided by steering element calibration components 612 on alignment steering elements 628 and data steering elements 640. In some implementations, steering element calibration components 612 are capacitors attached to each element of alignment steering elements 628 and each element of data steering elements 640.

[0098] In some implementations, alignment receivers 636 are combined with alignment sources 632 resulting in elements capable of sensing and creating alignment signals. In some implementations, alignment receivers 636 and alignment sources 632 are separate elements that are physically linked. In some implementations, alignment receivers 636 and alignment sources 632 are physically linked to the output port arrays, input port arrays, and / or steering arrays that hold alignment steering elements 628 and data steering elements 640 of the OCS. In some implementations, alignment receivers 636 and alignment sources 632 are located at the corners of the output port arrays, input port arrays, and / or steering arrays. In some implementations, alignment receivers 636 and alignment sources 632 are attached to the four corners of the input port array, the output port array, and / or the alignment steering array containing alignment steering elements 628 and data steering elements 640.FLOWCHARTS

[0099] FIGS. 7A-7B are a flowchart of an example method for aligning out-of-channel sensing elements and updating the nominal command vector. While the method below describes the use of fibers at the optical data channel ports, other light sources and detection methods can be used. In various embodiments, an input and / or output port includes a single optical data source. In various embodiments, an input port includes two or more optical data sources. Forexample, an input and / or output port may include a bundle of optical sources (such as fiber, lasers, etc.) pointed at a single steering element, which results in more sources than steering array elements. In various embodiments, each source of the bundle of optical sources corresponds to a collimator unique to the source.

[0100] The method begins at 704 where control selects a source alignment port (i.e., an alignment signal source) from a first port array (for example, an alignment port physically linked to the input port array). At 708, control selects a destination alignment port in the second steering array (for example, the output steering array). At 712, control determines if the alignment element (that corresponds to the source alignment port) in the first steering array (for example, the input steering array) is aligned with the destination alignment port in the second steering array. If alignment has been achieved, control transfers to 720. If alignment has not been achieved, control transfers to 716 where the position of the alignment element in the first steering array is adjusted and control transfers to 712.

[0101] At 720, control captures the alignment measurement (i.e., the position of the alignment element the enables the alignment signal to travel from the source alignment port to the destination alignment port). At 724, control determines if there are unmeasured destination alignment ports in the second steering array (i.e., has the port alignment signal source been aligned with all destination alignment ports in the second steering array). If there are unmeasured destination alignment ports in the second steering array, control transfers to 726 where the next destination alignment port is selected. If all destination alignment ports on the second steering array have been aligned with the selected source alignment port, control transfers to 728. At 728, control determines if all source alignment ports from the first port array have been measured. If there is an unmeasured source alignment port remaining, control transfers to 732. If there are no unmeasured source alignment ports remaining, control transfers to 736.

[0102] At 732, control selects the next source alignment port (for example, the next alignment port in the input port array) and control continues to 708. At 736 the nominal command vector is updated based on the collected measurements. At 740, control adjusts the position the alignment elements of the first steering array (using the updated nominal command vector) to align with the corresponding alignment elements in the second steering array (i.e., so that the signals from the source alignment ports are reflected from the corresponding alignment steering elements in the first steering array to the corresponding alignment steering elements in the second steering array).

[0103] At 744, control selects a source alignment port from the first port array (for example, an alignment port physically linked to the input port array). At 748, control selects a destination alignment port from the second port array (for example an alignment port physically linked to the output port array). At 752, control determines if the steering element in the second steering array (for example an alignment steering element in the output steering array) is aligned with the alignment port in the second port array (for example the output port array). If not, control transfers to 756 adjusts the position of the alignment steering element in the second steeringarray and control returns to 752. If the alignment steering element in the second steering array is aligned with the alignment port in the second port array, control transfers to 760. At 760, control captures the alignment measurement (i.e., the position of the alignment element in the second steering array the allows the signal from the source alignment port to reach the destination alignment port in the second port array).

[0104] At 764, control determines if there are unmeasured destination alignment ports in the second port array (i.e., has the currently elected source alignment port been aligned with all destination alignment ports). If the selected source alignment port has not been aligned with all destination alignment ports, control transfers to 766. At 766, control selects the next destination alignment port from the second port array. If the selected source has been aligned with all destination alignment ports, control transfers to 768. At 768 control determines if there are source alignment ports that have not been measured. If yes, control transfers to 772 and the next source alignment port is selected and control continues to 748. If there are no unmeasured source alignment ports (i.e., all source alignment ports in the first port array have been aligned with all destination alignment ports in the second port array), control transfers to 776. At 776 control updates the nominal command vector based on the captured measurements and control ends.

[0105] In some implementations, the method is executed by aligning alignment ports physically linked to the input ports with alignment ports physically linked to the output steering array (via alignment steering elements on the input steering array) and then finally with alignment ports physically linked to the output port array (via alignment steering elements on the output steering array). In some implementations, the method is executed in the opposite direction by aligning alignment ports physically linked to the output port array with alignment ports physically linked to the input steering array (via alignment steering elements on the output steering array) and then finally with alignment ports physically linked to the input port array (via alignment steering elements on the input steering array). In some implementations, the process is completed in both directions. In some implementations, the method is executed repeatedly, for example once every 0.1, 0.2, 0.5, 1, 2, 3, 4, 5 and / or 10 seconds. For systems in high vibration environments, the measurements might be carried more quickly, such as every 0.001 seconds.

[0106] FIG. 8 is a flowchart of an example method of maintaining steering element calibration within the OCS. At 804 control begins and determines if a position input for an element as begin received (as part of a connection request). If no position input as been received, control remains at 804. If a position input for an element has been received, control transfers to 808. At 808, control sends the position command to the element. At 812, control measures the element position (via continuous feedback). At 816, control determines if the element position is correct. If the element position is correct, control transfers to 812. If the position is incorrect, control transfers to 820 and adjusts the element position. Control then returns to 812.

[0107] FIG. 9 is a flowchart for an example method for aligning an optical input port with an optical output port. The method begins at 904 and control determines whether a connection request has been received. If yes, control transfers to 908. If no, control remains at 904. In some implementations, a connection request is a request to connect a particular input port with aparticular output port (for example, input A to output B). At 908, control accesses the nominal command vector. At 912, control accesses the relative command vector for the commands to move the steering elements associated with the requested input port (input A) and the requested output port (output B) so that light from the input port is received at the output port. Control continues to 916, where the complete command vector is determined by summing the nominal command vector and the relative command vector. Then control continues to 920 and executes the command vector, aligning the input steering elements so that the optical beam is directed to the correct output steering element. At 924, control continues to execute the command vector and aligns the output steering element with the request output port so that the requested input and requested outputs are optically connected (i.e., light from input port A is received at output port B). Control then returns to 904.IMAGE TRANSFER INTRODUCTION

[0108] The present disclosure describes a system for achieving high stability and low optical loss alignment between light source arrays (such as fibers, lasers, LEDs, and / or other light sources) and steering arrays. For example, this may be used in optical networking devices, such as optical circuit switches (OCSs). The proposed system uses a combination of passive optical elements and correction adjustments on the steering elements to transfer the image of the collimated port outputs to the steering elements while ensuring that the beam collimation is maintained after passage through the fiber-steering (combined fiber and steering) system. This image transfer approach significantly reduces manufacturing complexity as orientation misalignment is mitigated, resulting in a simpler, lower cost and more robust fiber array coupling to steering arrays. While optical fibers (including fiber arrays used as input and output sources) are frequently described throughout the disclosure, the approaches, structures, methods, and systems described below are equally applicable to systems that use alternatives to fiber including various light sources and light detectors. Together, light sources and light detectors are generically referred to as light stations. In various embodiments, each light station includes a single optical data source / detector. In various embodiments, a light station includes two or more optical data sources / detectors. For example, a light station may include a bundle of optical fibers or lasers pointed at a single steering element, which results in more light stations than steering array elements. In various embodiments, each source / detector of the bundle of optical sources / detectors corresponds to a collimator unique to the source / detector.IMAGE TRANSFER

[0109] The proposed system uses a passive optical relay assembly to carry out an image transfer from the collimator array to the steering elements on both the input and output ports which preferably conserves the Lightfield Vector Irradiance Distribution (LVID). During image transfer, all the light from any single point on the object plane is collected by a lens (or other optical structures) and focused onto the image plane at a single point (i.e., light from a plane is transferred to a parallel plane in a reverse orientation). This is the image transfer process enabledby lenses. Light transfer is possible on light emanating over a range of angles as long as angle of the light falls within the aperture of the lens. Image transfers is insensitive to angle as long as the light falls within the necessary range. This insensitivity is the key insight to the proposed invention. The image transfer approach avoids generating any displacement error at the steering element. Instead, the initial orientation angle error of the collimated beam through the image transfer approach only generates orientation angle error at the steering array. Since the initial orientation angle error of collimated beam tends to be of small scale, this is easily correctable by the steering element. Although the specification refers to a lens or lenses as the optical transfer elements for simplicity, a variety of optical structures that generate a focus can be used (such as curved reflectors).

[0110] In standard OCS architecture, the commonly used free space propagation approach between fibers and steering elements translates an initial orientation angle error (of the light source array) into both orientation angle error and displacement error at the steering element (i.e., a slight angular misalignment results in both the beam wandering off the desired steering element and a change of the impinging beam angle onto the steering element). For example, in FIG. 10, input port array 104 is misaligned with input steering array 108, resulting in misaligned light 120 not reaching the desired destination on input steering array 108. Similar misalignments can occur at an individual light source level. FIG. 11 A depicts an ideal light source array (including light sources 1104-1, 1104-2, 1104-3, and 1104-4) and collimator array (including collimator 1108-1, 1108-2, 1108-3, and 1108-4). In FIG. 11B, light source 1104-4 has become misaligned with collimator 1108-4. In the image transfer approach, a slight angular misalignment results only in a change of the impinging beam angle onto the steering element. The image transfer approach thus reduces the manufacturing complexity of the OCS and most importantly reduces its environmental sensitivity during operation, both of which directly impact cost and performance.BLOCK DIAGRAM

[0111] FIG. 12 is a functional block diagram of networking device 1244 using image transfer lenses (or other optical structure) to reduce the required level of manufacturing precision and insertion loss from misaligned OCS elements. Switch control module 1204 receives a routing signal (for example, a signal that identifies an input and output pairing) and controls which light sources of input ports 1216 are currently transmitting data signals (i.e., light) via an on / off command signal. Switch control module 1204 transmits input / output port connection commands (i.e., which input port should transmit data to a particular output port) based on a received connection request. In some implementations, switch control module 1204 receives a connection control signal that includes a request to transmit information. In some implementations, switch control module 1204 generates input / output port commands and on / off commands based on the connection control signal. In some implementations, the connection control signal is not received from an outside source but is generated by switch control module 1204. Angle determination module 1208 receives the input / output port connection commands and determinesthe angles of the individual steering elements of data steering elements 1228 that enable the transmission of light from the requested data source (light source) to the requested output source.

[0112] Input ports 1216 receive an input signal and emit light which pass through input port collimators 1220 and input image transfer lens set 1224. In some implementations, input image transfer lens set includes a single lens. In some implementations, input image transfer lens set 1224 includes a pair of lenses. Light is transferred onto data steering elements 1228. Data steering elements 1228 have moved into the positions received from angle determination module 1208. Data steering elements 1228 include an input steering array and an output steering array. In some implementations, data steering elements 1228 includes additional steering arrays and / or lenses. Light then passes through output image transfer lens set 1240. In some implementations, output image transfer lens set 1240 includes a single lens. In some implementations, output image transfer lens set 1240 includes a pair of lenses. Light is transferred from output image transfer lens set 1240 to output port collimators 1236 and output ports 1232. In some implementations, input ports 1216 includes two or more input ports that are directed to a single output port of 1232. In some implementations, the multiple input signals from the two or more input ports are distinguished by a digital identifier in each signal segment. In some implementations, the signals are distinguished via a time-division multiple access approach coordinated among the sources. In various embodiments, an input and / or output port includes a single optical data source. In various embodiments, an input and / or port includes two or more optical data sources. For example, an input and / or output port may include a bundle of optical sources (such as fiber, lasers, LEDs, etc.) pointed at a single steering element, which results in more sources than steering array elements. In various embodiments, each source of the bundle of optical sources corresponds to a collimator unique to the source.ADVANTAGES OF THE IMAGE TRANSFER APPROACH

[0113] The reduction in beam center displacement error at the steering element owing to the use of the image transfer approach has significant implications for manufacturing. The angular alignment from a collimated array is generally the most difficult alignment to achieve, as it is extremely sensitive to small errors in the location of the fiber and collimation lens. The image transfer approach removes the system sensitivity to angular misalignment. The output from the collimation array need not all be perfectly aligned when using the image transfer approach, which reduces the difficulty of aligning fibers as well as manufacturing the collimator array. The image transfer will replicate any angular errors of the beam onto the surface of the steering elements but will do so without adding in any beam-center displacement errors found in free- space propagation transfer. These displacement errors are the most difficult errors to remove owing to their sensitivity to extremely small motions / deviations of the fiber and collimation lens. The beam orientation angle error imaged onto the steering element can now be resolved with a tip / tilt correction applied by the steering elements so that the all the beams emitted from the steering elements are parallel by default, producing the desired on-axis collimated beamoutputs from the steering elements. In other words, the same desired parallelism of the various channels in the array can be achieved but it is done with an optical element and a tip / tilt offset correction by the steering element instead of through high precision manufacturing, assembly, and / or control of the collimator and fiber array. The shift of the challenge from the manufacturing space to the steering element control space means the system can be produced with lower cost, corrected to equivalent if not improved performance, and updated in-situ.

[0114] The tip / tilt correction terms for each steering element to achieve maximum coupling to the fiber can be determined by running the system after assembly and adjusting the steering element until a reference target is hit by the generated beam. For example, the output from the steering array can be passed through a lens and onto an array detector (such as a CCD) one focal length behind the lens. This detector at the Fourier plane will provide a measure of the angle of the light incident on the lens, as the pixel location maps to beam angle. Any deviation in the angle of the received light will appear as a change in the pixel location of the light on the detector. Each light source port can be activated and the associated steering element adjusted until the light impinging on the detector reaches the desired location on the detector corresponding to the desired primarily optical axis for the emitted beam. The calibration process can also be run in reverse, illuminating the steering array with a collimated beam and adjusting each steering element to maximize the signal at each of the corresponding ports. The calibration process will identify the correction term for each steering element in order to achieve optimal alignment to the corresponding fiber.

[0115] Another approach for running the calibration procedure in-situ relies on the use of calibration ports in the array. A designated calibration port in the output can be used as the reference port whose signal is interrogated for proper alignment. This reference port corresponds to a reference steering element on the output steering array. The input steering element can each be steered to point at the reference element, while the reference element is likewise steered to point at the reference port. The input steering element is then adjusted to maximize the signal output from the reference port. This calibration will identify the steering element angles required to achieve a known output angle, meaning that the beam angle required to optimally couple into each fiber can be calculated. The correction term can then be extracted from the known optimal coupling angle. The correction terms modify the equilibrium orientation of the steering element to produce an on-axis collimated beam output from each steering element. These correction terms account for the manufacturing and assembly variation in the fiber alignment to the collimator array, as both variations display as emissions angle errors from the collimator. Both variations should remain largely stable after manufacturing, however the assembly may shift over time. The calibration process can be run throughout the lifetime of the assembly, especially if in-situ calibration methods are used. The correction terms are defined here as modifying the equilibrium location of the steering elements. An on-axis collimated beam output from the steering element is generated when the elements are returned to the corrected equilibrium.

[0116] The image transfer approach applies to the optical transfer between the collimator to the steering element in both directions, collimator to steering and vice versa. The presentdescriptions has focused on describing the direction of light propagation from collimator element to steering element to support the light entering the OCS through the input ports. The same issues and solutions apply to the light propagation occurring in reverse, from the steering element to the collimator element for the light leaving the OCS through the output ports.SINGLE LENS IMAGE TRANSFER APPROACH

[0117] FIG. 13 is an example of an OCS architecture using a lens to carry out the image transfer approach. In this example, a single lens (image transfer element 1324) is placed two focal lengths from data signal light source 1304, which is a collimated output. This results in the re-creation (transfer) of light from data signal light source 1304 (at lightfield plane 1316-1) to two focal lengths from image transfer element 1324 (at lightfield plane 1316-2). In this approach, magnification is not constrained and can be tuned to other values if non-unity magnification is desired between the collimator array image and the steering array 1308 image. Light from data signal light source 1304 (collimator output HCGF LVID 1328) is transferred to steering array 1308 and steering element 1312 as long as the emitted light is between range of transferred angles 1320-1 and 1320-2.

[0118] Lightfield plane 1316-1 is the measure of all the light rays passing through the collimator array plane through all points on the plane. Lightfield plane 1316-1 has a 2D distribution of light ray vectors for each point on the 2D plane. The LVID for a given point describes where light is headed when leaving from the given point on the plane. The LVID is not the summation of the vectors to create a single value but rather it is a 2D plot (in Ox and 3y) showing the distribution of optical flux over the two angular degrees of freedom. It is generally desired that when steering elements are near their mid-range the steering array output LVID be approximately an On-Axis Delta-like Gaussian Function (OADGF), as the beam is composed of rays which are almost entirely parallel to the primary optical axis, so the 2D plot of the LVID would look like a delta function at the origin. The Gaussian vector distribution captures the nonzero divergence generated by the finite beam aperture. The OADGF LVID corresponds to an on- axis collimated beam which is parallel to the primary optical axis. While it is possible for designs to use later optics to fix beam divergence after the steering array (a non-zero divergence beam is useful in some applications), the collimated beam is more common and generally preferred. The OCS design works best with an on-axis collimated beam for each steering element output to support the transfer from the input steering elements to the output steering elements.

[0119] The single lens image transfer approach has performance variation dependent on the steering element LVID. It is preferred that the LVID is conserved (unchanged) through image transfer. If the LVID is not conserved through image transfer (as shown in FIG. 13), the light is imaged onto steering element 1312 in a manner such that it is not impinging normally onto steering element 1312 and is diverging. The LVID is shifted from centered around a horizontal ray, also called a horizontally centered Gaussian function (HCGF) (collimator output HCGF LVID 1328) to being centered around an angled ray (e.g., ITE LVID 1332). Collimator outputHCGF LVID 1328 is an OADGF LVID, so it is the ideal form for convenient use in an OCS. ITE LVID 1332 is no longer centered at the origin of the 2D angle plot, and is now distributed over a range of angles corresponding to a diverging beam. The initial primarily horizontal distribution of the vector irradiance of the data signal light source 1304 has not been conserved. Instead, it is now an off-normal vector irradiance when reaching steering element 1312. This poses a challenge for OCS design as the array of collimated beams are no longer collimated when they reach steering array 1308, thus requiring significant adjustment of the steering elements to compensate and return the array to its initial collimated state. The required compensation reduces the range of motion of the steering elements, limiting their range of motion left to steer the beam around the second steering array. The diverging light can be compensated via later optics or curvature of the steering elements, but it is an additional challenge. This challenge can be addressed in several ways as discussed below.

[0120] The single lens image transfer approach fails to produce the desired on-axis collimated beam output from steering element 1312. This can be fixed either by i) modifying the collimator output angles to be non-parallel and a function of port location or ii) by starting with a collimator output of an on-axis collimated beam and conserving it through the image transfer. In both solutions the image transfer approach requires a particular property of symmetry with regards the array of collimator output beams. If each collimator output beam is replaced with a single ray passing down the center of the beam, then these collimator rays must all substantially intersect at a point on the optical axis. The requirement for intersection results in forced symmetry around the optical axis. The requirement of collimator ray intersection on the optical axis is a necessary pre-requisite for the image transfer approach. Satisfying the imaging criteria for all ports requires radially symmetric optics. The image transfer approach can be done with collimator outputs that are parallel or non-parallel, but regardless of which approach is used, the pattern of the collimator output angle must be axi- symmetric around the optical axis to confirm to the symmetry of the image transfer elements. The symmetry requirement can be captured by the concept of the ray intersection point. The collimator rays must intersect at a point on the optical axis, whether this point is at a finite location (as seen in FIGS. 14A-14B) or at an infinitely distant location (as seen in FIG. 16).FINITE INTERSECTION COLLIMATOR: HYBRID COLLIMATOR ELEMENT

[0121] The challenges of the single lens image transfer approach (i.e., the data signal arrives as an off-normal vector irradiance at steering element 1312) can be overcome with a modified collimator element which cancels out the image transfer changes so that after the image transfer the steering element output is an on- axis collimated beam. The benefit of the pre-distorted approach is that the image transfer optical approach can be implemented with a single lens thus reducing optical losses from reflective interfaces. Two optical effects are required. First, the diverging light from the light source must be focused onto the lens’s focal plane. Second, all ports should focus exactly onto the optical axis at the shared focal point. This is not feasible using a single on-axis conventional lens as all ports’ diverging light LVIDs are effectivelyimaged to the same location. Two approaches are presented below for how to achieve the finite intersection of all the collimator outputs at the shared focal point. First, (shown in FIGS. 14A and 14B) is a hybrid optical element approach and second is a non-parallel collimator array approach (shown in FIG. 15).

[0122] As shown in FIGS. 14A and 14B, the finite intersection collimator output approach can be achieved with a hybrid optical element, for example an optic (e.g., modified collimator 1404) with collimating convex microlenses on the side facing the light sources that generate the necessary imaging focal length that supports object distance di and image distance d2, and a segmented flat side (e.g., tilting face 1412) facing away from the fibers that approximates an array of micro prisms, where the micro-prism array tilts each fiber imaged beam until that beam is pointed towards shared focal point 1408. Light is then transferred by image transfer element 1416.FINITE INTERSECTION COLLIMATOR OUTPUT: ALIGNMENT

[0123] The finite intersection collimator output approach can also be achieved with modification of the alignment of the light sources (data signal light source 1504 and data signal light source 1508) and (optionally non-parallel) collimator elements (collimator element 1516 collimator element 1520) as shown in FIG. 15. In this design, the role played by the microprisms in FIGS. 14A-14B is instead covered by the orientation of data signal light source 1504 and data signal light source 1508), set so that each collimator element output beam is aligned to intersect with shared focal point 1512. Collimator element 1516 and collimator element 1520 should be set to focus the beam to a point at shared focal point 1512. Collimator element 1516 and collimator element 1520 should all lie on lightfield plane 1316-1 corresponding to the image plane of the image transfer element because this ensures the image transfer is done without distortion. The collimator elements can, but do not need to be rotated non-parallel to one another to match the alignment of data signal light source 1504. Instead, the location only (and not rotation) of each data signal light source 1504 relative to collimator element 1516 can be controllably adjusted to be sufficient to steer the collimator output towards the shared focal point.

[0124] In some implementations, orientation of data signal light source 1504 is determined by the distance of the data signal light source 1504 from the center axis of the light source array and / or the center axis of image transfer element 1324. For example, data signal light source is more tilted than data signal light source 1508 because data signal light source 1504 is a further from the light source array than data signal light source 1508. In some implementations, the angle of tilt of data signal light source 1504 is determined by whether data signal light source is above or below the center axis of image transfer element 1324 and / or the center axis of the light source array. For example, data signal light source 1504 tilts downward because it is above the center axis of image transfer element 1324.

[0125] In some implementations, data signal light source 1504 is offset from collimator element 1516. In some implementations, the amount of offset is determined by the distance ofthe center axis of data signal light source 1504 from the center axis of the light source array and / or the center axis of image transfer element 1324. For example, data signal light source 1504 is more offset from collimator element 1516 than data signal light source 1508 is from collimator element 1520 because data signal light source 1504 is a further distance from the center axis of the light source array than data signal light source 1508. In some implementations, the offset direction is determined by whether data signal light source 1504 is above the center axis of image transfer element 1324 and / or the center axis of the light source array. For example, if data signal light source 1504 is above the center axis of image transfer element 1324 and / or the center axis of the light source array, data signal light source 1504 will be offset higher than collimator element 1516. If data signal light source 1504 is below the midpoint of image transfer element 1324, data signal light source 1504 will be offset lower than collimator element 1516.

[0126] The distance of data signal light source 1504 from the collimator element can also be adjusted for each port to ensure the light is focused onto the shared focal point. It is important to note that the image transfer approach ensures that the beam center displacement error at steering element 1312 is insensitive to small angle misalignment of the light at the collimator output. This small angle misalignment can take the form of small variation in the exact fiber location and misalignment of the collimator output focus vs the shared focal point. Both changes only produce proportional errors in arrival angle of the beam onto the steering element, they do not produce error displacement moving the beam off the steering element.INFINITE INTERSECTION COLLIMATOR OUTPUT

[0127] The second approach to achieving an on-axis collimated beam output from steering element 1312 is by starting with an infinite intersection collimator output and conserving it through the image transfer as noted earlier. This LVID conservation approach can be achieved through using symmetry in the design of the image transfer element (ITE). A symmetric ITE design cancels the LVID modification generated by the first half of image transfer element (1604-1) with the LVID modification generated by the second symmetric half of the image transfer element (1604-2). The net result is a steering element LVID which largely matches the collimator element LVID. The LVID conserving approach supports the use of a conventional parallel horizontal collimator array.

[0128] One example of a symmetric image transfer element using two lenses (image transfer element 1604-1 and image transfer element 1604-2) is shown in FIG. 16, however it should be noted that the symmetric image transfer element may use other optically symmetric layouts. As before, the actual collimator element output LVID must lie within the bounds (range of transferred angles 1620-1 and range of transferred angles 1620-2) of the transferrable LVID to reach steering element 1312. FIG. 16 is drawn specifically for the dl = fl = d2 = f2 scenario of a 4f telescopic lens system. LVID conservation can be achieved with non-4f approaches as long as dl and d2 are set to satisfy the image transfer criterion for the optical system. The two lenses need not have the same focal length, the ratio of focal lengths fl vs f2 can be tuned if magnification is desired between the collimator array image and the steering array image.

[0129] In certain ITE designs like the symmetric ITE in FIG. 16, stray light filters 1608-1 and 1608-2 may be placed at the Fourier plane to block stray light passing from the collimator element to steering element 1312. In some implementations, stray light filters 1608-1 and 1608- 2 are placed surrounding shared focal point 1612. Stray light filters 1608-1 and 1608-2 can be implemented by a single plate with a hole cut in the middle through which the shared focal point 1612 is passed. In some implementations, stray light filters 1608-1 and 1608-2 are a single piece of light blocking material with a hole or other light permeable material surrounding shared focal point 1612. A large hole in the filter corresponding to a large transmitted LVID enables compensation for large manufacturing and assembly error, but allows stray light in from a wide range of angles. A small hole in the filter corresponding to a small transmitted LVID limits the acceptable range of manufacturing and assembly error, but reduces stray light to a tightly controlled range of angles. The benefit of controlling the transmitted LVID through the stray light filter is that the lenses can be sized to accept the full aperture of the steering and collimator arrays, then the transmitted LVIDs can be separately tuned through the stray light filter to optimally balance manufacturing and assembly error vs stray light acceptance.TILTED COLLIMATOR ARRAY AND STEERING ARRAYS

[0130] The ITE designs shown above (e.g., FIGS. 13, 14A, 15, and 16) have shown steering array 1308 as being perpendicular to the optical axis (i.e., parallel to lightfield plane 1316-2 and lightfield plane 1316-1). In practice, steering array 1308 will likely be tilted to reflect the steered output from the array along a path that does not return through the ITE. The image transfer approach can be achieved with a tilted steering array by tilting the collimator array so that each matched set of collimator element and steering element are imaged to one another, as shown in FIG. 17.

[0131] In FIG. 17, light passes from the light source array containing data signal light source 1704-1, data signal light source 1704-2, data signal light source 1704-3, and data signal light source 1704-4 through image transfer element 1708-1 and image transfer element 1708-2 to steering array 1712. Steering array 1712 is tilted. Data signal light source 1704-1, data signal light source 1704-2, data signal light source 1704-3, and data signal light source 1704-4 are parallel to each other, but offset so that each light source is the same distance from steering array 1712. For example, data signal light source 1704-1 and steering element 1716-1 are the same distance apart as data signal light source 1704-4 and steering element 1716-2. In some implementations, the offset is achieved by tilting the light source array (e.g., so that it is no longer parallel to lightfield plane 1316-1). In some implementations, the light source array is parallel to lightfield plane 1316-1, but data signal light source 1704-1, data signal light source 1704-2, data signal light source 1704-3, and data signal light source 1704-4 are set at different distances within the array (e.g., data signal light source 1704-4 is further forward than data signal light source 1704-1).NON-IMAGING APPROACH

[0132] In various implementations, a non-imaging approach can be used for optical transfer between the input and output steering elements, similarly to why the imaging approach is used for transfer from collimator to steering elements. An imaging approach makes the beam location insensitive to angle, meaning that tip / tilt steering elements would be unable to modulate the beam location, which would render the OCS inoperable. Instead, a non-imaging optical transform can be used, and is represented in FIG. 18 by optical element 1804. In some implementations, optical element 1804 is any flat reflective surface. In some implementations, optical element 1804 is a curved reflective surface. In some implementations, the non-imaging optical transform uses free space propagation where input steering array 1808 points the beam towards output steering array 1812. In some implementations, the non-imaging optical transform approach uses a single reflector such as optical element 1804 where input steering array 1808 points the beam towards output steering array 1812.

[0133] In some implementations, the non-imaging optical transform approach uses a Fourier element, where an optical element (such as a lens, reflector, or mirror) is placed at the midpoint between the steering elements, designed so that its focal length matches the distance between the optical element and steering element on either side. The Fourier element maps the same input beam orientation angle to the same displacement at the output steering array. This ensures that all input steering elements of input steering array 1808 can hit the center of the output steering array 1812; for example, by pointing the beam along the primary optical axis (sharing the same input beam orientation angle). In some implementations, optical element 1804 is a Fourier element, which counteracts an expanding light beam by contracting the beam. In some implementations, optical element 1804 includes multiple Fourier elements. The beam displacement on output steering array 1812 is controlled by changing the input steering element angle. Thus, all steering elements can uniformly access the full output steering array without constraints on edge elements. Additionally, the Fourier element can be sized to mitigate beam expansion due to propagation, keeping it to a controlled and small beam waist and thus avoid power losses when reflecting off the output steering element.

[0134] In FIG. 18, light from input collimator array 1824 passes through input image transfer elements 1816-1 and input image transfer element 1816-2 and reflects on input steering array 1808, optical element 1804, and output steering array 1812. Light then passes through output image transfer element 1820-1 and 1820-2 before arriving at output collimator array 1828. In some implementations, input collimator array 1824 and output collimator array 1828 are tilted as described with respect to data signal light sources 1704-1 - 1704-4 described with respect to FIG. 17. In some implementations, output collimator array 1828 is tilted in the opposite direction as input collimator array 1824. The OCS will generally have a symmetric layout with a plane of symmetry around the midpoint of the optical path between the input and output steering elements. The output steering elements will route the beam through an image transfer to the output collimator and into the output fiber (or other light detector).

[0135] While this description of the invention has focused on the optical circuit switch application, the image transfer approach should be understood as a broad solution for resolving precision alignment issues when mapping a fiber array to a steering array in a range of applications including optical circuit switching. As a first example, it can be used for noncommunications beam combination when using an array of fiber sources (or other light sources and detectors) as might be desired to generate a long range lidar beam. As a second example, it can be used in reverse when running the steering array as a multiplexed passive sensor with many channels as might be desired when tracking the returns from multiple lidar beams. In all cases the need for alignment between the collimator array and steering array presents a challenge that is simplified by the use of the image transfer approach.CONCLUSION

[0136] The foregoing description is merely illustrative in nature and is in no way intended to limit the disclosure, its application, or uses. The broad teachings of the disclosure can be implemented in a variety of forms. Therefore, while this disclosure includes particular examples, the true scope of the disclosure should not be so limited since other modifications will become apparent upon a study of the drawings, the specification, and the following claims. In the written description and claims, one or more steps within a method may be executed in a different order (or concurrently) without altering the principles of the present disclosure. Similarly, one or more instructions stored in a non-transitory computer-readable medium may be executed in a different order (or concurrently) without altering the principles of the present disclosure. Unless indicated otherwise, numbering or other labeling of instructions or method steps is done for convenient reference, not to indicate a fixed order.

[0137] Further, although each of the embodiments is described above as having certain features, any one or more of those features described with respect to any embodiment of the disclosure can be implemented in and / or combined with features of any of the other embodiments, even if that combination is not explicitly described. In other words, the described embodiments are not mutually exclusive, and permutations of one or more embodiments with one another remain within the scope of this disclosure.

[0138] Spatial and functional relationships between elements (for example, between modules, circuit elements, semiconductor layers, etc.) are described using various terms, including “connected,” “engaged,” “coupled,” “adjacent,” “next to,” “on top of,” “above,” “below,” and “disposed.” Unless explicitly described as being “direct,” when a relationship between first and second elements is described in the above disclosure, that relationship encompasses a direct relationship where no other intervening elements are present between the first and second elements as well as an indirect relationship where one or more intervening elements are present between the first and second elements.

[0139] As noted below, the term “set” generally means a grouping of one or more elements. However, in various implementations a “set” may, in certain circumstances, be the empty set (in other words, the set has zero elements in those circumstances). As an example, a set of searchresults resulting from a query may, depending on the query, be the empty set. In contexts where it is not otherwise clear, the term “non-empty set” can be used to explicitly denote exclusion of the empty set — that is, a non-empty set will always have one or more elements.

[0140] A “subset” of a first set generally includes some of the elements of the first set. In various implementations, a subset of the first set is not necessarily a proper subset: in certain circumstances, the subset may be coextensive with (equal to) the first set (in other words, the subset may include the same elements as the first set). In contexts where it is not otherwise clear, the term “proper subset” can be used to explicitly denote that a subset of the first set must exclude at least one of the elements of the first set. Further, in various implementations, the term “subset” does not necessarily exclude the empty set. As an example, consider a set of candidates that was selected based on first criteria and a subset of the set of candidates that was selected based on second criteria; if no elements of the set of candidates met the second criteria, the subset may be the empty set. In contexts where it is not otherwise clear, the term “non-empty subset” can be used to explicitly denote exclusion of the empty set.

[0141] In the figures, the direction of an arrow, as indicated by the arrowhead, generally demonstrates the flow of information (such as data or instructions) that is of interest to the illustration. For example, when element A and element B exchange a variety of information but information transmitted from element A to element B is relevant to the illustration, the arrow may point from element A to element B. This unidirectional arrow does not imply that no other information is transmitted from element B to element A. Further, for information sent from element A to element B, element B may send requests for, or receipt acknowledgements of, the information to element A.

[0142] In this application, including the definitions below, the term “module” can be replaced with the term “controller” or the term “circuit.” In this application, the term “controller” can be replaced with the term “module.” The term “module” may refer to, be part of, or include: an Application Specific Integrated Circuit (ASIC); a digital, analog, or mixed analog / digital discrete circuit; a digital, analog, or mixed analog / digital integrated circuit; a combinational logic circuit; a field programmable gate array (FPGA); processor hardware (shared, dedicated, or group) that executes code; memory hardware (shared, dedicated, or group) that is coupled with the processor hardware and stores code executed by the processor hardware; other suitable hardware components that provide the described functionality; or a combination of some or all of the above, such as in a system-on-chip.

[0143] The module may include one or more interface circuits. In some examples, the interface circuit(s) may implement wired or wireless interfaces that connect to a local area network (LAN) or a wireless personal area network (WPAN). Examples of a LAN are Institute of Electrical and Electronics Engineers (IEEE) Standard 802.11-2020 (also known as the WIFI wireless networking standard) and IEEE Standard 802.3-2018 (also known as the ETHERNET wired networking standard). Examples of a WPAN are IEEE Standard 802.15.4 (including the ZIGBEE standard from the ZigBee Alliance) and, from the Bluetooth Special Interest Group(SIG), the BLUETOOTH wireless networking standard (including Core Specification versions 3.0, 4.0, 4.1, 4.2, 5.0, and 5.1 from the Bluetooth SIG).

[0144] The module may communicate with other modules using the interface circuit(s). Although the module may be depicted in the present disclosure as logically communicating directly with other modules, in various implementations the module may actually communicate via a communications system. The communications system includes physical and / or virtual networking equipment such as hubs, switches, routers, and gateways. In some implementations, the communications system connects to or traverses a wide area network (WAN) such as the Internet. For example, the communications system may include multiple LANs connected to each other over the Internet or point-to-point leased lines using technologies including Multiprotocol Label Switching (MPLS) and virtual private networks (VPNs).

[0145] In various implementations, the functionality of the module may be distributed among multiple modules that are connected via the communications system. For example, multiple modules may implement the same functionality distributed by a load balancing system. In a further example, the functionality of the module may be split between a server (also known as remote, or cloud) module and a client (or, user) module. For example, the client module may include a native or web application executing on a client device and in network communication with the server module.

[0146] Some or all hardware features of a module may be defined using a language for hardware description, such as IEEE Standard 1364-2005 (commonly called “Verilog”) and IEEE Standard 1076-2008 (commonly called “VHDL”). The hardware description language may be used to manufacture and / or program a hardware circuit. In some implementations, some or all features of a module may be defined by a language, such as IEEE 1666-2005 (commonly called “SystemC”), that encompasses both code, as described below, and hardware description.

[0147] The term code, as used above, may include software, firmware, and / or microcode, and may refer to programs, routines, functions, classes, data structures, and / or objects. Shared processor hardware encompasses a single microprocessor that executes some or all code from multiple modules. Group processor hardware encompasses a microprocessor that, in combination with additional microprocessors, executes some or all code from one or more modules. References to multiple microprocessors encompass multiple microprocessors on discrete dies, multiple microprocessors on a single die, multiple cores of a single microprocessor, multiple threads of a single microprocessor, or a combination of the above.

[0148] The memory hardware may also store data together with or separate from the code. Shared memory hardware encompasses a single memory device that stores some or all code from multiple modules. One example of shared memory hardware may be level 1 cache on or near a microprocessor die, which may store code from multiple modules. Another example of shared memory hardware may be persistent storage, such as a solid state drive (SSD) or magnetic hard disk drive (HDD), which may store code from multiple modules. Group memory hardware encompasses a memory device that, in combination with other memory devices, stores some or all code from one or more modules. One example of group memory hardware is astorage area network (SAN), which may store code of a particular module across multiple physical devices. Another example of group memory hardware is random access memory of each of a set of servers that, in combination, store code of a particular module. The term memory hardware is a subset of the term computer-readable medium.

[0149] The apparatuses and methods described in this application may be partially or fully implemented by a special-purpose computer created by configuring a general-purpose computer to execute one or more particular functions embodied in computer programs. Such apparatuses and methods may be described as computerized or computer-implemented apparatuses and methods. The functional blocks and flowchart elements described above serve as software specifications, which can be translated into the computer programs by the routine work of a skilled technician or programmer.

[0150] The computer programs include processor-executable instructions that are stored on at least one non-transitory computer-readable medium. The computer programs may also include or rely on stored data. The computer programs may encompass a basic input / output system (BIOS) that interacts with hardware of the special-purpose computer, device drivers that interact with particular devices of the special-purpose computer, one or more operating systems, user applications, background services, background applications, etc.

[0151] The computer programs may include: (i) descriptive text to be parsed, such as HTML (hypertext markup language), XML (extensible markup language), or JSON (JavaScript Object Notation), (ii) assembly code, (iii) object code generated from source code by a compiler, (iv) source code for execution by an interpreter, (v) source code for compilation and execution by a just-in-time compiler, etc. As examples only, source code may be written using syntax from languages including C, C++, C#, Objective-C, Swift, Haskell, Go, SQL, R, Lisp, Java®, Fortran, Perl, Pascal, Curl, OCaml, JavaScript®, HTML5 (Hypertext Markup Language 5th revision), Ada, ASP (Active Server Pages), PHP (PHP: Hypertext Preprocessor), Scala, Eiffel, Smalltalk, Erlang, Ruby, Flash®, Visual Basic®, Lua, MATLAB, SIMULINK, and Python®.

[0152] The term non-transitory computer-readable medium does not encompass transitory electrical or electromagnetic signals propagating through a medium (such as on a carrier wave). Non-limiting examples of a non-transitory computer-readable medium are nonvolatile memory circuits (such as a flash memory circuit, an erasable programmable read-only memory circuit, or a mask read-only memory circuit), volatile memory circuits (such as a static random access memory circuit or a dynamic random access memory circuit), magnetic storage media (such as an analog or digital magnetic tape or a hard disk drive), and optical storage media (such as a CD, a DVD, or a Blu-ray Disc).

[0153] The term “set” generally means a grouping of one or more elements. The elements of a set do not necessarily need to have any characteristics in common or otherwise belong together. The phrase “at least one of A, B, and C” should be construed to mean a logical (A OR B OR C), using a non-exclusive logical OR, and should not be construed to mean “at least one of A, at least one of B, and at least one of C.” The phrase “at least one of A, B, or C” should be construed to mean a logical (A OR B OR C), using a non-exclusive logical OR.

Claims

CLAIMS1. A system for directing optical signals for optical networking, the system comprising: a light source; a collimator element corresponding to the light source; a steering array including a steering element corresponding to the light source; and a set of optical transfer elements, wherein: the set of optical transfer elements is positioned between the steering array and the light source, the set of optical transfer elements forms a light image from a first plane positioned at a first location onto a second plane positioned at a second location, the first location coincides with the collimator element, the second location coincides with the steering array, the set of optical transfer elements is defined by at least one of a first characteristic or a second characteristic, the first characteristic is that an optical signal from the light source passes through each element of the set of optical transfer elements at most once, and the second characteristic is that the set of optical transfer elements is a single element that forms the light image from the first plane onto the second plane.

2. The system of claim 1 wherein the set of optical transfer elements directs light from the light source to the steering element.

3. The system of claim 1 wherein a distance between the single element and the light source is based on a focal distance of the single element.

4. The system of claim 1 wherein a focal distance of the collimator element and focal distances of the set of optical transfer elements are equal.

5. The system of claim 1 wherein the set of optical transfer elements includes at least two elements that form the light image from the first plane onto the second plane.

6. The system of claim 5 wherein a focal distance of the at least two elements is equal.

7. The system of claim 5 wherein: a distance between the at least two elements and the light source is based on one or more focal distances associated with the at least two elements, and a distance between the at least two elements is based on one or more focal distances associated with the at least two elements.

8. The system of claim 1 wherein a center axis of the collimator element is offset from a center axis of the light source.

9. The system of claim 8 wherein an offset distance between the center axis of the collimator element and the center axis of the light source is based on a distance of the light source from a center axis of a light source array that includes the light source.

10. The system of claim 1 wherein a center axis of the light source is angularly offset from parallel with respect to a center axis of the collimator element.

11. The system of claim 10 wherein an angle offset between the center axis of the light source and a center axis of the collimator element is based on a distance of the light source from a center axis of a light source array that includes the light source.

12. The system of claim 1 wherein each element of the set of optical transfer elements has a fixed position.

13. The system of claim 1 further comprising an optical element that is physically distinct from the set of optical transfer elements.

14. The system of claim 13 wherein the optical element is a Fourier lens element.

15. The system of claim 1 further comprising a second steering array.

16. The system of claim 15 further comprising: a light detector; a second collimator element corresponding to the light detector; and a second set of optical transfer elements.

17. The system of claim 16, wherein: the second set of optical transfer elements is positioned between the second steering array and the light detector, the second set of optical transfer elements forms a second light image from a third plane positioned at a third location onto a fourth plane positioned at a fourth location, the third location coincides with the second steering array, the fourth location coincides with the second collimator element, the second set of optical transfer elements is defined by at least one of a third characteristic or a fourth characteristic, the third characteristic is that the optical signal from the second steering array passes through each element of the second set of optical transfer elements at most once, and the fourth characteristic is that the second set of optical transfer elements is a single element that forms the second light image from the third plane onto the fourth plane.

18. The system of claim 17 wherein the second set of optical transfer elements is defined by the third characteristic.

19. The system of claim 17 wherein the second set of optical transfer elements is defined by the fourth characteristic.

20. The system of claim 17 wherein the second set of optical transfer elements directs light from a second steering element of the second steering array to the light detector.

21. The system of claim 1 wherein the set of optical transfer elements is defined by the first characteristic.

22. The system of claim 1 wherein the set of optical transfer elements is defined by the second characteristic.

23. The system of claim 1 wherein the single element is a lens.

24. The system of claim 1 wherein the single element is a curved reflector.

25. The system of claim 1 further comprising: a set of light sources that includes the light source and a second light source, wherein the steering element corresponds to the light source and the second light source.

26. The system of claim 25 further comprising a third collimator element that corresponds to the second light source.

27. An optical networking device comprising the system of claim 26.

28. The system of claim 16 wherein: the light source is a first light station that is configured to transmit and detect light simultaneously, and the light detector is a second light station that is configured to transmit and detect light simultaneously.

29. An optical networking device comprising the system of claim 28.

30. An optical networking device comprising: the system of claim 18; a set of light sources including the light source; and a set of light detectors including the light detector.

31. A method for directing optical signals for optical networking, the method comprising: shining an optical signal onto a set of optical transfer elements, wherein: a collimator element directs the optical signal from a light source, the set of optical transfer elements is positioned between a steering array and the light source, the set of optical transfer elements forms a light image from a first plane positioned at a first location onto a second plane positioned at a second location, the first location coincides with the collimator element, the second location coincides with the steering array,the set of optical transfer elements is defined by at least one of a first characteristic or a second characteristic, the first characteristic is that an optical signal from the light source passes through each element of the set of optical transfer elements at most once, and the second characteristic is that the set of optical transfer elements is a single element that forms the light image from the first plane onto the second plane.

32. The method of claim 31 wherein the set of optical transfer elements is defined by the first characteristic.

33. The method of claim 31 wherein the set of optical transfer elements is defined by the second characteristic.

Images