WO2001095009A2

WO2001095009A2 - Method and apparatus for mirror control in an optical cross-connect switch

Info

Publication number: WO2001095009A2
Application number: PCT/US2001/017767
Authority: WO
Inventors: Peter Hunt
Original assignee: Calient Networks, Inc.
Priority date: 2000-06-05
Filing date: 2001-05-31
Publication date: 2001-12-13
Also published as: WO2001095009A3

Abstract

An embodiment of the present invention provides for a method and a system for updating a table of mirror positions to compensate for drifts experienced by the mirrors. The drifts for at least one surrogate mirror in a first mirror are measured. The first mirror array comprises a plurality of working mirrors and surrogate mirrors. The working mirror carries working light in a working path and the surrogate mirror carries test light in a calibration path. The drifts for the working mirrors are estimated using the drifts measured from the surrogate mirrors. The mirror positions of the surrogate mirrors and the working mirrors are updated to compensate for the measured drifts and the estimated drifts respectively. The mirror positions point each mirror in the first mirror array to a plurality of target mirrors in a second mirror array.

Description

METHOD AND APPARATUS FOR MIRROR CONTROL IN AN OPTICAL CROSS-CONNECT SWITCH

FIELD OF THE INVENTION

The present invention relates generally to field of signal processing in an optical network. More specifically, the present invention is directed to a method and an apparatus for maintaining mirror positions in an optical cross-connect switch. BACKGROUND

Gimbaled mirrors are usually used to reflect light. Initially, the mirrors are calibrated so that mirror positions are known for all possible light paths. Over time, properties of the system will be changed. This may cause the mirrors to drift from their positions. The drifts experienced by the mirrors can be attributed to many factors including physical factors as well as electrical factors. For example, the drift caused by the change to the geometry of the switch assembly may include drift associated with a relaxation of stresses in the assembly (and possible aging of parts) and drift due to temperature, barometric pressure, or humidity.

These changes will affect pointing positions of the mirrors. These changes also affect the properties of the mirrors. To accurately position the mirrors, one method is to use extra sensors to find the true position of the mirror. However, this leads to extra complexity, and yet another set of sensors that are also subject to changes of the properties of the system. SUMMARY OF THE INVENTION

An embodiment of the present invention provides for a method and a system for updating a table of mirror positions to compensate for drifts experienced by the mirrors. The drifts for at least one surrogate mirror in a first mirror are measured. The first mirror array comprises a plurality of working mirrors and surrogate mirrors. The working mirror carries working light in a working path and the surrogate mirror carries test light in a calibration path. The drifts for the working mirrors are estimated using the drifts measured from the surrogate mirrors. The mirror positions of the surrogate mirrors and the working mirrors are updated to compensate for the measured drifts and the estimated drifts respectively. The mirror positions point each mirror in the first mirror array to a plurality of target mirrors in a second mirror array. BRIEF DESCRIPTION OF THE DRAWINGS

The present invention is illustrated by way of example in the following drawings in which like references indicate similar elements. The following drawings disclose various embodiments of the present invention for purposes of illustration only and are not intended to limit the scope of the invention.

Figure 1 illustrates an exemplary optical switching system for practicing the invention.

Figure 2 illustrates a top view of one embodiment of a micro- electro-mechanical system (MEMS) mirror device for a MEMS mirror array.

Figure 3 illustrates an exemplary mirror array having nine mirrors.

Figure 4 is an exemplary a bell curve illustrating the different levels optical power output.

Figure 5A is an exemplary flow diagram illustrating a drift identification process.

Figure 5B is an exemplary flow diagram illustrating one embodiment of a process of providing the voltages necessary to point a mirror to the required X, Y coordinates. Figure 6 illustrates an embodiment of a computer-readable medium.

Figure 7 is an exemplary block diagram illustrating one embodiment of an optical cross-connect switch. DETAILED DESCRIPTION

In one embodiment, an adaptive mirror positioning method and system that keep all mirrors in their correct positions are disclosed. Changes in mirror position are recorded to update mirror position tables on a continuous basis.

The exponential growth of the Internet and Intranet traffic has created tremendous demands upon the bandwidth transport capacity of Service Provider networks in recent years. Advances in fiber throughput and optical transmission technologies have enabled the Service Providers to increase backbone network capacity. One component of the optical transmission technologies is the optical cross-connect switch.

The optical cross-connect switch is used to connect high-capacity incoming fiber optic communication channels to outgoing fiber optic communication channels. Traditionally, the optical cross-connect switch is not truly optical because the incoming optical signals are converted from optical-to-electrical-to-optical (OEO) to switch wavelengths. This is expensive and requiring vast amounts of synchronous optical network (SONET) de-multiplexors /multiplexors. In a true optical cross-connect switch, multiple microscopic mirrors on two planes are used to pass the optical signals. A beam of light going to a mirror in a first plane is directed to another mirror in a second plane. This allows optical signals from one incoming fiber to be directed to an outgoing fiber. An optical connection is made when the mirror in the first plane, illuminated by the incoming fiber, directs the light to the mirror on the second plane, which directs this light to the outgoing fiber. To maintain this connection, the mirrors on the first plane and on the second plane have to be perfectly aligned. Mirror positions are important to maintain the alignment.

Figure 1 illustrates an exemplary optical switching system 100 for practicing the invention. For example, optical switching system 100 may represent a 3-dimensional optical switching system. A 3-dimensional optical switching system allows for optical coupling between input fibers and output fibers in different planes using lens arrays and mirror arrays. The lens arrays and mirror arrays provide proper angle and position of light beams traveling from input fibers to output fibers. That is, a light beam must leave and enter a fiber in a direct path. In the following description of Figure 1, mirror arrays are described as micro-electromechanical-system (MEMS) mirror arrays. The MEMS mirror arrays are arrays of microscopic mirror devices formed with a substrate. The mirror devices can redirect beams of light to varying positions.

Referring to Figure 1, optical switching system 100 includes input fiber array 150, first lens array 130A, first MEMS mirror array 120A, second MEMS mirror array 120B, second lens array 130B, and output fiber array 160. The first MEMS mirror array 120A is also referred to as first mirror plane 120A. The second MEMS mirror array 120B is also referred to as second mirror plane 120B or target mirror plane 120B.

Input fiber array 150 provides a plurality of optical fibers 140 for transmitting light to first lens array 130A. First lens array 130A includes a plurality of optical lenses (not shown), which are used to direct collimated beams of light from input fiber array 150 to individual MEMS mirror devices on first MEMS mirror array 120A. First MEMS mirror array 120A includes a plurality of electrically addressable MEMS mirror devices such as, for example, MEMS mirror device 110. MEMS mirror device 110 may be a gimbaled mirror device having a cross shape with an elliptical mirror component. Alternatively, MEMS mirror device 110 may be a gimbaled mirror device having different shapes and configurations. For example, the mirror component may be a circular or rectangular mirror component. First MEMS mirror array 120A may have an array of MEMS mirror device 110 that can pivot a mirror component to redirect or reflect light beams to varying MEMS mirror devices on second MEMS mirror array 120B. Second MEMS mirror array 120B may also include a plurality of MEMS mirror devices such as MEMS mirror device 110, which are used to redirect and reflect light beams to varying lenses on second lens array 130B. Second lens array 130B accepts collimated light beams from second MEMS mirror array 120B and focuses the light beams to individual output fibers 170 of output fiber array 160.

Optical switching system 100 allows light beams from any input fiber 140 of input fiber array 150 to be redirected to any output fiber 170 of output fiber array 160. For example, a light beam following the path "A" is outputted from one input fiber and is redirected using first and second lens arrays 130A and 130B and first and second MEMS mirror arrays 120A and 120B to a different output fiber. The lens and MEMS mirror arrays may also be used in scanning systems, printing systems, display systems, and other systems that require redirecting beams of light.

Figure 2 illustrates a top view of one embodiment of a MEMS mirror device 110 for a MEMS mirror array. MEMS mirror device 110 illustrates an elliptical mirror component 101. MEMS mirror device 100 may have different shaped mirror components such as, for example, a rectangular or circular mirror component. MEMS mirror device 110 includes a frame 105 supporting mirror component 101 and other suspended components. Gimbals 102A, 102B, 102C and 102D allowing mirror component 101 to pivot. Gimbals 102A and 102B reside in the same axis. Gimbals 102C and 102D reside in the same axis. The axis of gimbals 102A and 102B is perpendicular to the axis of gimbals 102C and 102D. Gimbals 102A and 102B allow the mirror component 101 to rotate in the Y direction. Gimbals 102C and 102D allow the frame 105 to rotate in the X direction.

Mirror component 101 is capable of having an angular range of motion with respect to the Y axis and the X axis along gimbals 102A, 102B, 102C and 102D caused by electrodes 205 and 210 respectively. Thus, the arrangement of the MEMS mirror device 110 allows the mirror component 101 to direct light in the X or Y directions. Frame 105 provides support for mirror component 101.

Although not shown in Figure 2, MEMS mirror device 110 may also include ground planes and wiring lines, which have not been shown for purposes of clarity.

An optical connection is made when a mirror in the first mirror array 120A, illuminated by an input fiber 140 of the input fiber array 150, directs the light to a mirror in the second mirror array 120B, which directs this light to an output fiber 170 of the output fiber array 160. To establish the connection, the mirror in the first mirror array 120A and the mirror in the second mirror array 120B are moved to point to each other. Each of the mirrors in the first mirror array 120A forms a path with each of the mirrors in the second mirror array 120B. The path may be a working path carrying working light, a standby path serving as a backup for the working path, or a calibration path serving as a test path carrying test light for mirror calibration purposes. In one embodiment, the first mirror array 120A and the second mirror array 120B each contain 256 mirrors redirecting light from 256 input fibers 140 to 256 output fibers 170. Figure 3 illustrates an exemplary mirror array having nine mirrors. Each mirror 300 in the first mirror array 120A may be moved to point to any of the nine mirrors on the second mirror array 120B. The mirror pointed to is also referred to as target mirror. Figure 3 shows an exemplary working path formed between the mirror 300 and the target mirror 306. Initially, the mirror 300 in the first mirror array is calibrated and its positions to each of the nine target mirrors are kept in a mirror position table. When the mirror 300 is to be moved and form a different path with, for example, the target mirror 302, the information required to move the mirror 300 to point to the target mirror 302 is retrieved from the mirror position table. The target mirror 302 also has a mirror position table to help moving it to point to the mirror 300. When the two mirrors 300 and 302 are perfectly aligned, they are said to be in optimal positions and the light path formed by the two mirrors 300 and 302 provides peak or optimal optical power output with zero or minimum optical loss.

Figure 4 is an exemplary a bell curve illustrating the different levels optical power output. At the peak 405 of the curve, the optical output power is at its maximum and there is no optical loss. Over time the properties of the mirrors in the mirror arrays 120A and 120B change resulting in mirror drifts and mirror misalignment. Since the mirrors do not have any feedback mechanism to provide current mirror position, it is unknown how much the drifts affect the mirror position. The drifts may cause the mirrors to not accurately point at each other, and, as a result, create pointing errors in the mirrors. When there are pointing errors, the light may not be correctly directed from the incoming lens to the outgoing lens resulting in optical loss. Referring to Figure 4, optical loss reduces the optical power output and lowers it to a level 410, 412, 414 below the optimal level 405.

Generally there are two cross-connect switches in the system. A first switch serves as a working switch and a second switch serves as a protect or standby switch. The first switch carries working light (i.e., user's data) in working paths. The second switch carries non-working light or test light in standby path. In an event of a failure to the first switch, the second switch resumes the switching activities of the first switch. Alternatively, a mixture of the first switch and the second switch may be used as the working switch and the standby switch.

All mirrors are maintained in paths. In one embodiment, spare mirrors (i.e., mirrors not in the working paths or the standby paths) are used to determine the drift data so that accurate mirror position information can be obtained. The spare mirror is also referred to as available mirror. These spare mirrors carry test light and are used in calibration paths. The spare mirrors are also referred to as surrogate mirrors because they provide calibration or drift data to the working mirrors. The surrogate mirrors may also provide drift data for the standby mirrors. The test light is injected by the system into the standby paths and the calibration paths. Since the surrogate mirrors are spare mirrors (available), the calibration paths may be broken to form new working paths whenever necessary.

When a new working path is to be formed, the mirror on the first mirror array 120A and the mirror on the second mirror array 120B are moved to point to each other to form the new connection. In one embodiment, a technique is used to calculate the optimal or peak position of the mirrors in the new path. This technique is referred to as a hill climbing technique. When the light is detected through the path after initial set up, small changes are made to the mirror positions and power readings are taken. The change in power gain is observed from the power readings and the optimal mirror position is calculated using a mathematical approximation. The optimal mirror position is found when the power reading is at its optimal or peak level. The hill climbing technique helps keep the paths focused.

The mirrors in the working paths are frequently focused to keep them aligned for optimum performance. Drift for a particular mirror, at its current pointing position, is updated whenever a focus operation is performed on the connection that includes the mirror. Thus while the working path remains active, the system knows that the mirrors in the working path are aligned. The mirrors in the standby paths are also frequently focused but may occasionally be used for calibration. When a mirror has been in a working path for a long time, the drift from the mirror position is available. However, the drift is known only for this position. While the working path carries the working light, there is no opportunity to test the working mirrors at any position other than the position they have been in. The working path may last for years, and it is unknown how much the working mirror drifts in the other positions. Because of this unknown, when the working path is to be changed, an estimate of the correct path to the new target mirror needs to be done. In one embodiment, the surrogate mirrors can be used at several possible connections to estimate drifts data experienced by the working mirrors in the other positions.

In one embodiment, the drift data includes a common drift, a voltage drift, a torsion drift and a zero point drift. The common drift includes drifts caused by changes to the geometry of the switch assembly. These drifts impact all mirrors in the mirror array. However, the effect on an individual mirror is dependent on the mirror's position in its mirror array. In one embodiment, the common drift consists of two components, permanent drift and cyclic drift. Permanent drift is associated with a relaxation of stresses in the assembly and possibly is associated with the aging of parts. The cyclic drift is associated with temperature, barometric pressure, humidity, etc. The cyclic drift varies back and forth. The common drift is approximately a linear error across the second mirror array. An assessment of the drift in all the mirrors will provide a reasonable estimate of drifts due to the changes to the geometry of the switch assembly. When there is a common drift experienced by all the surrogate mirrors, the drift values for the working mirrors can be interpolated from those surrogate mirrors. It would be appreciated to note that there are times when the drifts experienced by the surrogate due to the changes to the geometry of the switch assembly are so different that no common drift can be detected. The common drift is also referred to as a global drift.

The voltage drift includes drifts caused by changes to the voltages (e.g., voltage gain) applied to the deflection electrodes and can be compensated using a voltage scale factor. The voltage drift creates a nonlinear error in mirror position. The voltage drift arises in zero voltage and high voltage situations. Zero voltage stresses some of the amplifier components, while high voltage stress other components. The torsion drift is also referred to as a position drift and includes change in any parameter that is linear with mirror rotation. The zero point drift includes changes in the position of the mirror when there is no voltage applied. The voltage drift, torsion drift and zero point drift are local to a particular mirror or one of the amplifiers associated with that mirror. These drifts are also referred to as local drifts. Several mirror positions are necessary to confirm the strength of the various effects of the local drifts. It will be apparent to one skilled in the art that there may be drifts caused by other factors such as, for example, drift caused by sudden external impact to the system, that can be used to estimate the drift data without departing from the scope of the present invention.

In one embodiment, when a move to a new path is to be made, the information used to calculate the drift on the new target position includes the geometry changes that are common to all the mirrors subtracted from the observed mirror drift that occurred during the life of the connection. An assumption is made that the ratio of deterioration due to the other factors are governed by the same laws, and decay at the same rates as other mirrors on the same mirror plane. A continuous assessment of the cyclic drift and permanent drift is maintained. For example, the cyclic drift is assessed approximately every 10 minutes for each path (approx. one mirror calibration per 2 seconds), and the permanent drift is assessed every 24 hours (annual cyclic drift patterns will be considered as permanent). In one embodiment, the assessment is performed based on the drift on the established paths such as, for example, the calibration paths. An available mirror will go through a series of tests, each designed to evaluate the degree of each type of drift, so that the deterioration ratios can be computed.

Figure 5A is an exemplary flow diagram illustrating a drift identification process. At block 505, all common or global drifts are extracted from all the mirrors. As mentioned earlier, the common drift includes drifts caused by changes to the geometry of the switch assembly. The drift due to the changes to the geometry of the switch assembly is approximately linear error across the mirror array. For a particular mirror in the mirror array, the pointing errors at any point will be derived from a linear interpolation of drift values measured at all the mirrors.

At block 510, local drifts are extracted from the surrogate mirrors. To determine the local drifts, multiple mirror positions are necessary to confirm the strength of the various effects of the voltage drift and position drift. In one embodiment, the surrogate mirrors are changed to a different calibration path on a regular basis such as, for example, every 24 hours. This calibration path change allows the mirror drift data (i.e., voltage drift and position drift) to be collected for each surrogate mirror over a whole domain of movement. This way each surrogate mirror can be exercised to periodically point to a different target mirror to collect the particular components of the drift data.

In one embodiment, each surrogate mirror is taken through a sequence of calibration paths, changing to a new path at every cycle such as, for example, 24 hours. The sequence of calibration paths helps to isolate the voltage drift, the torsion drift and the zero point drift. The sequence of calibration paths requires each surrogate mirror to make a minimum of six well-placed connections. For example, the sequence of paths may include the four corner mirrors, the middle mirror, the mirrors between the middle mirror and the four corner mirrors, etc. For each path in the sequence of paths, drift components are extracted and used to update drift positions in a drift table for that corresponding surrogate mirror. The drifts are calculated by using the hill climbing technique discussed earlier and by interpolating the extracted data.

In block 515, the global and the local drifts are used to estimate drifts for the working mirrors. The estimation is based on the average of all the drifts experienced by the surrogate mirrors. At block 520, the estimated drifts are applied to the drift table for the working mirror.

In one embodiment, the mirror position table is a table of position to deflection values - a two dimensional control surface that provides the voltages required by each mirror to point to a particular target mirror. The position of the target mirror in the second mirror array 120B is identified using Cartesian coordinates (e.g., X and Y values). Voltage is applied to four electrodes attached to the mirror assembly (frame axis 102C and 102D, and mirror axis 102A and 102B to move or deflect the sending mirror to a position, using electrostatic attraction, so that it can direct a light ray to a specified target mirror (identified by the X, Y Cartesian coordinates) on the second mirror array 120B. The system is symmetrical, in that the mirror 302 (Figure 3) can be seen as aiming at the mirror 300 to receive the light ray from the mirror 300.

The mirror position table provides a control surface for each mirror by specifying the voltages required to make the mirror reflect the light beam to (from) any mirror in the second mirror array 120B. In one embodiment, the mirror position table is addressed by the integer values of the target mirror Cartesian coordinates and provides the voltages that are needed to point to the particular target mirror. The mirror position table may also provide the voltages to point midway between the target mirrors (interpolated values) and to positions outside the second mirror array (extrapolated values). In another embodiment, the voltages that are needed to point to the particular target mirror are specified as two signed 16 bit integer voltages. One voltage is for the two frame electrodes and the other voltage is for the two mirror electrodes. It would be appreciated to note that two electrodes are controlled at one time. The other two electrodes are set to zero. When the mirror is not held tightly by the hinges, an offset voltage may be applied to all the electrodes sufficient enough to tighten up the assembly. In this situation, there would always be a nominal voltage applied to at least one of the electrodes, instead of the usual zero voltage. The negative voltages are seen as a positive voltage on one electrode, positive voltages on the other electrode.

The drift table provides the pointing error for each mirror position. The drift data in the drift table is used to update the mirror position table to adjust the pointing direction of the mirrors. This update, for example, may be made periodically such as, for example, every 24 hours. This way, the mirror position table is dynamically updated on a continuous basis to provide accurate mirror positions. In one embodiment, there is a drift table for each mirror. As the drift data is accumulated for each mirror, it is added to the drift table. Some of the entries in the drift table are from direct measurement of the surrogate mirrors in the calibration paths. The other entries are based on prediction because the corresponding mirrors are in the working paths. In one embodiment, the drift table is addressed using the target mirror Cartesian coordinates, the same way the mirror position table is addressed. In another embodiment, the drift data (e.g., deltas) in X and Y are signed fixed point values of the same scale as the X, Y Cartesian values used to access the drift table. In another embodiment, for each drift data there is a status word defining whether this position was estimated, or calibrated. For example, the status word may include the values to indicate that the drift data is from new calibration, old calibration, estimated or null. The drift table is table assumes that changes (drift and error) to the mirror position table are seen as position changes, rather than voltage changes. Figure 5B is an exemplary flow diagram illustrating one embodiment of a process of providing the voltages necessary to point a mirror to the required X, Y coordinates. In block 525, the local drift for the target mirror is retrieved from the drift table. The integral part of the X and Y coordinates is used to access the drift values from the drift table for the four points surrounding the X, Y position. The fractional parts of the coordinates are used to perform a bilinear interpolation of the drift values.

In block 530, the drift is added to the target mirror coordinates. The drift values DX, and DY are added to the original X,Y, to make the corrected values X', and Y'. In block 535, using the corrected values X', and Y', the V_x and V_γ bipolar voltages are retrieved. The integral part of the X' and Y' coordinates are used to access the two voltage values from the MV table for the four points surrounding the X, Y position. The fractional parts of the X' and Y' coordinates are used to perform a bilinear interpolation of the two bipolar voltages.

In block 540, the bi-polar voltages are converted to to the 4 deflector voltages. Positive voltages are scaled using the positive voltage gain value for the particular axis, and are written to the positive electrode. Zero is written to the negative electrode. Negative voltages are inverted, scaled with the negative voltage gain value for the particular axis, and are written to the negative electrode. Zero is written to the positive electrode. In block 545, the electrode voltages are written to the voltage control buffer and the appropriate voltages are applied to the electrodes to move the mirror.

The bilinear interpolation mentioned above is now discussed in more detail using the following example. For a point having the target mirror coordinates X, Y, the two drift values are. D_X(X,Y) and D_y(X,Y). Both X, Y are 16 bits. The top 8 bits indicates a table entry (signed). The bottom 8 bits represent a fractional part. To interpolate a function D_X(X,Y), split into integer part X', and fractional part »X

D_X(X,Y) = (D_X(X',T+1) *-Y + D_x(X',Y) *(1-»Y))(1-»X)

+ (D_X(X'+1N'+1)*-»Y + D_X(X'+1,Y')*(1-Y))-»X To interpolate a function D_y(X,Y), the approach is the same. In one embodiment, the two interpolations can be done in parallel by using low and high halfwords for the D_x andD_y.

Alternatively, the interpolation for the function D_X(X,Y), can be performed as:

D_x(X,Y) = ( (D (X'+1X+1) - OJ(X'+1X))*^»Υ + O_X(X'+1,Y)

- (D_x(X',r+l) - D_X(X',Y'))*-»Y - D_x(X',Y') )-X

+ (D_X(X'N'+1) - D_X(X',T))*-Y + D_x(X'N')) In another embodiment, when memory is cheap, and background time is available, the delta can be stored in the table.

D_X(XN) = (D_x.(X',r) -Y + D_X(X'N') )(1-.X)

+ ((Ό X'+IX) *»Y + D_x(X'+ι,r) )-x

To create the voltage, add D_X(X,Y) to X, D_y(X,Y) to Y. Then use the new X and Y values to access voltages V_X(X,Y) & N_y(X,Y) in the same way as discussed above. In one embodiment, 16 bits are used for voltages and the two voltages are paired into one 32 bit word. The x and y components can be processed in parallel and used to make the four scaled electrode voltages.

Figure 6 illustrates an embodiment of a computer-readable medium 600 containing various sets of instructions, code sequences, configuration information, and other data used by a computer or other processing device. The embodiment illustrated in Figure 6 is suitable for use with the mirror positioning method described above. The various information stored on medium 600 is used to perform various data processing operations. Computer-readable medium 600 is also referred to as a processor-readable medium. Computer-readable medium 600 can be any type of magnetic, optical, or electrical storage medium including a diskette, magnetic tape, CD-ROM, memory device, or other storage medium.

Computer-readable medium 600 includes interface code 602 that controls the flow of information between various devices or components in the computer system. Interface code 602 may control the transfer of information within a device (e.g., between the processor and a memory device), or between an input/ output port and a storage device. Additionally, interface code 602 may control the transfer of information from one device to another.

Computer-readable medium 600 also includes the code to perform the determination of the different drift components including the code to determine the common drift 604, the voltage drift 606 and the position drift 608. The computer-readable medium 600 may also contain the code that maintain the drift table 610 and the code that dynamically update the mirror position table 612.

Figure 7 is an exemplary block diagram illustrating one embodiment of an optical cross-connect switch 700. Switch 700 includes fiber and mirror array assembly 705, which is coupled to control circuitry 710. Fiber and mirror assembly 705 receives as inputs an array of fiber optic cables 715. An array 720 of fiber optic cables is provided as an output from fiber and mirror array assembly 705.

Control circuitry 710 includes digital signal processors (DSPs) 725 and 730. DSP 725 is coupled to fiber and mirror array assembly 705 via lines 735. DSP 730 is coupled to fiber and mirror array assembly 705 via lines 740. For one embodiment of the invention, DSPs 725 and 730 are each a TMS320C6211 fixed-point digital signal processor supplied by Texas Instruments of Dallas, Texas. Memory 745 is coupled to DSP 725. Memory 750 is coupled to DSP 730. DSP 725 controls the first array of mirrors (not shown). DSP 730 controls the second array of mirrors (not shown). The first array of mirrors and the second array of mirrors are in the fiber and mirror array assembly 705. DSP 725 controls the voltages and currents sent to position the mirrors within the first mirror array. DSP 730 controls the voltages and currents sent to position the mirrors within the second mirror array. Each DSP 725 and 730 receives optical power values from optical detectors (not shown) residing within the fiber and mirror array assembly 705.

DSP 725 is able to generate a mathematical approximation of the relationship of optical power versus mirror or reflector position based on a set of optical power values received by DSP 725 for the first mirror array. DSP 725 is able to choose a position of the movable mirror or reflector of a mirror within the first mirror array based on a point in the mathematical approximation of the relationship of optical power versus reflector position. DSP 725 is able to do this running code stored within memory 745 and DSP 725.

Likewise, DSP 730 can generate a mathematical approximation of a relationship of optical power versus mirror or reflector position based on a set of optical power values with respect to a mirror within the second mirror array. Likewise, DSP 730 can choose a position of a movable mirror or reflector within the second mirror array based on a point in the mathematical approximation of the relationship of optical power versus reflector position. DSP 730 runs code stored within the memory 750 and DSP 730.

DSP 725 is able to run the mathematical approximation used to choose the optimal mirror position in parallel for all the mirrors within the first mirror array. Likewise, DSP 730 is able to generate the mathematical approximation for choosing the optimal mirror position in parallel for all the mirrors within the second mirror array.

Processor 755 is coupled to DSP 725 via lines 760 and to DSP 730 via lines 765. Processor 755 is also coupled to fiber and mirror array assembly 705 via lines 770. Processor 755 is also coupled to volatile memory 775 and non-volatile memory 780 via bus 785. Processor 755 is also coupled to memory 745 via bus 785 and to memory 750 via bus 790.

Processor 755 oversees the operations of DSP 725 and 730. In one embodiment, processor 755 is a PowerPC. Processor 755 oversees the operation of maintaining the optimal signal path through fiber and mirror array assembly 705. Processor 755 runs code that ensures that the optimal mirror position is chosen in one mirror array for a particular light path before the optimal mirror position is chosen in the second mirror array in that same light path. In other words, processor 755 oversees the shifting of processing from DSP 725 to DSP 730 for a particular light path optimization procedure. One mirror in a light path is optimized before the second mirror in the light path is then optimized. Processor 755 allows the shifting of optimization from one mirror array to the other. It should be kept in mind, however, that the processing for the mirrors on a particular mirror array is done in parallel, so there is parallel global shifting between a first mirror array and a second mirror array that is coordinated by processor 755. From the above description and drawings, it will be understood by those of ordinary skill in the art that the particular embodiments shown and described are for purposes of illustration only and are not intended to limit the scope of the invention. Those of ordinary skill in the art will recognize that the invention may be embodied in other specific forms without departing from its spirit or essential characteristics. References to details of particular embodiments are not intended to limit the scope of the claims.

Claims

CLAIMSWhat is claimed is:

1. A method, comprising: measuring one or more drifts for at least one surrogate mirror in a first mirror array comprising of a plurality of working mirrors and surrogate mirrors, wherein the working mirror carries working light in a working path and the surrogate mirror carry test light in a calibration path;

estimating drifts for the working mirrors to form estimated drifts using the drifts measured from the surrogate mirrors; and

updating mirror positions of the working mirrors to compensate for the estimated drifts respectively, the mirror positions pointing each mirror in the first mirror array to a plurality of target mirrors in a second mirror array.

2. The method of claim 1, wherein the drifts comprise drift due to geometry and assembly change, drift due to voltage change and drift due to position change, and wherein all mirrors in the first mirror array and in the second mirror array are connected in the working paths and the calibration paths.

3. The method of claim 2, wherein the drift due to the geometry and assembly change is determined using all the surrogate mirrors, wherein when there is a common change in all the surrogate mirrors due to the geometry and assembly change, a common drift is extracted.

4. The method of claim 3, wherein the drift due to the geometry and assembly change is measured by periodically focusing the surrogate mirrors in the calibration paths.

5. The method of claim 2, wherein the drift due to the voltage change and the position change are determined using at least one surrogate mirror in a predetermined sequence of calibration paths comprising of selected target mirrors in the second mirror array.

6. The method of claim 5, wherein the selected target mirrors in the second mirror array are positioned at different locations in the second mirror array to provide representative samplings of the drift due to the voltage change and the position change of all mirrors in the second mirror array.

7. The method of claim 5, wherein using at least one surrogate mirror in the predetermined sequence of calibration paths comprises:

changing the surrogate mirror to a different calibration path at a predetermined time period to enable the drift data to be measured for each different calibration path, wherein the different calibration path is formed by pointing the surrogate mirror at a different target mirror not in the working path;

deriving drift components at each different calibration path; and generating an estimate of drifts for a complete domain of movement for the working mirror.

8. The method of clai 7, wherein the predetermined sequence of calibration paths comprises paths from the surrogate mirror to four corners mirrors in the second mirror array, and wherein the drift components comprise cyclic drift, voltage drift, zero voltage drift, and linear drift.

9. The method of claim 1, wherein estimating drifts for the working mirrors using the drifts measured from the surrogate mirrors comprise using a linear interpolation of the drifts measured from the surrogate mirrors.

10. The method of claim 9, wherein the mirror positions of the working mirrors are updated by combining the estimated drifts to the mirror positions of the working mirrors.

11. The method of claim 1, further comprising using a linear interpolation of the updated mirror position of the working mirror to retrieve voltages necessary to move the working mirror to point to a specified position.

12. A computer readable medium having stored thereon sequences of instructions which are executable by a digital processing system, and which, when executed by the digital processing system, cause the system to perform a method comprising: measuring one or more drifts for at least one surrogate mirror in a first mirror array comprising of a plurality of working mirrors and surrogate mirrors, wherein the working mirror carries working light in a working path and the surrogate mirror carry test light in a calibration path;

13. The computer readable medium of claim 12, wherein the drifts comprise drift due to geometry and assembly change, drift due to voltage change and drift due to position change, and wherein all mirrors in the first mirror array and in the second mirror array are connected in the working paths and the calibration paths.

14. The computer readable medium of claim 13, wherein the drift due to the geometry and assembly change is determined using all the surrogate mirrors, wherein when there is a common change in all the surrogate mirrors due to the geometry and assembly change, a common drift is extracted.

15. The computer readable medium of claim 14, wherein the drift due to the geometry and assembly change is measured by periodically focusing the surrogate mirrors in the calibration paths.

16. The computer readable medium of claim 13, wherein the drift due to the voltage change and the position change are determined using at least one surrogate mirror in a predetermined sequence of calibration paths comprising of selected target mirrors in the second mirror array.

17. The computer readable medium of claim 16, wherein the selected target mirrors in the second mirror array are positioned at different locations in the second mirror array to provide representative samplings of the drift due to the voltage change and the position change of all mirrors in the second mirror array.

18. The computer readable medium of claim 16, wherein using at least one surrogate mirror in the predetermined sequence of calibration paths comprises:

extracting the drift data measured for each calibration path; and

storing the drift data corresponding to the surrogate mirror for each different calibration path.

19. The computer readable medium of claim 18, wherein the predetermined sequence of calibration paths comprises paths from the surrogate mirror to four corners mirrors in the second mirror array.

20. The computer readable medium of claim 12, wherein estimating drifts for the working mirrors using the drifts measured from the surrogate mirrors comprise using a linear interpolation of the drifts measured from the surrogate mirrors.

21. The computer readable medium of claim 20, wherein the mirror positions of the working mirrors are updated by combining the estimated drifts to the mirror positions of the working mirrors.

22. A system comprising:

a first logic configured to calculate drifts for at least one surrogate mirror in a first mirror array, the surrogate mirror carrying a test light in a calibration path to a first target mirror in a second mirror array;

a second logic configured to use the drifts for the surrogate mirror to predict drifts for one or more working mirrors carrying a working light in a working path to a second target mirror in the second mirror array; and

a third logic configured to combine the predicted drifts for the working mirrors with a mirror position for the working mirror, the mirror position being in a mirror position table, the combined mirror position and the predicted drifts for the working mirror enabling the working mirror to correctly point to a third target mirror to form a new working path.

23. The system of claim 22, wherein the a first logic configured to calculate the drifts for the surrogate mirrors comprise: logic to calculate drift due to geometry and assembly change, logic to calculate drift due to voltage change and linear change.

24. The system of claim 23, wherein the logic to calculate drift due to the geometry and assembly change comprises logic to measure the geometry and assembly change of all surrogate mirrors to extract a common geometry and assembly change experienced by all the surrogate mirrors.

25. The system of claim 23, wherein the logic to calculate drift due to the voltage change and linear change comprises:

logic to periodically change the surrogate mirror to a different calibration path following a predetermined calibration path sequence to enable the drift data to be measured for each different calibration path; and

logic to extract the drift data measured for each different calibration path.

26. The syste of claim 25, wherein the predetermined calibration path sequence is selected to enable the drift data measured at each different calibration path to be representative for all possible calibration paths.

27. The system of claim 22, wherein the second logic configured to use the drifts for the surrogate mirror to predict drifts for the working mirrors comprises logic to interpolate drifts for the surrogate mirrors to predict drifts for the working mirror.

28. The system of claim 22, wherein the third logic configured to combine the predicted drifts for the working mirrors with a mirror position for the working mirror comprises logic to interpolate voltages required to move the working mirror to point to the third target mirror.