WAVELENGTH FILTER USING LIGHT COUPLER AND OPTICAL RESONATOR
RELATED APPLICATIONS
The present application claims the benefit of U. S. Provisional Patent Application Serial Number 60/201,957, entitled "WAVELENGTH FILTER USING LIGHT COUPLER AND OPTICAL RESONATOR," filed May 5, 2000 and U.S. Provisional Patent Application Serial Number 60/200,883, entitled "OPTICAL DEVICE HAVING POLARIZATION ELEMENTS," filed May 1 , 2000, which are hereby incorporated herein by reference. The present application is also related to co-pending U.S. Provisional Application Serial Number 60/201,648 entitled "WIDE TRANSMISSION OPTICAL COMB FILTER WITH WIDE PASS BAND AND WIDE STOP BAND," filed May 1, 2000; U.S. Patent Application Serial Number 09/614,782 entitled "WIDE TRANSMISSION OPTICAL COMB FILTER WITH WIDE PASS BAND AND WIDE STOP BAND," filed July 12, 2000; U.S. Provisional Application Serial Number 60/200,969 entitled "LOW LOSS ULTRA STABLE FABRY-PEROT ETALON," filed May 1, 2000; U.S. Patent Application Serial Number 09/642,985 entitled "LOW LOSS ULTRA STABLE FABRY-PEROT ETALON," filed August 21, 2000; and U.S. Patent Application Serial Number 09/643,559 entitled "OPTICAL DEVICE HAVING
POLARIZATION ELEMENTS," filed August 21, 2000, which are hereby incorporated herein by reference.
TECHNICAL FIELD
The present application relates in general to optical communications, and in specific a wavelength filter in wavelength division multiplex communications.
BACKGROUND
Optical wavelength division multiplexing has gradually become the standard backbone network for fiber optic communication systems. WDM systems employ signals consisting of a number of different wavelength optical signals, known as carrier signals or channels, to transmit information over optical fibers. Each carrier signal is modulated by one or more information signals. As a result, a significant number of information signals may be transmitted over a single optical fiber using WDM technology. These optical signals are repeatedly amplified by erbium-doped fiber amplifiers (EDFA) along the network to compensate for transmission losses. The amplified signals reach the receiving end and are detected using WDM filters followed by photo receivers. Fiber optic communications networks are typically arranged with a plurality of terminals in any of a number of topological configurations. The simplest configuration is two terminals communicating data over an optical link. This can be extended to a daisy- chain configuration in which three or more terminals are connected in series by a plurality of optical links. Ring configurations are also used, as well as other two-dimensional mesh networks. In each case, the optical link between two terminals typically includes a plurality of optical fibers for bidirectional communications, to provide redundancy in the event of a fault in one or more of the optical fibers, and for future capacity.
Despite the substantially higher fiber bandwidth utilization provided by WDM technology, a number of serious problems must be overcome, for example, multiplexing, de-multiplexing, and routing optical signals, if these systems are to become commercially viable. The addition of the wavelength domain increases the complexity for network management because processing now involves both filtering and routing. Multiplexing involves the process of combining multiple channels (each defined by its own frequency spectrum) into a single WDM signal. De-multiplexing is the opposite process in which a single WDM signal is decomposed into individual channels or sets of channels. The individual channels are spatially separated and coupled to specific output ports. Routing differs from de-multiplexing in that a router spatially separates the input optical channels to
output ports and permutes these channels according to control signals to create a desired coupling between an input channel and an output port.
Note that each carrier has the potential to carry gigabits of information per second. Current technology allows for about forty channels or optical carriers, each of a slightly different wavelength, to travel on a single-mode fiber using a single WDM signal. The operating bands are limited by the EDFA amplifier (C) band, thus the increase in the number of channels has been accomplished by shrinking the spacing between the channels, and by adding new bands. The current standard is 50 andlOO GHz between optical channels, whereas older standards were 200 and 400 GHz spacings. Another characteristic of the WDM signal is the modulation rate. As the modulation rate is increased, more data can be carried. Current technology allows for a modulation rate of 10 Gigabits per second (Gbs). This has been recently increased from 2.5 Gbs. The 10 Gbs standard is SONET OC-192, wherein SONET is synchronized optical network and OC is optical carrier. The increase in the modulation rate translates into a wider signal in the spatial domain. Consequently, the wider signal and smaller spacing means that the signals are very close together (in the spatial domain), and thus are very hard to separate. As a result, crosstalk may occur from adjacent signals.
One prior art separation method is to use a Fourier based filter to pass a particular wavelength from the input signal and block the other wavelengths on the signal. Such a filter 100 is depicted in FIGURE IA, wherein the filter 100 receives a WDM signal 101, which comprises λl, λ2, and λ3. The filter 100 blocks λl and λ3, and passes λ2 as output signal 102. The filter 101 has the transmission characteristics 103 shown in FIGURE IB. Note that this filter 101 has a low peak to valley ratio, i.e. the peak is not much higher than the floor. Thus, filter will have a (poor) low signal-to-noise ratio. To provide a higher (better) signal-to-noise ratio, several identical filters 100a, 100b, 100c, can be cascaded together as shown in FIGURE 1C. These filters also receive WDM signal 101, which comprises λl, λ2, and λ3, and blocks λl and λ3, while passing λ2 as output signal 102. The cascaded filters 100a, 100b, 100c have the transmission characteristics 104 shown in FIGURE ID. Note that the cascaded filters have a higher peak-to-valley ratio than the single filter of FIGURE IA. Thus, the cascaded filters will have higher (better) signal-to-
noise ratio. However, also note that this filter has a narrower width than the filter of FIGURE IA, thus this arrangement has better isolation but at a cost of having a narrower pass band.
Another prior art separation method is to use a Fourier based, Mach-Zehnder filter to divide the input signal into two periodic, inter-digitated sub-signals, each carrying an odd or even set of alternating wavelength signals, see United States Patent number 5,680,490 issued to Cohen et al., the disclosure of which is hereby incorporated herein by reference. As shown in FIGURE 2A, the WDM input signal 201 comprises a plurality of wavelengths, λl, λ2, λ3, and λ4. The filter 200 separates the input signal 201 into two sub- signals, which have complementary, inter-digitated wavelengths, one signal 202a with the odd wavelengths, λl and λ3, and the other signal 202b with the even wavelengths λ2 and λ4. Note that even and odd do not literally mean even and odd numbers, but rather indicate that alternating wavelengths in the input stream are separated into two streams. This usage will become apparent in the discussion of FIGURE 2C. The filter 200 has the transmission characteristics 205 and 206, for outputs 202a and 202b respectively, as shown in FIGURE 2B. Several filters can be cascaded to isolate single wavelengths, as shown in FIGURE 2C. The second stage filters 203 a, 203b have pass bands that are twice the size of first stage filter 200, as shown in FIGURE 2D, which depicts characteristic 205 which corresponds to signal 202a, and characteristic 207 which corresponds to signal 204a of filter 203a. The other characteristics of filters 203 a and 203b are not shown for the sake of simplicity. Note that the transmission profiles shown in FIGURE 2 are idealized, as the peaks are actually not flat, which reduces both the transmission band and the stop band of the filters.
A further type of filter incorporates a Michelson interferometer. FIGURE 3 A depicts a standard arrangement for a Michelson interferometer 300. An interferometer 300 is a device that can be used to measure lengths or changes in length with great accuracy by means of interference fringes, and operates as follows.
Light leaves particular point of a source 301 and falls on half-silvered mirror (or ' "beam splitter") 302. This mirror 302 has a silver coating just thick enough to transmit half the incident light and to reflect half of the light, in other words, at mirror 302, the light
divides into two portions. One portion proceeds by transmission toward mirror 303, and the other portion proceeds by reflection toward mirror 304. The waves are reflected at each of these mirrors and are sent back along their directions of incidence, each wave eventually leaving the interferometer via output 305. Since the light portions are coherent, being derived from the same point on the source 301, they will interfere, either constructively or destructively.
If the mirrors 303 and 304 are exactly perpendicular to each other, the effect is that of light from source 301 falling on a uniformly thick slab of air, between glass, whose thickness is equal to d2 - dl. Interference fringes appear, caused by small changes in the angle of incidence of the light from different points on the source 301 as it strikes the equivalent air film. For thick films a path difference of one wavelength can be brought about by a very small change in the angle of incidence.
If mirror 304 is moved backward or forward, the effect is to change the thickness of the equivalent air film. Suppose that the center of the (circular) fringe pattern appears bright and that mirror 304 is moved just enough to cause the first bright circular fringe to move to the center of the pattern. The path of the light beam striking mirror 304 has been changed by one wavelength. This means (because the light passes twice through the equivalent air film) that the mirror must have moved one-half a wavelength. By such techniques the lengths of objects can be expressed in terms of the wavelength of light. FIGURE 3B depicts a bandpass filter 300' using a Michelson interferometer. The band filter 300' operates similarly to the interferometer of FIGURE 3 A, except that mirror 303 has been replaced with GT (Gires-Tournois) mirror 306. As shown in FIGURE 3C, the GT mirror 306 is a resonator cavity comprising mirror 307, and moveable mirror 308. Mirror 308 is controlled by controller 309 to vary the distance d between the mirrors 307 and 308. This filter 300' is explained further in Dingle et al., "Properties of a Novel Noncascaded Type, Easy-to-Design, Ripple-Free Optical Bandpass Filter", Journal of Lightwave Technology, volume 17, number 8, August 1999, the disclosure of which is hereby incorporated herein by reference.
The GT mirror 306 introduces a non-linear phase into the transmission function which widens both the transmission band and the stop band, in other words, forming a more square-wave like transmission function. The light remains trapped within the cavity d, reflecting back and forth between the two mirrors 307, 308. Mirror 307 is a partially reflecting mirror, which allows a portion of the incident light from mirror 306 to pass through and continue to mirror 302. The amount of light passing through is a factor of the reflectivity of the mirror 307. The reflectivity also determines how long a portion of light entering the cavity will remain in the cavity, i.e. the cavity introduces a time delay. For example, if mirror 307 has a reflectance of 50%, then after five reflections in the cavity 3.125% of the light will remain in the cavity, with 96.875% of the light having left the cavity. The time delay is the time taken for the four additional reflections off of mirror 308. The phase of any one portion of light leaving the cavity depends on the distance d and the number of reflections in the cavity. This phase affects the interference of the light on the BI and B2 arms of the interferometer, which in turn changes the transmission pattern of the filter 300'. With a low reflectance of mirror 308, the non-linear effects are reduced such that filter 300' acts like interferometer 300, and produces a sine-squared transmission function. With a reflectance of .34, a relatively flat topped transmission function is produced. With a high reflectance a rippled top transmission function results.
However, Michelson interferometers are difficult to properly construct. Mirrors 304 and 306 of FIGURE 3B, as well as mirrors 303 and 304 of FIGURE 3A, must be precisely perpendicular to each other, at all points of the mirror. Similarly, mirror 302 must precisely aligned at a 45 degree angle with respect to mirrors 303,304 or 304,306. Any misalignment of any of the mirrors causes the filter to function improperly, as the alignment is critical for the interference to occur. Interferometers are particularly sensitive to alignment changes from vibrations and temperature changes. Note that building a Michelson interferometer using fiber waveguides does not overcome the alignment problems of bulk optics interferometers.
Therefore, there is a need in the art for a WDM filter that has good transmission characteristics while being easy to construct and maintain.
SUMMARY OF THE INVENTION
The present invention is directed a system and method which uses a fiber coupler with a GT resonator.
The inventive filter uses an integral GT resonator and fiber coupler to form a Michelson interferometer. The coupler will divide the light into two arms, with one arm being reflected by a high reflectance mirror, and the other arm being reflected by a GT resonator. The size of the arms and the spacing of the resonator determines the filtering characteristics of the filter. Tuning the filter is accomplished by illuminating a photosensitive fiber material that is used to form a portion of the filter. Illumination causes the refractive index of the material to change, and thereby changes the effective length of the fiber, which in turn adjusts or tunes the characteristics of the filter.
The foregoing has outlined rather broadly the features and technical advantages of the present invention in order that the detailed description of the invention that follows may be better understood. Additional features and advantages of the invention will be described hereinafter which form the subject of the claims of the invention. It should be appreciated by those skilled in the art that the conception and specific embodiment disclosed may be readily utilized as a basis for modifying or designing other structures for carrying out the same purposes of the present invention. It should also be realized by those skilled in the art that such equivalent constructions do not depart from the spirit and scope of the invention as set forth in the appended claims. The novel features which are believed to be characteristic of the invention, both as to its organization and method of operation, together with further objects and advantages will be better understood from the following description when considered in connection with the accompanying figures. It is to be expressly understood, however, that each of the figures is provided for the purpose of illustration and description only and is not intended as a definition of the limits of the present invention.
BRIEF DESCRIPTION OF THE DRAWING
For a more complete understanding of the present invention, reference is now made to the following descriptions taken in conjunction with the accompanying drawing, in which:
FIGURES 1 A- ID depict prior art filters that isolate a single wavelength from an input signal;
FIGURES 2A-2D depict prior art filters that separate multiple wavelength input signals into odd/even sub-signals;
FIGURES 3A-3C depict arrangements using Michelson interferometers;
FIGURE 4 depicts one embodiment of the inventive filter; FIGURE 5 depicts a filtering system using the inventive filter;
FIGURE 6 depicts a second embodiment of the inventive filter;
FIGURE 7 depicts a preferred embodiment of the inventive filter;
FIGURE 8 depicts another embodiment of the inventive filter of FIGURE 7;
FIGURE 9 depicts the filtering characteristics of the inventive filter; FIGURE 10 depicts an alterative embodiment for the GT mirror;
FIGURE 11 depicts an inventive arrangement having two inputs and two outputs; and
FIGURE 12A and 12B depict an inventive arrangement having one input and a plurality of outputs .
DETAILED DESCRIPTION
A WDM signal consists of multiple channels with each channel having its own range of wavelengths or frequencies. As used herein, the terms "channel" or "spectral band" refer to a particular range of frequencies or wavelengths that define a unique information signal. Each channel is usually evenly spaced from adjacent channels, although this is not necessary. For example, the wavelength slicers shown in FIGURE 2 can be configured to separate channels based on a 50 GHz spacing between adjacent channels. Uneven spacing may result in some complexity in design, but, as will be seen, the present invention can be adapted to such a channel system. This flexibility is important in that the channel placement is driven largely by the technical capabilities of transmitters (i.e., laser diodes) and detectors and so flexibility is of significant importance.
FIGURE 4 is a detailed schematic diagram of a wavelength filter 400. Filter 400 is a three port device, with one input port 401 and two output ports 402, 403. Note that the output port 403 is connected to the input port 401. Thus, on that fiber, input light flows in one direction, and output light flows in the opposite direction. Thus, a light separator may be utilized to physically separate the light paths of the input and output lights. This separator could be either integral with the filter or separate from the filter. As shown in FIGURE 4, a circulator 404 is used as the input/output light separator, however other types of separators could be used. Also note that the circulator could be replaced with an isolator, this would effectively change filter 400 into a two port device, with one input port and one output port. The isolator would block or filter the light signal being sent down the input port.
Wavelength filter 400 uses a standard 3dB fiber coupler 405 to separate the light into two portions. The coupler will evenly split input light into equal portions for as many output arms as are connected to the coupler 405. In this case, the two output arms, upper arm 406 and lower arm 407, each receive substantially 50% of the input light. Thus, if another filter has 4 output arms, then each arm would receive 25% of the input light. Note that the designation of "upper" and "lower" for the arms or paths is only to allow visual
correspondence with the depiction of the FIGURES and is not meant for actual orientation of the filter as the filter may be oriented in any manner.
After the input light signal 401 has been split into two portions, one portion travels on lower arm 407 until encountering minor Rl 408, which has a reflectance of substantially 100%. The term substantially as used herein is to indicate as close to the value as technically and/or commercially possible. The distance from the coupler 405 to mirror Rl 408 is LI. After reflecting off of the mirror 408, the light portion travels back to the coupler 405. Thus, this light portion has traveled 2*L1.
The other portion of the input signal 401 travels on the upper arm 406, until encountering collimator 409, which coUimates the light portion. The distance from the coupler 405 to collimator 409 is L2. The collimated light is then incident onto the GT resonator 410. The GT resonator 410 is comprised of mirror R2 411 and mirror R3 412, which are separated from each other by distance d. In a preferred embodiment, mirror R2 411 has a reflectance of approximately 16% and mirror R3 412 has a reflectance of substantially 100%. The distance from the collimator 409 to minor 411 is L3. In a preferred embodiment, controller 413 is operable to vary the location of GT resonator 410, which thereby changes the distance L3.
Controller 414 is operable to vary the location of mirror R3 412 which thereby changes the distance d. Controller 414 may be implemented within FT resonator 410 or be implemented external to such GT resonator 410. Note that the distances LI, L2, L3, and d are not physical distances but rather effective distances which includes the refractivity of the medium through which the light is passing. Thus an alternative foπn of control would be to place variable refractive index mechanisms in the path of L3 and d. Thus, the effective distances of L3 and d could be independently controlled by varying the refractive index of the medium material over at least a portion of the distances L3 and d, respectively. Further note that controllers 413, 414 are optional, and, a fixed distance filter can be implemented instead, for example. The light exiting the cavity 410 is inputted back on to fiber 406 via collimator 409.
The relationships of LI, L2, L3, and d, as well as the reflectivity of mirror R2 411 determines the characteristics of the filter 400. The reflectivity of mirror R2 411 will control the profile of the transmission function of the filter 400. If the reflectivity of mirror R2 411 is low, meaning that most of the light reflecting off of mirror R3 412 passes through mirror R2 411, then the filter 400 will act as a Mach-Zehnder interferometer. A reflectivity of about 16% (reflectance of .4) results in a wide, flat-topped or square-shaped transmission function, which is the preferred embodiment. As the reflectivity is increased, ripple patterns will begin to appear in the top of the transmission function. Thus, the reflectance of the mirror R2 411 controls the shape of the transmission characteristics of the filter.
The lengths of LI, L2, L3, and d determine the pass band width, and the spacing between the peaks of the pass bands. Setting d equal to (Ll-L2-L3)/2 results in interdigitated transmission passbands similar to those shown in FIGURE 2B. In a preferred embodiment, channel spacing, which is defined as "λi - (λi-1)," is determined by the value of d. Controller 414 is used to change the length of d to change the size and spacings of the passbands. In one embodiment, controller 413 may be used to change the length of L3 to maintain the relationship of d-=(Ll-L2-L3)/2, as d is changed. This relationship needs to be accurately controlled to within about 1% of the equation's value, otherwise the passband will be degraded. The light returning from the two branches 406 and 407 interfere with each other in the coupler 405. The light on branch 406 has a phase difference with respect to the light on branch 407 from the cavity 410. The phase difference causes destructive and constructive interference of the light from the two branches 406, 407, such that the light outputted onto branch 402 has constructive interference occurring for one of the even wavelengths and the odd wavelengths, and destructive interference occurring for the other one of the even wavelengths and the odd wavelengths. The light outputted on branch 403 would have the opposite interference occurring. For example, branch 402 could output the even wavelengths through constructive interference and block the odd wavelengths through destructive interference. Branch 403 would then output the odd wavelengths and block or
filter the even wavelengths. The light output on each branch would be an inter-digitated wavelengths, as shown in FIGURE 2B.
The noise floor, or the distance between the tops of the transmission peaks and the valleys between the transmission peaks, is about -15 dB in a preferred embodiment. Cascading several filters together, with each filter having substantially identical characteristics with the other filters, will increase the noise floor. For example, the arrangement 500 shown in FIGURE 5 has three filters 501, 502, 503, cascaded together. The output from filter 503 would have a noise floor of -45 dB (three times -15 dB). The arrangement 500 includes isolators 504, 505 to block noise from flowing back up the input line. Element 506 may be a circulator that separates the alternate channels from the input line, and passes those channels to system 507, which may comprise two additional filters arranged similarly as filters 502, 503, for example. Alternatively, element 506 may be an isolator.
FIGURE 6 depicts an alternative aπangement for the filter of FIGURE 4. The filter 600 is fiber-based, in that no air gaps are present in the system. Tins filter includes GT resonator 601, which comprises partially reflective mirror 602, reflective mirror 603, and gap 606, which operate as described in FIGURE 4. The filter also includes reflective mirror 604, and coupling area 605, which operate as described in FIGURE 4. This filter is more stable than that of FIGURE 4. Since all of the elements are physically connected together, particular elements cannot have their alignments changed relative to other elements from vibration or other affects. However, this arrangement is difficult to tune. The elements must be precisely aligned, and then held in place while the connecting glue sets. During setting, the elements may slip out of alignment, or changes caused by the glue setting may cause movement of the elements. FIGURE 7 depicts a preferred embodiment for the filter of FIGURES 4 and 5. The filter 700 includes the same elements as the filter of FIGURE 6, but incorporates photosensitive fiber portions 701, 702, 703, and 704. Each of these portions can have their respective refractive index (n) changed by illuminating the portion with light of a particular wavelength, e.g. ultraviolet light. By changing the refractive index, the effective path
length of the fiber is changed, as effective path length is equal to nl, where n is the refractive index and 1 is the length of the fiber or fiber portion. The material used for each portion may be selected such that the refractive index increases with exposure to the light or the refractive index decreases with exposure to the light. The portions 701, 702, 703, and 704 may all comprise a single type, e.g. all increasing or all decreasing, or they may comprise different combinations of the types, e.g. 701,702 increasing and 703,704 decreasing. Note that each portion, e.g. 701, may comprise at least two portions, with one portion having an increasing change from the light, and another portion with a decreasing change from the light. These portions may all be used, or only some may be needed, for example the fiber 700 may only include portion 703. The portions allow the fiber to be tuned to achieve the desired optical filtering profile.
It should be recognized that portion 701 is utilized to couple two originally separate fibers, which are shown in FIG. 7 as exemplary fibers 705 and 706. That is, originally separate fibers 705 and 706 are coupled by coupling portion 701 to form fiber 700. The coupling ratio of the fiber coupler depends on the coupling distance (i.e., the length of portion 701) between the two fibers 705 and 706. The coupling ratio of the fiber coupler determines the isolation of the spectrum (i.e., determines the filtering). Therefore, the coupling length can be tuned, for example, by ultraviolet (UV) trimming the length of portion 701. Thus, the coupling ratio may be improved as well as the isolation of the spectrum (i.e., the filtering), as shown in FIGURE 9 hereafter.
An alternative embodiment to that of FIGURE 7 is shown in FIGURE 8. The filter 800 has a substantial portion of the filter 801 comprising photosensitive fiber material. Note that the substantial portion 801, may comprise a plurality of segments, with particular ones of the segments comprising increasing refractive index material, and other ones comprising decreasing refractive index material. Also note that the portion 801 may be comprised all of one material, i.e. all increasing or all decreasing material. It should be recognized that because collimator 409 of FIGURE 4 is not included in the exemplary implementations of FIGURES 6, 7, and 8, length L3 discussed above in conjunction with FIGURE 4 is omitted (or is equal to zero). Thus for example, setting d equal to (L1-L2V2 results in inter-digitated transmission passbands similar to those shown in FIGURE 2B. In
a preferred embodiment, the photosensitive fiber portions of FIGURES 7 and 8 may be used to change one or more of effective distance d, effective length of LI, and effective length L2 to maintain the relationship of d=(Ll-L2)/2. As discussed above, this relationship is most preferably accurately controlled to within about 1% of the equation's value.
FIGURE 9 depicts an exemplary optical filtering profile for the filters of FIGURES 4, 6, 7, and 8 of a preferred embodiment. The profile includes first output 901 and second output 902. These outputs depict only a portion of their response spectrum. Note that the outputs are odd/even oriented. The fully reflective mirror of the GT resonator (e.g., mirror 412 of FIGURE 4 and mirror 603 of FIGURE 6) has been shown to be flat. However, as shown in the exemplary implementation of FIGURE 10, the fully reflective mirror of resonator 1000 can include a curved surface 1001. The curved mirror reduces insertion loss of the reflected signal, as the curvature of the mirror is selected to approximately match the Gaussian propagating wavefront emerging from collimator 310 (FIGURE 3). This is described further in the related and co-pending application U.S. Provisional Patent Application Serial Number 60/201,648, entitled "WIDE TRANSMISSION OPTICAL COMB FILTER WITH WIDE PASS BAND AND WIDE STOP BAND," filed May 1, 2000, and co-pending U.S. Patent Application Serial Number 09/614,782 entitled "WIDE TRANSMISSION OPTICAL COMB FILTER WITH WIDE PASS BAND AND WIDE STOP BAND," filed July 12, 2000, the disclosure of which are hereby incorporated herein by reference.
FIGURE 11 depicts another alternative embodiment for the inventive system. The inventive filter 1100 includes two GT resonators 1101, 1102. This filter would receive two input spectrums, namely inputl 1103 that comprises fa, fb, fc, fd,..., and input2 1104 that comprises fl ,f2, f3, f4, ..., wherein both GT resonators operate as described above. The outputs are separated from the input lines via circulators 1107, 1108. The first output, outputl 1105 would comprise fa,f2,fc,f4..., while the second output, output2 1106, would comprise fl, fb, f3, fd...., which is the complement of outputl 1105.
FIGURE 12A depicts an exemplary filter that comprises a plurality of GT resonators 1201, e.g. four, a single input 1202, and plurality of outputs 1203, e.g. four. Note that one of the outputs is formed by placing a circulator 1204 on the input line 1202. Thus for an input signal of fl, f2, £3 ,f4...fN, each output will comprise a portion of the input signals, related to the number of outputs. For example, with four outputs, the first output 1203 A, shown in FIGURE 12B, comprises fl, f5, f9..., the second output 1203B comprises £2, f6, flO..., the third output 1203c (not shown in FIGURE 12B) comprises f3, f7, fl 1..., and the fourth output 1203D (not shown in FIGURE 12B) comprises f4, fS, fl2..., wherein the number of nulls between the peaks is N-l of the fibers fused together to form the coupling device.
Note that adding additional circulators to the output lines 1203, each output line could then receive an input signal. Thus, the arrangement of FIGURE 12A would in such an implementation have four inputs, four outputs. This forms an NxN filter. Each output would be comprised of the inputs as described with regards to FIGURE 11. For input signals one: fl,f2,f3,f4...; two: fix, f2x, £3x, f4x...; three: fa,fb,fc,fd..., four: fA, fB, fC, fD...; then the first output would comprise fl, f2x, fc, fD..., the remaining outputs would be similarly formed.
Each of the inventive filters can include at least one modulator that is located within the filter, e.g. in the cavity of the GT resonator, such that output going to ports 1 and 2 can be switched by changing the phase of the modulator. An example of such a modulator that may be implemented is a lithium niobate electro optic modulator. For the fiber based GT resonators, a thermal material can be used, such that a change in temperature will cause a change in the phase of the light passing through the modulator.
The GT resonator in this system can be implemented in different ways. As one example, it can be made from a solid etalon which is a solid block of optical material polished at both surfaces and one side coated with a partial reflecting mirror, and the other side coated with a 100% reflecting mirror. Such an etalon may be made integral in a fiber as shown in FIGURES 6 and 7. Another possibility is to use an air gapped resonator formed by a pair of mirrors such as the mirrors described in the related co-pending U.S.
Provisional Patent Application Serial Number 60,200,969, entitled "LOW LOSS ULTRA STABLE FABERY-PEROT ETALON," filed May 1, 2000, and co-pending U.S. Patent Application Serial Number 09/642,985 entitled "LOW LOSS ULTRA STABLE FABRY- PEROT ETALON," filed August 21, 2000, the disclosures of which are hereby incorporated herein by reference.
Although the present invention and its advantages have been described in detail, it should be understood that various changes, substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims. Moreover, the scope of the present application is not intended to be limited to the particular embodiments of the process, machine, manufacture, composition of matter, means, methods and steps described in the specification. As one of ordinary skill in the art will readily appreciate from the disclosure of the present invention, processes, machines, manufacture, compositions of matter, means, methods, or steps, presently existing or later to be developed that perfonn substantially the same function or achieve substantially the same result as the corresponding embodiments described herein may be utilized according to the present invention. Accordingly, the appended claims are intended to include within their scope such processes, machines, manufacture, compositions of matter, means, methods, or steps.