Optical waveguides – With optical coupler – Switch
Reexamination Certificate
2000-10-06
2002-06-25
Healy, Brian (Department: 2874)
Optical waveguides
With optical coupler
Switch
C385S018000, C324S097000, 36
Reexamination Certificate
active
06411751
ABSTRACT:
FIELD OF THE INVENTION
The invention relates generally to optical cross-connects and, more particularly, to training the steerable switching elements, such as micro-electromechanical systems (MEMS)-based mirror arrays, used in optical cross-connects.
BACKGROUND OF THE INVENTION
MEMS steerable elements are being used in a variety of applications. For example, clusters of MEMS mirrors are being contemplated for use in optical cross-connects to provide all-optical switching of optical signals through a network. In one such cross-connect application, an optical signal (i.e., light beam) is launched from an input optical fiber through an input lens that focuses the beam onto an input mirror. Control voltages are applied to tilt the mirror on its axes to direct the beam to an output mirror. In a similar manner, the output mirror is tilted on its axes by control voltages to direct the beam through an output lens to an output optical fiber. By controllably tilting the input and output mirrors, optical signals can therefore be switched between the various inputs and outputs of the optical cross-connect.
Performance of optical cross-connects that use steerable mirrors largely depends on how well the input and output mirrors are aligned for a particular cross-connection; that is, whether the steerable mirrors are tilted to provide maximum optical signal power for the optical signal being switched. One approach for ensuring that mirrors can be aligned properly is to measure and store the control voltages as factory settings for each possible cross-connection, e.g., for each possible input-to-output pairing. These stored values are then used to control the establishment of cross-connections. However, a problem arises during operation of the cross-connect that may cause the mirrors to be out of alignment even when the factory settings are used to establish cross-connections. For example, small changes in the system may occur due to temperature, aging of components, drifts in control voltages, mirror characteristics, and so on. These changes can cause improper alignment of the mirrors and subsequent signal losses.
So-called indirect methods are known for attempting to determine the actual position of a MEMS mirror in such an arrangement. Indirect methods typically use a light source (e.g., infrared radiation) and a camera (e.g., charge coupled device), whereby the light source is used to illuminate a mirror and the camera is used to record the position of the reflected light. The position of the MEMS mirror is then calculated based on knowing where the light source is located as well as the measured position of the light recorded by the camera. This information along with information about the control voltage levels associated with the respective positioning of the mirrors is then stored, e.g., in a database. Among other disadvantages, this method becomes unwieldy as the number of mirrors in the cluster become large. Moreover, this approach can be very disruptive if used during the cross-connect operation, e.g., when illuminating a mirror that is being used for an active cross-connection. Consequently, this technique is limited to use before a system is placed in operation, e.g., “coarse” training of the MEMS mirror array during factory setup, system initialization, and so on.
SUMMARY OF THE INVENTION
Signal losses caused by changes in the alignment of steerable switching elements used for cross-connecting an optical signal are substantially reduced according to the principles of the invention by controllably and selectively training the steerable switching elements as a function of measured input and output power of the cross-connected optical signal. More specifically, training of one or more steerable switching elements associated with a particular cross-connection is performed by dithering (or re-aligning) the tilt position of the one or more switching elements to increase the optical signal power in the optical signal without disrupting the active cross-connection of that optical signal. Because measurement of optical signal power and control of switching elements is performed as a function of only the particular cross-connected optical signal, the training technique according to the principles of the invention requires less extensive processing resources as compared to the prior art methods. Moreover, for a large capacity optical cross-connect having clusters of steerable switching elements, the individual cross-connections can be independently and simultaneously trained according to the principles of the invention.
In one illustrative embodiment, optical monitoring arrangements monitor the input and output optical signal power of optical signals coupled to the cross-connect inputs and outputs. The cross-connect includes a switching fabric comprising a plurality of steerable MEMS mirror elements used as switching elements for directing the light beams within the cross-connect. By comparing the actual optical power loss (e.g., measured input power minus measured output power) with a previously stored value representing the expected optical power loss for a particular cross-connection, small adjustments can then be made, as appropriate, to optimize the alignment of the mirrors associated with the cross-connection. For example, if the difference between the measured and expected optical power loss exceeds a prescribed threshold, then a dithering process is initiated whereby the individual mirrors are “walked through” alternate tilt positions until the measured optical signal power has been optimized, e.g., increased. By scaling the optical signal power measured at the output as a function of the optical signal power measured at the input, the output optical signal power can be normalized with changes in input optical signal power to provide more accurate alignment adjustments and to avoid unnecessary adjustments. For example, when a legitimate or otherwise acceptable change in input power is correspondingly measured, output power may be expected to change such that unnecessarily changing positions of the mirrors would not achieve any greater output power.
REFERENCES:
patent: 4560925 (1985-12-01), Niven et al.
patent: 5333117 (1994-07-01), Ha et al.
patent: 5960132 (1999-09-01), Lin
Giles Randy Clinton
Gottlieb Albert M
Neilson David Thomas
Healy Brian
Lucent Technologies - Inc.
Murgia Gregory J.
Woo Sarah
LandOfFree
System and method for training an optical cross-connect... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with System and method for training an optical cross-connect..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and System and method for training an optical cross-connect... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2981385