Patent Application
20060133721
The present invention relates to optical or photonic switches, and in particular to the monitoring, preferably in real time, of the status of optical switches that include an electronic element that controls the switching actions. Such optical switches include, but are not limited to, a Mach Zehnder Interferometer (MZI) structure in which one or more heating elements are employed to affect the refractive characteristics of a waveguide, thereby enabling redirection of light passing through the optical or photonic switch.
A network of such switches may be used to route an optical stream from a source to a destination on the network. It is imperative that each switch function properly, i.e. all its paths are in working order, and to facilitate this it is desirable to monitor all the switches (preferably in real time). The use of optical switches is well known in the communication arts. Network monitoring is also well known in the networking arts, and the use of MZI switches is also well established.
Optical methods for monitoring the status of an optical switch are known. For example, U.S. Pat. No. 6,198,856 to Schroeder et al. shows an optical switch having test ports. In this teaching it is pointed out that not all optical switches are compatible with test or monitoring equipment, e.g. the switch presented in U.S. Pat. No. 5,699,462. In '856, Schroeder discloses a variation of the switch having modifications specific to supporting a monitoring function.
Electrical methods for monitoring the status of an optical switch are also known. U.S. Pat. No. 6,750,655 to Staple discloses methods for affirming switched status of MEMS-based devices. A configuration is provided by which it may be determined whether a MEMS device is in a select state. The select state is defined by a position of a moveable element, which is moved with electrostatic forces upon activation of an electrode. The select state is detected with a sensing configuration that has first and second regions. The regions are generally separated such that they are electrically uncoupled unless the moveable element is in the position that defines the select state. A detector may be provided to indicate whether the first and second regions are coupled electrically.
Staple's and similar testing methods can provide only a partial status of a switch, in the sense that even if the tested switch “works” in its present “select” state, an underlying problem goes undetected. For example, assume an optical switch with two states, 1 and 2. In state 1 the switch directs light to a first port 1, and in state 2 the switch directs light to a second port 2. According to Staple, testing the switch (in his case a MEMS mirror) in select state 1 provides the following status information: the switch is now in state 1 and not in state 2 and b) the switch “works” in state 1, i.e. directs the light to port 1. This is by no means a complete knowledge of the switch status. For example, assume that the MEMS mirror is stuck in state 1, and therefore unable to be moved to state 2. As long as the mirror stays voluntarily in state 1, the switch is “healthy” in the sense that it performs its current function of sending light to port 1, but it is not healthy in the sense that it will not be able to move to state 2. This problem is further illustrated in
a shows a basic crossbar MEMS switch matrix 100 comprising 4 inputs 102a-d, 4 outputs 104a-d and 16 mirrors. In the left view, all mirrors are shown in an OFF position (letting light from the input pass through). To create a connectivity, 4 mirrors A, B, C and D are actuated and create the exemplary configuration shown in the right view. Staple's methods can periodically test each mirror in its present select state and detect its present status. However, as mentioned, this testing will not detect if a mirror is stuck in its current position. Suppose a mirror E connecting input 102c to output 104b in the left view is stuck at OFF. Periodic testing of mirror E indicates it is in state OFF, as expected (and from this point of view “healthy”). If an operator tries now to connect input 102c to output 104b by actuating mirror E, the operation will fail. Even though the current status of each mirror is known from the periodic testing,
Staple's testing cannot identify the potential problem that will prevent the switching of mirror E to a next configuration or state. In essence, Staple's method has no “functional prediction” capability as to the ability of an optical switch to switch to another configuration, and no capability to detect an underlying problem that does not affect the operation in the current select state.
U.S. Pat. No. 6,317,526 to Shirasaki et al. shows a specific example of a photonic switch employing an MZI configuration that incorporates heating elements and monitoring elements. Two methods for monitoring the heating elements are described. U.S. Pat. No. 6,470,106 issued to McClelland et al. shows a “bubble” based optical switch incorporating a failure monitor. U.S. Pat. No. 6,636,668 issued to Al-Hemyari et al. teaches an optical means of monitoring optical switches in real time.
None of the methods described above can monitor a switch path that is not currently passing light, i.e. monitor one of the possible paths of a switch not passing light while the monitoring takes place. There is thus a widely recognized need for, and it would be highly advantageous to have, a monitoring method that can check continuously, preferably in real time, the status of any switch path, regardless of whether it is passing light or not at the time of the monitoring. Such monitoring will provide a complete status report of the entire optical system.
The present invention discloses a system and methods for monitoring the complete status of any switch path assumable by a switch, regardless of the switch's current select state. As used herein, “complete” includes a predictive aspect: the monitoring of the present invention can, in contrast with all known monitoring methods, identify an underlying problem in each switch, thereby predicting a potential switching problem ahead. In particular, the system and methods disclosed herein can provide a “functionality prediction”, i.e. predict whether a switch can switch from the current select state to another select state based on a measurement performed in the current select state.
The present invention provides a circuit that is used to monitor, in real time, electrically addressable elements embedded in optical switches, particularly heaters embedded in thermo-optical switches. Monitoring of such elements facilitates the early detection of the failure of each such element, and thereby the failure of the switch as a whole.
While the following description deals in detail with thermo-optical switches, the circuit, system and method disclosed herein can be implemented in all switch types that use an electronic control method for triggering the optical switching.
According to the present invention there is provided asystem for monitoring the status of an optical switch that includes at least one electrically addressable element operative to provide the switching action, the system comprising an electrical monitoring circuit operative to periodically test electrically the at least one electrically addressable element while the switch is set to a current switching configuration, and a mechanism for effecting the periodic electrical testing, the mechanism operative to provide an optical switch status that includes a functionality prediction, whereby the periodic electrical testing does not interfere with optical traffic passing through the optical switch.
According to one feature in the system of the present invention, the functionality prediction includes a prediction of the switching capability from the current switching configuration to another switching configuration
According to the present invention there is provided a system for monitoring a non-active optical switch having at least one electrically addressable element, comprising a first mechanism for electrically testing the at least one electrically addressable element at a current switching configuration and a second mechanism for obtaining a current switch status and for predicting a capability to assume another switching configuration based on the electrical testing at the current switching configuration, whereby the electrical testing is done periodically without affecting any regular function of the at least one electrically addressable element.
According to the present invention there is provided a method for monitoring the status of an optical switch that has at least one electrically addressable element, comprising the steps of periodically addressing the at least one electrically addressable element to obtain periodic output signals while the switch is in a current switching configuration and processing the output signals to obtain continuous, real-time information that includes a present status of the optical switch and a functionality prediction on a switching capability from the current switching configuration to another switching configuration.
According to the present invention there is provided a method for providing real time status information on an optical communications system that includes a plurality of switches, each switch having at least one electrically addressable element, and a plurality of optical components, the method comprising the steps of electrically monitoring each optical switch in a current switching configuration through the its at least one electrically addressable element to provide a present status and a functionality prediction on a switching capability from the current switching configuration to another switching configuration and optically monitoring one or more of the optical components to provide optical communications system status information.
The invention is herein described, by way of example only, with reference to the accompanying drawings, wherein:
a illustrates the basic problem with existing status monitoring methods as applied to a 4×4 optical switch;
b illustrates the operation of a Mach Zehnder Interferometer (MZI) optical switch;
The present invention provides a monitoring circuit that is used to monitor electrically addressable elements embedded in optical switches. These elements are responsible for the optical switching action, for example through the electro-optic, thermo-optical effect or a mechanical actuation effect. The monitoring is based on electrical signals passed through an electrically addressable element such as a heater that is integral to the switch, the signal response monitored and evaluated. In other words, in contrast with prior art, the monitoring system and method disclosed herein provide electrical monitoring of the status of an optical switch. Furthermore, the method disclosed herein differs from prior art methods in that it monitors the actual switching mechanism, and not only the switching command and its processing, or the current switch position. The method also monitors all independent switching elements, verifying that they operate in all their positions needed to operate the optical switch system. The description focuses on a particular type of optical switch—the thermo-optical MZI type switch, shown schematically in
Optionally, in addition to the electrical monitoring mechanism of switches, switch parts, switch arrays, etc., the present invention also provides optical monitoring of the light paths that complements in certain cases the electrical monitoring, to provide a complete monitoring system at the level of both optical chip and optical switching system. One optical monitoring mechanism involves taps and photodetectors for tapping into, and detecting an optical signal present in the fibers, or in other words, external monitoring of the light signals. Another optical monitoring mechanism involves a laser (or any other light source with a light characteristics range available and supported in the system) providing a signal that monitors a pigtail connection. This mechanism monitors all pigtails of the optical switch by passing light through additionally added or already present unused fibers to make sure that a pigtail is not disconnected or degraded. This is done by making sure that each pigtail connected to the optical switch, which can have multiple fibers in an array, passes the light from the laser used into the switch and back out to a photodetector making sure the pigtail is secured properly to the optical switch and is passively passing the light from all fibers in the pigtail, not just the ones used for the monitoring. Both these methods are optionally added to the electrical monitoring to provide a complete monitoring solution.
The present invention provides an inventive monitoring circuit, system and method that can monitor every part of an optical switch at all times, thus providing an extremely reliable (system-wise) switch, which can be used for high availability demanding applications, for example for protection switching applications. As with other monitoring methods, the monitoring or testing method disclosed herein can monitor the light coming into all ports and going out of all ports of a switch or of a switch array (“chip”), making sure that the present configuration of a tested switch is working and is within required parameters. This is simply done by tapping into the input and output fibers and measuring the signals that are present, making sure that the effect the switch has on these signals is the desired one. However, in contrast with known optical monitoring methods, the method disclosed herein can go beyond such simple “active” monitoring, and can monitor the (theoretical) configurations not used at present but potentially used in the future. In other words, the monitoring of the present invention ensures that a light path is operational at all times, even if that path is not in use (not passing light at present). The present monitoring can thus be applied to a “non-active” switch, switching configuration or light path, as well as to a switch in an intermediate position between ON and OFF configuration. A “non-active” switch as used herein means a switch that does not pass light through a certain path or paths of that switch when tested. Known monitoring systems and methods, which are exclusively optical, monitor only “active” switches, configurations or light paths, i.e. only switches that pass light at the time of the monitoring.
As mentioned, the optical switches which can be monitored using this present invention include regular ON/OFF (binary) switches as well as “intermediate” switches, i.e. switches that not only pass the entire light signal to one port or another, but can multicast the signal between different ports at different weighted ratios. This means for example that in an 8×8 switch system in which there are 8 inputs and 8 outputs, input 3 may be switched to outputs 2, 5, 6 and 7 at 30%, 25%, 15%, 30% power respectively. In another example of thermo-optical switches based on MZI, the control of such switches can be done by applying an electrical power level somewhere between the ON level and the OFF level of the switching element in the optical switch, according to the desired ratio of the optical power split needed by the MZI being controlled. The present invention can also monitor the status of all such possible configurations, ensuring that they are functioning appropriately both in the present configuration and in future configurations to be switched by the optical switch. In addition the present invention also includes status monitoring of variable optical attenuators (VOA), which may be basically implemented using a multicast 1×2 switch having one input and two outputs, where the second output is either disregarded or discarded. The signal from input 1 to output 1 (1→1) is attenuated by transferring some part of the optical power to output 2 (1→2) according to the desired ratio. This is done, for example in thermo-optical MZI switches controlled by PWM, by setting the PWM duty cycle, or in other words the heating delivered by the heater to the optical chip on one of the arms of the MZI to an intermediate level between that of an ON level and that of an OFF level, according to the desired attenuation.
In a particular example, the monitoring system and method of the present invention is applied to a 1×2 switch passing light in a 1→1 configuration. The present monitoring can ensure that the 1→2 path will be operational when we want to reroute the light, while the switch is still in the 1→1 configuration, without light actually passing through the 1→2 configuration. In other words, the monitoring described herein can indicate or “predict” if the switch will be able to switch from the 1→1 configuration to the 1→2 configuration, for example in a protection scenario. The monitoring supplied not only information on the current switch status (i.e. in state 1 and not state 2), but also on the switch functionality (ability to switch from state 1 to state 2). Suppose the switch is used to protect a service going through a port 1 by switching to a port 2 (protection mode) if a facility connected to port 1 fails. The switch is tested when in select state 1. Prior art methods cannot predict that when the facility connected to port 1 will fail, the service will indeed switch to port 2. In other words, prior art methods cannot predict if the switch will switch to the protection mode. The present invention can make this prediction, allowing an operator to replace the faulty switch before an actual failure.
To clarify, as used herein, “prediction” refers to prediction of switch functionality and not necessarily to prediction of a future failure. In an optical switch comprising a plurality of switches (e.g. a switch matrix), all the switches can be continuously monitored in real time, making sure they are functional, whether they are currently being used to pass light or whether they are unused at the moment but potentially required in future configurations requested from the optical switch.
Referring now to the figures,
In operation, module 212 receives a command to set the optical switch to a chosen configuration, and translates the request to a set of control PWM values to each heater. PWM systems and methods are known. For a recent reference, see for example U.S. Pat. No. 6,647,318 to Salsbury. The module then sends the PWM command to the gates of the driving FETs. Each PWM command opens the gate for a specific time slice (out of the control cycle of the heaters), essentially passing a 12V current from the “12V” power supply to the appropriate heater/s on the PLC for a specific duty cycle. For example, the 12V connection is “HI” x % of the time, while it is LOW (disconnected from the heater) (100-x)% of the time, thus achieving a certain heating differential between the two MZI paths. The switch thereby executes the requested configuration until a new PWM command is received.
The monitoring and sensing of the operational status of the optical switch is done through the monitoring of the status of each heater. The heater is the only active component in the optical switch, and if the heater is functioning properly and has a closed circuit it will generate the required heat, which will perform the switch operation due to the inherent physical characteristics of the material. Therefore, a good heater with a good connection to the driver circuit validates a good optical switching operation. Module 212 does this monitoring and sensing continuously. It reads the ADC at a number of time slots along the PWM cycle, ensuring that the correct value is read by the FPGA, or any other electronic device that can determine and report the status from the ADC both when PWM is HI (the FET is passing an electrical current) and when it is LOW (the FET is not passing an electrical current). This makes sure that the FET is in working order, since the ADC reading checks that the FET can be used to both connect and disconnect the electrical current from the heater in the MZI. The module also checks the heater resistance, making sure the heater is able to carry the current needed to drive the optical switch with the required resistance. This check makes sure that the heater has not deteriorated in any manner. The procedure described above essentially operates the entire mechanism needed to drive the optical switch, making sure everything is in working order. Note that this can be done when the switch's monitored path is OFF (no light passing), unlike in any known optical monitoring method and system.
In cases when no PWM is driven to a certain heater (due to the nature of the request sent to the switch control and its translation into PWMs for the “n” MZIs comprising the optical switch) a very short PWM duty cycle is enforced on the unused heaters for a time long enough to measure the ADC readings. This procedure ensures that the circuit is operational, but is not long enough to interfere with the thermo-optical effect of the optical switch. The thermo-optical effect typically requires around Ims to cause a switching from ON to OFF, whereas a short PWM duty cycle used to measure the operational status of an unused MZI and its heater according to the present invention will typically require about 1 μs (or less) and will therefore not affect the optical signal passing through the switch. The same is true for a circuit in a full duty PWM control at HI. In this case, if applicable, one short burst of LOW control will be forced on the switching mechanism to ensure the capability to shut off the FET running electrical current to the heater, thus ensuring complete functionality. This will also be done for a time short enough not to affect the performance of the optical switch on the optical signal being passed through the switch, but long enough to measure and assure the operational status of the switching mechanism.
Optional Additional Optical Testing
The electrical testing through an electrically addressable element of a switch permits testing of the type that could not be done in prior art, specifically testing of a non-active switch path or paths, i.e. paths not passing light at time of testing or monitoring. Optionally, the testing method disclosed herein may comprise additional optical testing of other components of the optical switch system, such as fibers and pigtails. These are passive components of the optical switch system that need to be monitored as well to ensure the complete functionality status of the optical switch system.
To make sure the working signals are working according to the desired specifications, each fiber in
In optical switching systems that include pigtails, a final status check that needs to be done on both working and non-working (potentially future working) paths includes checking the pigtail (
In summary, the present invention discloses mechanisms and methods of monitoring the status of any optical switch, whether in use or not at the time it is monitored, using electrical means. All switches in a system, MZI switches being just exemplary, can be continuously monitored in real time, making sure they are functional, whether they are currently being used to pass light or whether they are unused at the moment but potentially required in future configurations requested from the optical switch. If an unexpected value is read by the sensing mechanism at any time, an appropriate notification and alarm is sent to the management of the optical switch, thus reporting the operational status of the optical switch to its user at all times.
While the description above refers in detail to MZI switches, the monitoring described is of course general and applicable to any type of optical switch. The only requirement is that the optical switch has attached an electrical or electronic element (such as a heater or actuator) that determines a switching configuration or state. For example, in MEMS optical switches, the switching may occur by micromirror movements, see e.g. U.S. Pat. No. 6,750,655 to Staple and U.S. Pat. No. 6,633,694 to Tew and the references therein, or by waveguide movements, see e.g. U.S. Pat. No. 6,647,168 to Hsu et al., U.S. Pat. No. 6,577,785 to Spahn et al., and PCT application PCT/IL01/00787 to Haronian. The movement results from actuation, normally electrostatic, therefore involving electrical/electronic drivers (e.g. comb drives), which, according to the present invention, represent an electrically addressable element. Tew '694 specifically mentions electrically addressable circuitry associated with his switches. The monitoring method described herein is then applicable to all such switches that incorporate an electrically addressable element by (for example) using PWM or other signals on the electrically addressable circuitry, checking the switch status at all times, whether the switch is in use or not.
Example for Electrical Monitoring of an Optical Switch Using PWM Sense Control
This example is a case in which an ARM central processor is controlling the entire optical switch, and an FPGA is used to control the PWM generation and monitoring. The example is taken from one already implemented embodiment of the invention. Others embodiments (not shown) have been implemented using a discrete logic circuit, a gate array, a dedicated fully custom chip, a DSP, a microprocessor and other control modules instead of an FPGA.
Sense Function
PWM control is achieved through a Sense Function that has the purpose of providing detection of errors in the heaters. The following describes an exemplary error detection algorithm.
Each PWM FPGA collects data from sensors relating to its drive pulses that pass through A/Ds or any other value comparing or measuring electronic circuit. Referring now to
If the PWM is meant to fall within a certain proximity (defined in the PWM Settling Time register) to one of the samples (“uncertainty window” below), the algorithm allows either value of the sample. The uncertainty windows define low and high PWM thresholds (measured in clocks). If a PWM width is shorter than the PWM Low Threshold, the PWM will be known as Lo. If a PWM width is longer than the PWM High Threshold, the PWM will be known as Hi. Otherwise, the PWM will be known as medium (see also the section on PWM Low/High Threshold registers below).
In addition, the algorithm checks the amplitude of the sample. The thresholds defining the expected amplitude of a sample are programmable in registers. Three thresholds are defined:
A sample below TH off is considered to be ‘0’. A sample between TH on and TH ill is considered to be ‘1’. A sample between TH off and TH on is considered “mid” and is only legal if the PWM is meant to fall within the uncertainty window. A sample above TH ill always reports an error. The terms ON and OFF refer to logic levels of the FET. Each range includes the upper threshold value but not the lower
(i.e. THoff<mid<=TH on)
The table below shows what the expected sampled values are for each type of PWM:
When a sample is found not to match the expected value, an error is generated for that switch in an Error Bit Register. In addition to the Error Bit Register, each switch will send notification to the managing entity to report the error to a higher level to be dealt with there as needed by the specific application.
All publications, patents and patent applications mentioned in this specification are herein incorporated in their entirety by reference into the specification, to the same extent as if each individual publication, patent or patent application was specifically and individually indicated to be incorporated herein by reference. In addition, citation or identification of any reference in this application shall not be construed as an admission that such reference is available as prior art to the present invention.
While the invention has been described with respect to a limited number of embodiments, it will be appreciated that many variations, modifications and other applications of the invention may be made.
The present invention is a continuation-in-part of U.S. patent application Ser. No. 10/744,016, filed Dec. 24, 2003, which is hereby incorporated herein by reference.
| Number | Date | Country | |
|---|---|---|---|
| Parent | 10744016 | Dec 2003 | US |
| Child | 11339626 | Jan 2006 | US |
