BRIEF DESCRIPTION OF THE DRAWINGS
Embodiments of the present invention will be further described with reference to the accompanying exemplary drawings, in which:
FIG. 1 illustrates a network comprising optical nodes interconnected by wavelength-division-multiplexed (WDM) links, devices for modulating optical carriers with distinctive identifying signals, and devices for controlling the power levels of the modulated optical carriers in accordance with an embodiment of the present invention;
FIG. 2A illustrates a set of links connecting two optical nodes where each link has only one optical carrier, for use in illustrating an embodiment of the present invention;
FIG. 2B illustrates a WDM link connecting two optical nodes in the network of FIG. 1 where crosstalk due to scattering is significant;
FIG. 3 illustrates a WDM link in the network of FIG. 1 connecting two optical switching nodes and traversing two passive transmission nodes;
FIG. 4 illustrates a segment of the link in network 100 provided with a wavelength selective gain-control device coupled to a local processing unit for determining carrier power in accordance with an embodiment of the present invention;
FIG. 5 illustrates crosstalk power components within a WDM link transporting tone-modulated composite carrier signals for use with an embodiment of the present invention;
FIG. 6 illustrates application of selective gain to a composite carrier signal to determine carrier power at a monitoring point along a WDM link in accordance with an embodiment of the present invention;
FIG. 7 illustrates the contribution of optical-amplifier crosstalk to the overall crosstalk in a WDM link in which crosstalk due to optical-power scattering is present;
FIG. 8 is a flow chart illustrating a method of measuring optical power of individual optical carriers in the presence of crosstalk in accordance with an embodiment of the present invention;
FIG. 9 details a step, in the method of FIG. 8, of measuring response to selective optical-power perturbation and determining carrier power and crosstalk power;
FIG. 10 details an alternative step, in the method of FIG. 8, of measuring response to selective optical-power perturbation and determining carrier power and crosstalk power;
FIG. 11 illustrates computation of carrier power and crosstalk power of a selected carrier in accordance with an embodiment of the present invention;
FIG. 12 illustrates an alternate computation of carrier power and crosstalk power of a selected carrier in accordance with an embodiment of the present invention;
FIG. 13 is a flowchart describing a process executed by a network controller for optical power control of optical-fiber spans across an optical network employing WDM fiber-optic links in accordance with an embodiment of the present invention;
FIG. 14 details a process executed by the network controller for optical-power control of all wavelength channels in a selected span in accordance with an embodiment of the present invention; and
FIG. 15 illustrates a process executed by a local controller of a fiber span for estimating carrier-signal power and determining a correction factor in accordance with an embodiment of the present invention.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
Terminology
Optical switching node: An optical switching node is a device for receiving optical signals from at least two input channels and directing each received optical signal to one of at least two output channels. The optical switching node considered herein may take the form of a space switch or a wavelength-channel add-drop multiplexer. The input channels may comprise channels arriving from another switching node and sharing an input wavelength-division multiplexed (WDM) link as well as channels from local data sources. The output channels may comprise channels directed to another switching node and sharing an output WDM link as well as output channels leading to local data sinks.
Optical transmission node: An optical transmission node is a passive device used for processing signals received from input channels and placing the processed signals to output channels having one-to-one correspondence to the input channels. The signal-processing function may include overall amplification or wavelength-selective gain control.
Wavelength-tracking unit: A wavelength-tracking unit used in the network of the present invention has a primary function of ensuring proper routing of each carried optical signal. The unit may be further provided with a capability to measure optical power.
Wavelength-selective gain-control device: The wavelength-selective gain-control device used in the network of the present invention is provided with a capability to apply a predetermined gain (or attenuation) to a selected carrier signal or to any subset of carrier signals.
Carrier signal: The term carrier signal is used to indicate an optical signal of a single frequency (single wavelength).
Information signal: An information signal refers to a “payload” signal, modulating a carrier signal. A typical information signal is a digital signal of 2.5 or 10 Gigabits per second.
Identifying signal: An identifying signal is a specific signal used to modulate a carrier signal in order to identify the carrier signal as it propagates through switching node to ensure proper routing.
Signature: An identifying signal is often referenced as a “signature” to emphasize its role in tracking (modulated) carrier signals.
Tone: An identifying signal is conveniently devised as a low-frequency sinusoidal signal called “tone”. The term is also used herein to refer to an identifying signal comprising multiple low-frequency sinusoidal signals which may be time-interleaved.
Tone band: A tone comprising multiple-sinusoidal signals, however organized, may occupy a (low) frequency band which may be contiguous or spread.
Composite carrier signal: A carrier signal modulated by both an identifying signal and an information signal is called a “composite carrier signal”, or “a composite signal”. Modulation by the identifying signal should not interfere with detection of the information signal.
Wavelength channel: A composite carrier signal occupies an optical-frequency band (a wavelength band) within the bandwidth of a fiber link. The occupied band is loosely called a “wavelength channel” or a “channel”.
Link: A link is a transmission medium connecting one switching node to another switching node. A WDM link may comprise several wavelength channels.
Span: A link may comprise a concatenation of fiber-optic spans joined at passive transmission nodes.
A path: A path comprises a wavelength channel in each of a series of links connecting a first (source) switching node to a second (destination) switching node.
Optical Network with Channel Monitoring Capability
FIG. 1 illustrates a network 100 in which the present invention may be deployed. The network 100 comprises a plurality of optical switching nodes 120 interconnected by optical wavelength-division-multiplexed (WDM) fiber links 125. A fiber link 125 exceeding a predetermined length may traverse passive transmission nodes for signal enhancement. The passive transmission nodes divide the link into a number of spans as will be illustrated in FIG. 3. A fiber link connecting a first switching node to a second switching node may comprise multiple spans where a first span connects the first switching node to a first passive transmission node. If the fiber link traverses only one passive transmission node, a second span connects the first passive transmission node to the second switching node. Otherwise, the second span connects the first passive transmission node to a second passive transmission node, and a third span connects the second passive transmission node to either the second switching node or to a third passive transmission node, if any, and so on.
An optical switching node 120 may be an edge node directly connected to signal sources and sinks or a core optical node for routing optical signals among edge nodes. The network 100 may comprise several tracking units 135 to ensure routing sanity. A tracking unit 135 may also be provided with power-measurement devices to ensure appropriate signal power and wavelength-selective gain-control devices to provide power control where needed. At least one tracking unit 135 may be placed along a path from one switching node 120 to another. A global network controller 160, which may be geographically distributed in the form of well-coordinated control nodes, is communicatively coupled to each of the tracking units 135 either directly through links 162 or through some external network. FIG. 1 illustrates two communicatively coupled network controllers 160-1 and 160-2.
A path from an optical switching node 120 to another may traverse several links 125 where the path includes one wavelength channel in each link. An optical node 120 may perform wavelength channel switching, including simple add-drop multiplexing, and may monitor and adjust optical-signal levels.
The channels of each link 125 are individually and uniquely modulated by identifying signals, also called tones, dither tones, or signatures. The identifying signals of wavelength channels within a link are distinct. An identifying signal associated with a wavelength channel may carry information identifying the channel for tracking purposes. An identifying signal may also indicate signal intensity at different points along a WDM link and may be devised to carry other control information. In a first channel-identification scheme, as illustrated in FIG. 1, each end-to-end path in the network 100 is given a distinct signature and, consequently, a specific signature may appear in two or more links traversed by an end-to-end path. The signatures are indicated as s1, s2, . . . , s804. The required number of distinct signatures is then at least equal to the number of end-to-end paths. In a second channel-identification scheme, a signature may not appear in more than one link. The number of required signatures is then substantially higher than that of the first scheme. Several techniques may be employed to increase the number of available signatures using combinations of identifiers. A simple identifier is a low-frequency tone having a frequency within a frequency band or occupying non-contiguous frequency bands carefully selected to meet several criteria.
In both the first and second channel-identification schemes, the global network control node 160 determines and assigns the individual signatures and distributes relevant information to optical switching nodes 120 of network 100 as well as to any intervening passive optical transmission nodes as illustrated in FIG. 3. In the second channel-identification scheme, the global network controller 160 would have the extra burden of interrelating signatures of channels constituting an end-to-end path.
A local controller operating in the electronic domain (not illustrated) is associated with each optical switching node and each optical transmission node that participate in the power-monitoring system. A network controller 160 may communicate directly with selected local controllers.
In FIG. 1, a path from node 120A to node 120D traverses nodes 120B and 120C. The path is identified by a signature s1 which appears in all links 125 connecting optical node 120A to node 120D. A path from node 120A to node 120E is identified by a signature s3 which appears in the three links connecting node 120A to node 120E.
Regardless of the scheme used for channel-identification, the channels of each link 125 are monitored to ensure proper routing and proper signal levels.
FIG. 2A illustrates m>1 wavelength channels 215 connecting a first optical node 120X to a second optical node 120Y in a hypothetical arrangement where each of the m channels exclusively occupies a fiber link 215 connecting a port 214 in node 120X to a port 218 in node 120Y. Thus, m links are needed to connect the first node 120X to the second node 120Y. The m channels are centered at wavelengths λ1, λ2, . . . , λm. A channel Cj, 1≦j≦m, carries u1 watts at the sending end (the first node 120X). Channel C delivers Wj watts to the second node 120Y. With uj>0, the ratio Wj/uj is determined by the intrinsic attenuation of link j and any possible amplification along link j. The link attenuation (or amplification) is considered time-invariant. There is no interaction among the m channels and, hence, power measurements at the second node may be used directly to adjust the power level at the input of any of the m links.
In FIG. 2B, spectral multiplexer 224 multiplexes the m channels onto a common link 225. If the attenuation and/or gain of each channel Cj, 1≦j≦m, in the common link 225 is the same as that of link j in the arrangement of FIG. 2A, and if the m wavelengths are spaced to render crosstalk negligible (implying a small number m of channels), then the power carried by channel Cj at the second node would be substantially equal to the power Wj delivered by link j in the arrangement of FIG. 2A. When crosstalk is present, the power—denoted W*j—carried by channel Cj at the second node 120 differs from Wj by an amount determined by the crosstalk power received from channels of wavelengths shorter than λj, if any, and crosstalk power transferred to channels of wavelengths greater than λj, if any. A spectral demultiplexer 228 may be used at node 120Y to separate the multiplexed carrier signals and a power-measuring device deployed at the second node 120Y would measure the values W*j, 1≦j≦m. The contribution of crosstalk to the carrier power W*j of channel Cj at the second node 120Y is not directly known.
A preferred method of measurement uses the identifying signals (also called tones or signatures) described above with reference to FIG. 1 to monitor power levels in any span in any link 125.
FIG. 3 illustrates a WDM link 125AB in the network of FIG. 1 connecting two optical switching nodes 120A and 120B and traversing two passive transmission nodes 320-1 and 320-2 dividing the link into three spans: a first span (span-1) connecting optical switching node 120A to passive transmission node 320-1; a second span (span-2) connecting the first transmission node 320-1 to the second transmission node 320-2; and the third span (span-3) connecting the second transmission node 320-2 to the second switching node 120B. Optical-power control may be applied separately for each of the three spans, to a concatenation of any two spans, or to the entire WDM link 125AB. Wavelength tracking units 135 may be provided at all or at a subset of nodes 120A, 320-1, 320-2, and 120B. A tracking unit 135 ensures proper routing of wavelength channels by detecting identifying signals of composite carrier signals propagating through a link and comparing with a reference list of expected identifying signals. A tracking unit may also include a power detector to determine signal power and a wavelength-selective gain-control device to adjust signal power levels. Signal power levels are adjusted either “permanently” for operational purposes or temporarily for executing the methods of the present invention for accurate power monitoring in the presence of crosstalk as will be detailed below. For example, the channel-routing function in network 100 assigns wavelength channels C1, C2, . . . , Cm, of carrier wavelengths λ1, λ2, . . . , λm, to link 125AB. The m wavelength channels are expected to be present and at appropriate power levels in each span in link 125AB because each of transmission nodes 320-1 and 320-B is a passive node which may perform signal processing but does not switch a signal to a different path. A tracking unit 135-1 provided at a designated output port of switching node 120A verifies that the m channels have been switched properly in switching node 120A and a tracking unit 135-2 provided at transmission node 320-2 verifies that the m channels have maintained appropriate power levels through span-1 and span-2.
FIG. 4 illustrates a segment of a link 125 in network 100 traversing fiber spans 430, 440, and 450 and including two optical amplifiers 432A and 432B. A first wavelength-selective gain-control device 428 receives a WDM optical signal from a fiber link 420 and may adjust the power level of any carrier signal if required. Device 428 is illustrated schematically to include a component 422 for isolating a specific (modulated) optical carrier signal, a component 426 for applying a desired gain or attenuation to the specific optical signal, and a combiner 424. An optical-signal tapping device 462 is positioned following the second span 440. A second wavelength-selective gain-control device 458 is provided in the vicinity of, but preceding, the tapping device 462. Components 452, 456, and 454 are similar to components 422, 426, and 424, respectively. The device 458 is controlled by a local controller 490 which selects a value of gain (or attenuation) to be applied to a selected wavelength channel. Local controller 490 has a control path 492 to the second wavelength-selective gain-control device 458. Local controller 490 communicates with the first wavelength-selective gain-control device 428 through a network controller 160.
The received signal from incoming fiber link 420 comprises a number of carrier signals each carrier signal modulated with an identifying signal in addition to an information signal. Modulating a carrier signal by an identifying signal may take place at a node, preceding fiber link 420, presented in FIG. 4 as node 416. Node 416 may not be directly connected to fiber link 420. Link 420 may, however, carry un-modulated carrier signals in addition to the modulated carrier signals. Device 428 may be incorporated in an optical switching node 120, which may be a space switch or an add-drop multiplexer. The output of device 428 comprises m modulated carrier signals propagating through fiber-optic spans 430, 440 and 460 towards some destination. The monitoring device 462 taps a portion of the optical signal and the tapped signal is supplied via a (short) link 464 to a low-frequency detector 470. The detected low-frequency signal, which comprises tone signals of individual carrier signals and other low-frequency components resulting from cross modulation, is supplied to a signal analyzer 480 through an internal path 474.
The ratio θ of tapped-signal power to the propagating signal power is precisely determined. A typical value of the ratio θ is 0.05. It is important to note that any tone-power measurement obtained by any means, analog or digital, is preferably automatically adjusted according to the ratio θ to yield the actual tone power of the corresponding composite signal propagating through the monitored link. Alternatively, the ratio θ may be incorporated in a ratio of tone-signal power in a tapped-signal to actual carrier power.
A low-frequency signal analyzer 480 may measure the power of any of the tone signals. The measurements are supplied to a local controller 490 through internal path 484, and the local controller 490 communicates the measurements, or a power correction factor based on the measurements, to a network controller 160. A power correction factor is determined by comparing a power measurement with a predetermined objective power level. Controller 160 has a control path (not illustrated) to the first wavelength-selective gain-control device 428, as well as control paths to other monitoring devices as illustrated in FIG. 1. The signal analyzer 480 may be an analog instrument or a digital instrument employing discrete spectral-analysis techniques. If an analog instrument is used, its output is converted to digital format which is directed to the local controller 490. If the instrument is digital, analog-digital conversion precedes the measurements. Any of available software programs for evaluating the discrete Fourier transform of a sampled signal may be used. A typical program uses a “Fast-Fourier-Transform” algorithm which lends itself easily to digital realization. The derivation of the discrete Fourier transform may be performed in either the signal analyzer 480 or the local controller 490. The local controller 490 further implements the methods of the invention to accurately estimate individual carrier power levels at the monitoring point and transmit the results to a respective network controller which instructs the first wavelength-selective gain-control device 428 to adjust carrier-signal power accordingly.
Crosstalk Spectrum in a Tone-Modulated WDM Link
It is well known that light scattering within a WDM fiber link results in crosstalk where energy is transferred from a carrier signal to another carrier signal of longer wavelength. This property is known as the Raman scattering phenomenon. The transfer of energy from a first carrier signal amplitude modulated by a first tone to a second carrier signal amplitude modulated by a second tone results in attenuation of the first carrier signal together with its modulating first tone and amplification of the second carrier signal together with its modulating second tone. In addition, the fluctuating amplitudes of the two carrier signals result in the first carrier signal being further modulated by the second tone and the second carrier signal being further modulated by the first tone. With a large number of carrier signals sharing a WDM link, the spectrum of the envelope of each carrier signal would comprise the frequencies of the tones of all carrier signals in addition to their cross-modulation products.
FIG. 5 illustrates crosstalk power components within a WDM link transporting tone-modulated composite carrier signals.
The crosstalk among the tone-modulated carrier signals produces composite carrier signals each having an envelope containing the tone-frequencies of all other carriers in addition to the pair-wise sums and differences of the tone frequencies, as illustrated in FIG. 5 which illustrates crosstalk products for a WDM link having six wavelength carriers of wavelengths λ1, λ2, λ3, λ4, λ5, λ6, where λ1>λ2>λ3>λ4>λ5>λ6. Each wavelength carrier is modulated with a respective information signal and further amplitude-modulated by a distinctive identifying tone to produce a composite carrier signal. For brevity, each composite carrier signal is said to occupy a “wavelength channel”, or simply a “channel”, where each channel corresponds to a wavelength band (an optical-frequency band) centered on a corresponding wavelength (a corresponding optical frequency). A channel centered on wavelength λj is referenced as Cj. For each composite carrier signal, the tone power bears a predetermined ratio to the carrier power. The tone signals for the six wavelength channels λ1 to λ6 have tone-frequency bands Φ1, Φ2, Φ3, Φ4, Φ5, and Φ6, respectively. Other resulting low frequencies are grouped in one column with entries marked ‘x’. A frequency band Φj, 1≦j≦6, may contain a single discrete frequency or a mix of multiple discrete frequencies, and the frequency band is selected to meet certain criteria; for example a frequency band should not overlap the frequency spectrum of information signals modulating the carrier signal. The discrete frequencies of different tone-frequency bands may, however, interleave.
Consider, for example, a composite carrier signal of wavelength λ4 where, at the sending end, the carrier power is W*4 and the tone power is p*4. Without loss of generality, the WDM link may be considered to be ideal, offering unity gain (zero decibels) for each of the composite-signals. This simplification will be reconsidered below. With unity gain, the power received at the monitoring point related to wavelength channel C4 assumes a value of W4, where the difference (W*4−W4) may be positive, if, due to scattering, channel C4 receives more power from channels C5 and C6 than it imparts to channels C3, C2, and C1, or negative otherwise. The tone power of the composite carrier signal received at the monitoring point would be p4 where the difference (p*4−p4) may be positive or negative. As illustrated in FIG. 5, wavelength channel C4 imparts to wavelength channels C3, C2, and C1 tone power, in frequency band Φ4, of values β4,3, β4,2, and β4,1, respectively. Likewise, channel C5 imparts to wavelength channel C4 tone power, in frequency band Φ5, of value β5,4, and channel C6 imparts to wavelength channel C4 tone power, in frequency band Φ6, of value β6,4.
Because the composite carrier signals of channels C1, C2, and C3 are tone-modulated with frequency bands Φ1, Φ2, Φ3, the transfer of power from channel C4 to channels C1, C2, and C3 results in amplitude modulation of the carrier signals of channel C4, with tones in frequency bands Φ1, Φ2, Φ3, and the power of channel C4 contains corresponding tone power of values δ1,4, δ2,4, and δ3,4, respectively. Likewise, because the composite carrier signal of channels C4 is tone-modulated with frequency band Φ4, the transfer of power, from channels C5 and C6 to channel C4, results in amplitude modulation of the carrier signals of channels C5, and C6 with tones in frequency band Φ4, with corresponding tone power of values δ4,5, δ4,6, respectively.
A measurement of tone power in frequency band Φ4 would then yield a power π4 where π4=(p4+q4), q4 being the combined crosstalk in frequency band Φ4 within channels C1, C2, C3, C5, and C6 (excluding channel C4), i.e.,
q
4=β4,3+β4,2+β4,1+δ4,5+δ4,6.
The tone power component p4 within channel C4 is indicative of the carrier power W4 received at the monitoring point.
In FIG. 5, the power components identified in row 520 constitute the total power within the wavelength-band (optical-frequency band) occupied by channel C4. The power components identified in column 530 constitute the total power π4 (reference 540) in low-frequency tone band Φ4 as seen by the low-frequency detector 470, and after accounting for instrumentation parameters, such as the proportion of optical-signal power tapped by optical-signal tapping device 462.
In order to isolate the crosstalk power qj, in accordance with an embodiment of the present invention, a temporary gain is applied to the signal of channel Cj and the corresponding change in the power of channel Cj is measured. It is essential, however, that the temporary gain be discriminatory, being applied to either of the components pj or qj, but not both. In a preferred embodiment according to the present invention, the total power of the composite carrier signal of channel C4 is multiplied by a predetermined factor (1+α) where −1<α<â, and |α|<{hacek over (a)}. The upper bound â is constrained by crosstalk tolerance and the lower bound {hacek over (a)} is based on the sensitivity of detection and measurement instruments.
The power πj, (πj=pj+qj), of channel Cj is measured at a monitoring point (the second optical node) under normal load condition, or generally under a current load condition. When a discriminating gain (attenuation) (1+α) is applied to only pj, the measured power of channel Cj at the second node changes to π*j=(1+α)pj+qj.
FIG. 6 illustrates application of selective gain, in accordance with an embodiment of the present invention, to a composite carrier signal to determine carrier power at a monitoring point along a WDM link. As illustrated in FIG. 6, a measurement of tone power in frequency band Φ4 after the induced temporary power change of channel C4 would yield a power π*4 where π*4=(1+α) p4+q4, where q4 is the combined crosstalk within frequency band Φ4 in channels C1, C2, C3, C5, and C6, excluding channel C4, as determined above.
In FIG. 6, the power components identified in row 620 constitute the total power within the wavelength-band (optical-frequency band) occupied by channel C4 after introducing the temporary gain (1+α). Each power component in row 620 has a value (1+α) times its counterpart in row 520 of FIG. 5. The power components identified in column 630 constitute the total power π*4 (reference 640) in low-frequency tone band Φ4 with only the component p4 being modified by the factor (1+α).
Estimating Carrier Power at Monitoring Point
The tone power pj recovered at the monitoring point is determined as pj=(π*j−πj)/α. The power of the carrier-signal propagating in span 460 (FIG. 4) at the monitoring point is then determined as Wj=pj/Rj, Rj being the ratio of identifying-signal power (tone power) to the power of the carrier-signal in a composite carrier signal occupying wavelength channel Cj. This ratio is set at a sending end and remains unchanged as the composite carrier signal gains or loses power due to intrinsic link attenuation and scattering.
The ratio ηj=(π*j/πj) may be used to determine a ratio ρ=(qj/pj) as:
ρ=(qj/pj)=(1+α−ηj)/(ηj−1),
and the tone power pj in band Φj within channel Cj is determined as:
p
j=πj/(1+ρ).
Effect of Channel Attenuation
In the above description, the WDM link is considered to offer unity gain to each composite carrier signal. In reality, the gain is determined by the combined effect of intrinsic attenuation, which may be wavelength dependent, and optical-amplifier gain. The estimation of carrier-signal power is determined by applying wavelength-selective gain (or attenuation) at the monitoring point and, hence, preceding gain or attenuation is irrelevant.
Crosstalk Contribution of Passive Optical Nodes
FIG. 7 illustrates the combined effect of crosstalk caused by optical-power scattering and crosstalk occurring within optical amplifiers of the passive optical nodes 320 (FIG. 3). An optical signal propagating in a link having intermediate optical amplifiers, such as Erbium-Doped Fiber Amplifiers (EDFA), suffers from both scattering-induced crosstalk (such as the Raman Stimulation crosstalk) and internal crosstalk within each amplifier. Thus, a detected tone signal at a monitoring point would be affected by the two crosstalk types. It is noted, however, that the power gain or loss of a tone signal is predominantly caused by the scattering crosstalk. Consider a tone of a single (low) frequency having an amplitude (intensity) normalized to unity. Scattering crosstalk may increase (or decrease) the amplitude of the tone with a relative (normalized) increment (or decrement) of ε1 to result in a corresponding power change proportional to (1±ε1)2. The crosstalk (of relative magnitude ε2) generated within an EDFA is orthogonal to the scattering crosstalk as illustrated in FIG. 7. The vector diagrams of FIG. 7 illustrate intensity level of a tone signal in the presence of the two types of crosstalk. With an orthogonal relative amplitude increment of ε2, (|ε1|<<1 and (|ε2|<<1), the total power is proportional to ((1±ε1)2+ε22) as illustrated in FIG. 7, with an insignificant relative power change due to EDFA crosstalk of approximately ε22. The relative power change due to scattering crosstalk is approximately (±2×ε1).
The contribution of scattering crosstalk and EDFA crosstalk is illustrated in FIG. 7 where tone-measurement error is plotted as a function of Raman crosstalk levels and EDFA crosstalk levels. It is seen that a contour 740 indicates minor measurement change as EDFA crosstalk varies from −10 decibels to −40 decibels. A contour 760 indicates a significant increase in measurement error as the Raman crosstalk increases from −40 decibels to −10 decibels.
The disparity of crosstalk contributions would be of the same order when a tone comprises multiple interleaved sinusoidal signals or other distinctive signals instead of the single sinusoidal wave discussed above.
Determining Carrier Power from Tone-Power Measurements
FIG. 8 illustrates the method, according to the present invention, of determining carrier power, at a monitoring point along a WDM link, of a selected composite carrier signal from among m composite carrier signals sharing a WDM link. The method is preferably implemented by local controller 490 (FIG. 4). However, variations of the procedure of FIG. 8 may divide the computational effort between local controller 490 and a respective network controller 160.
In step 810, tone power πj of each channel Cj, 1≦j≦m, is measured. Estimation of the individual carrier power for the m channels may be performed in any selected order. In step 812, a channel Cj is selected and in step 814 the carrier power πj of channel Cj, as determined in step 810, is compared with a predetermined threshold. If the tone power πj is found to be below the threshold, the tone power and carrier power of channel Cj are considered inadequate for further processing. In step 820, an indication of an indefinite measurement is formed and the processing is directed to step 840. If the tone power πj exceeds the threshold, step 825 is activated to execute the process of the selective power modification of the present invention and estimation of power in channel Cj. The result of step 825 is directed to step 840 and local controller 490 reports the computed tone power, carrier power, and crosstalk power to a respective network controller 160. Step 840 directs the power-estimation process to step 824 which determines whether to select a subsequent wavelength channel Cj (if any) to process a corresponding composite-carrier signal, or to indicate that the process is complete.
FIG. 9 details a first implementation of step 825 of FIG. 8. In step 920, a temporary power gain (1+α) is applied to channel Cj. The method of application of the gain depends on the structure of the detection and monitoring devices used at the monitoring point. The value of α is selected to be below a predetermined upper bound to limit the impact on other wavelength channels. A negative value of α, indicating attenuation, may be used, but should not be close to unity, otherwise subsequent measurements related to channel Cj would be unreliable. The magnitude of α, regardless whether α is positive or negative, should be above a certain lower bound {hacek over (a)} in order to produce a measurable effect. The value of {hacek over (a)} depends to a large extent on the sensitivity of the measuring devices.
In step 930, the tone power π*j in (low) frequency band Φj is measured and the measurement is used in step 940 to compute the power of both the tone and the carrier signals of the composite carrier signal of channel Cj.
FIG. 10 illustrates an alternate implementation of step 825 (referenced as 825A) of FIG. 8. In step 1020, a temporary attenuation (1−α) is applied to the composite carrier signal of channel Cj and in step 1022 the tone power in frequency band Φj is measured to yield a value πj−. In step 1024, a subsequent temporary gain (1+α) is applied to the unperturbed composite carrier signal of channel Cj and in step 1026 the tone power in frequency band Φj is measured to yield a value πj+. In step 1030, the change in tone power level in frequency band Φj is determined as π*j=(πj+−πj−)/2. In step 1040 the carrier power and crosstalk power of both the tone and the carrier signals of the composite carrier signal of channel Cj are computed. Step 1040, to be described below, is quite similar to step 940.
FIG. 11 details a first implementation of step 940 of FIG. 9. Step 940 is reached from step 930 of FIG. 9 which provides a measurement π*j of tone power after applying the temporary gain. In step 1140, the power pj of the tone in frequency band Φj is determined as pj=(π*j−πj), and the tone crosstalk power qj is determined as qj=πj−pj. In step 1180, the carrier power Wj is determined as pj/Rj, Rj being the ratio of tone power to carrier power of the composite carrier signal of channel Cj at the sending end. It is noted that the selection of the value of Rj may depend on the corresponding carrier wavelength λj, where the composite carrier signal of shortest wavelength (being predominantly a donor of scattered power) is given a higher ratio Rj, and generally Rj>Rk for λj<λk.
Step 1040 of FIG. 10 is similar to step 940 of FIG. 9. The main difference is that the value of π*j is determined as π*j=(πj++πj−)/2. The computation of pj in step 1040 may alternatively be determined as pj=(πj+−πj−)/(2×α).
FIG. 12 details a second implementation of step 940 (referenced as 940A) of FIG. 9. In step 1120, the ratio ρ of crosstalk power to carrier signal power is determined as: ρ=(απj+πj−π*j)/(π*j−πj). In step 1240 the carrier-signal power is determined as: Wj=πj/(Rj+ρRj), Rj being the ratio of tone power to carrier power of the composite carrier signal of channel Cj at the sending end as defined above.
Network-Wide Application of Carrier-Power Measurements
A WDM link connecting two optical switching nodes may comprise more than one span where a span connects any two nodes which may be switching nodes, transmission nodes, or one of each. The method described above applies to (1) a single span within a WDM link, (2) a path traversing several spans belonging to a single WDM link, (3) a path traversing several spans belonging to concatenated WDM links, or (4) to an end-to-end path from a source optical node 120 to a sink optical node 120.
The measurement process described in the present application requires an experimental temporary modification of a composite carrier signal over a link of interest. The level of power modification is judicially selected so as to avoid undue interference in any other carrier signal. Within a single link, only one carrier signal is processed at a time as described above. Applying the power-measurement procedure concurrently to any two composite carrier signals propagating in two different links where the two links have a common composite carrier signal may falsify the measurements, or at least reduce the accuracy of the respective carrier-power estimates. The accuracy of carrier-power estimation may also be reduced even if the two links do not have a common composite carrier signal. This occurs if at least one carrier signal in the first link and at least one carrier signal in the second link share an intermediate link. To ensure accurate measurements, the network controller 160 may schedule the measurements so that only one carrier signal is considered at a time. Alternatively, the network controller 160 may identify groups of links 125 (FIG. 1) each group comprising links 125 that are not directly or indirectly coupled. The links in each group may then be processed concurrently.
FIG. 13 describes a process executed by a network controller 160 for optical power control of optical-fiber spans across optical network 100. The controller 160 maintains a record for each link 125 with detailed information on the wavelength channels and spans of each link 125. The controller 160 determines an order of processing the spans of interest in network 100. In step 1320, controller 160 selects a link 125 and in step 1322, the wavelength channels of the selected links are identified. In step 1324, the controller 160 selects a span within link 125. In step 1326, controller 160 determines a specific order of processing the wavelength channels. In step 1328, controller 160 receives power measurements and/or an estimated carrier-signal power for a currently considered wavelength channel from a local controller (local controller 490 of FIG. 4) to be used for instructing a sending end to adjust power level of the transmitted carried signal if required.
If the link 125 under consideration has one or more spans to be considered, step 1324 is revisited (from step 1328). Otherwise, step 1328 directs the process to step 1320 to select a new WDM link in network 100. The process may execute continuously, thus, when all links 125 have been considered, the first link in the ordered list is revisited.
The links 125 may be arranged in a preferred order for cyclic processing. For example, the links 125 may be ranked according to the number of end-to-end paths traversing each link; an end-to-end-path being a path traversing links from a source optical node 120 to a sink optical node 120. A link shared by a large number of end-to-end-paths, or a link of high traffic occupancy, may be processed ahead of others.
The wavelength channels (equivalently, the composite-carrier signals) may be considered in step 1326 in a favorable order which takes into account the light-scattering properties. For example, a favorable order may be a descending order of wavelengths of the m composite carriers.
FIG. 14 details step 1326. In step 1420 controller 160 identifies a first local controller at the sending end of the selected span and a second local controller (local controller 490, FIG. 4) at the receiving end of the span. In step 1422, controller 160 selects a wavelength channel. In step 1426, controller 160 sends an index j of the selected channel Cj to the second local controller and in step 1430 it receives measurements, and/or an estimate of carrier-signal power for channel Cj, from the second local controller. In step 1432, controller 160 determines a power correction factor for channel Cj to be applied at the sending end of the span under consideration. If power correction is to be applied, the index j and the correction factor are sent to the first local controller. In step 1440, controller 160 receives a response from the first local controller indicating that a correction has been applied to channel Cj and step 1436 is executed. If, in step 1434 it is determined that a correction is not needed, step 1436 is executed. If step 1436 determines that all channels of the span under consideration have been considered a subsequent channel from the ordered list is selected in step 1424 and the steps starting with step 1426 and ending in step 1436 are executed again. If step 1436 determines that all wavelength channels have been considered, the process is directed to step 1328 of FIG. 13.
FIG. 15 illustrates a process executed by the second local controller (local controller 490 of FIG. 4) of the span under consideration for estimating carrier-signal power and determining the correction factor described above. The second local controller receives from a network controller 160 an index j of a carrier signal to be measured next (step 1512). In step 1514, the power πj of tone Φj is measured. In step 1516, if the measured power πj is found to be insufficient for further processing, step 1520 assigns a default value to pj and other variables to be reported in step 1540. In step 1516, if the measured power is found to be sufficiently large to permit further processing, the wavelength-selective gain-control device 458 applies a temporary gain to the identified carrier signal and the process described in step 825 of FIG. 8 is executed to compute the carrier-signal power. The second local controller maintains a table of objective carrier-signal power for each carrier signal routed through the monitored link or span. A ratio ηj of the computed carrier-signal power to a respective objective power is determined (step 1530) and reported (step 1540), together with the carrier-signal power Wj to the network controller 160 (step 1560). The network controller 160 instructs a controller at the sending end of the monitored link (or span) to adjust power level of the carrier signal.
Although specific embodiments of the invention have been described in detail, it should be understood that the described embodiments are intended to be illustrative and not restrictive. Various changes and modifications of the embodiments shown in the drawings and described in the specification may be made within the scope of the following claims without departing from the scope of the invention in its broader aspect.
Patent Application
20070280687
