The present invention relates generally to the telecommunications and optical networking fields. More specifically, the present invention relates to an optical amplifier pre-emphasis and equalization method and an optical communications system incorporating the same. Advantageously, the optical amplifier pre-emphasis and equalization method of the present invention efficiently and cost-effectively alleviates the optical amplifier gain ripple penalty experienced in conventional optical communications systems.
In optical communications systems that use wavelength division multiplexing (WDM), multiple wavelengths of light are used to support multiple communications channels on a single fiber. Optical amplifiers, such as erbium-doped fiber amplifiers (EDFAs), Raman amplifiers, and the like, are used in these optical communications systems to amplify optical signals that experience attenuation over multi-kilometer fiber optic links. Due to the length of such multi-kilometer fiber optic links, which may be on the order of hundreds or thousands of kilometers, many of these optical amplifiers may be used.
The intrinsic gain spectrum of an EDFA, for example, is not flat. Unless the signal spectrum is equalized, different wavelengths of light are amplified by different amounts as they propagate through multiple optical amplifiers. This results in degraded bit error rate (BER) performance for some communications channels. This degradation is referred to as the “optical amplifier gain ripple penalty.” It comes from two different sources, namely erbium-doped fiber (EDF)/gain flattening filter (GFF) mismatch and the temperature dependent change in the gain spectrum of the optical amplifiers, and it is compounded by the use of many optical amplifiers, increasing linearly with the number of optical amplifiers used. The gain ripple penalty stems from the optical amplifiers and, to a lesser extent, from dispersion compensating modules (DCMs), the fiber plant, etc., and it is usually a reach-limiting constraint for an optical communications system.
Conventional approaches for alleviating the optical amplifier gain ripple penalty involve the use of a dynamic gain equalizer (DGE) or the like. The DGE has a loss spectrum that tracks the gain spectrum of the optical amplifiers, thereby introducing losses to offset gain peaks. Conventional approaches for alleviating the optical amplifier gain ripple penalty also involve the use of an optical performance monitor (OPM) or the like for the active measurement of communications channel signal gain ripple at a receiver or an intermediate point in the optical amplifier chain and a feedback mechanism to provide this communications channel signal gain ripple information to a controlling element, such as a variable optical attenuator (VOA), DGE, or the like. The drawback to these approaches is that OPM and DGE elements are relatively expensive, especially when used in low channel count optical communications systems.
Thus, what is needed in the art is an approach for alleviating the optical amplifier gain ripple penalty that does not necessarily use OPM and DGE elements, or feedback mechanisms, thereby limiting optical communications system cost.
In various exemplary embodiments, the present invention provides an optical amplifier pre-emphasis and equalization method that efficiently and cost-effectively alleviates the optical amplifier gain ripple penalty experienced in conventional optical communications systems. This optical amplifier pre-emphasis and equalization method includes storing measured communications channel signal gain ripple information, acquired during factory calibration, in the internal memory of each optical amplifier module. When the optical amplifiers are assembled into a chain, system software retrieves this communications channel signal gain ripple information from each optical amplifier module and computes the pre-emphasis or equalization required for each channel in order to obtain a flat signal-to-noise ratio (SNR) at a receiver. The optical amplifier pre-emphasis and equalization method also includes measuring the ambient temperature of each optical amplifier module and applying a correction based on the expected change in gain response of each optical amplifier. The optical amplifier pre-emphasis and equalization method further includes, for Raman amplifiers and the like, applying a fiber type, gain setting, GFF error, etc. correction based on the expected change in gain response of each optical amplifier. The optical amplifier pre-emphasis and equalization method relies on using VOAs at the sources and/or DGEs/reconfigurable blocking filters (RBFs) that are already deployed at optical add/drop multiplexer (OADM) nodes in the optical signal path to pre-emphasize or equalize the optical channels. In other words, the gain ripple is corrected by pre-emphasizing the input spectrum of the in-line amplifier (ILA) chain, or equalizing points in between, thereby improving the SNR of weak channels and reducing non-linearity on the “hot” channels.
In one exemplary embodiment of the present invention, an optical amplifier pre-emphasis and equalization method for alleviating the optical amplifier ripple penalty experienced in an optical communications system, such as a WDM optical communications system carrying a plurality of channels or wavelengths, includes, using system software disposed outside of an optical amplifier module comprising an optical amplifier disposed within the optical amplifier module, retrieving gain ripple information related to the optical amplifier from a memory disposed within the optical amplifier module; using the system software, computing the pre-emphasis or equalization required for each channel or wavelength (or for the optical amplifier chain) in order to compensate for gain ripple using the retrieved gain ripple information; and, using one of a VOA, an RBF, and a DGE, pre-emphasizing or equalizing each channel or wavelength. The method also includes, using the system software, computing the pre-emphasis or equalization required for each channel or wavelength in order to obtain a flat SNR at a receiver using the retrieved gain ripple information.
The method further includes, using the system software, retrieving temperature dependent gain ripple information related to the optical amplifier from the memory disposed within the optical amplifier module. The method still further includes, using the system software, retrieving ambient temperature information related to the optical amplifier from a sensor disposed one of within and adjacent to the optical amplifier module. The method still further includes, using the system software, computing the pre-emphasis or equalization required for each channel or wavelength (or for the optical amplifier chain) in order to compensate for temperature dependent gain ripple using the retrieved temperature dependent gain ripple information and the retrieved ambient temperature information. The method still further includes, using the system software, retrieving other gain ripple information related to the optical amplifier from the memory disposed within the optical amplifier module. Finally, the method includes, using the system software, computing the pre-emphasis or equalization required for each channel or wavelength (or for the optical amplifier chain) in order to compensate for other gain ripple using the retrieved other gain ripple information.
In another exemplary embodiment of the present invention, an optical communications system, such as a WDM optical communications system carrying a plurality of channels or wavelengths, executing an optical amplifier pre-emphasis and equalization algorithm for alleviating the optical amplifier ripple penalty experienced in the optical communications system, includes system software disposed outside of an optical amplifier module comprising an optical amplifier disposed within the optical amplifier module, wherein the system software is operable for retrieving gain ripple information related to the optical amplifier from a memory disposed within the optical amplifier module; wherein the system software is also operable for computing the pre-emphasis or equalization required for each channel or wavelength (or for the optical amplifier chain) in order to compensate for gain ripple using the retrieved gain ripple information; and wherein the system software is further operable for, using one of a VOA, an RBF, and a DGE, pre-emphasizing or equalizing each channel or wavelength. The system software is further operable for computing the pre-emphasis or equalization required for each channel or wavelength in order to obtain a flat SNR at a receiver using the retrieved gain ripple information.
The system software is further operable for retrieving temperature dependent gain ripple information related to the optical amplifier from the memory disposed within the optical amplifier module. The system software is still further operable for retrieving ambient temperature information related to the optical amplifier from a sensor disposed one of within and adjacent to the optical amplifier module. The system software is still further operable for computing the pre-emphasis or equalization required for each channel or wavelength (or for the optical amplifier chain) in order to compensate for temperature dependent gain ripple using the retrieved temperature dependent gain ripple information and the retrieved ambient temperature information. The system software is still further operable for retrieving other gain ripple information related to the optical amplifier from the memory disposed within the optical amplifier module. Finally, the system software is operable for computing the pre-emphasis or equalization required for each channel or wavelength (or for the optical amplifier chain) in order to compensate for other gain ripple using the retrieved other gain ripple information.
The present invention is illustrated and described herein with reference to the various drawings, in which like reference numbers denote like method steps or system components, as appropriate, and in which:
As described above, conventional approaches for alleviating the optical amplifier gain ripple penalty involve the use of a gain equalization filter or the like. The gain equalization filter has a loss spectrum that tracks the gain spectrum of an optical amplifier, thereby introducing losses to offset gain peaks. Conventional approaches for alleviating the optical amplifier gain ripple penalty also involve the use of an OPM or the like for the active measurement of communications channel signal gain ripple at a receiver or an intermediate point in the optical amplifier chain and a feedback mechanism to provide this communications channel signal gain ripple information to a controlling element, such as a VOA, DGE, or the like. The drawback to these approaches is that OPM and DGE elements are relatively expensive, especially when used in low channel count optical communications systems.
In various exemplary embodiments, the present invention provides an optical amplifier pre-emphasis and equalization method that efficiently and cost-effectively alleviates the optical amplifier gain ripple penalty experienced in conventional optical communications systems. This optical amplifier pre-emphasis and equalization method includes storing measured communications channel signal gain ripple information, acquired during factory calibration, in the internal memory of each optical amplifier module. When the optical amplifiers are assembled into a chain, system software retrieves this communications channel signal gain ripple information from each optical amplifier module and computes the pre-emphasis or equalization required for each channel in order to obtain a flat SNR at a receiver. The optical amplifier pre-emphasis and equalization method also includes measuring the ambient temperature of each optical amplifier module and applying a correction based on the expected change in gain response of each optical amplifier. The optical amplifier pre-emphasis and equalization method further includes, for Raman amplifiers and the like, applying a fiber type, gain setting, GFF error, etc. correction based on the expected change in gain response of each optical amplifier. The optical amplifier pre-emphasis and equalization method relies on using VOAs at the sources or DGEs/RBFs in the optical signal path to pre-emphasize or equalize the optical channels. In other words, gain ripple is corrected by pre-emphasizing the input spectrum of the ILA chain, thereby improving the SNR of weak channels and reducing non-linearity on the “hot” channels.
As described above, the optical amplifier gain ripple penalty stems from the optical amplifiers and, to a lesser extent, from DCMs, the fiber plant, etc., and it is usually the limiting constraint for an optical communications system. All optical amplifiers use GFFs to flatten out their gain response as a function of wavelength. These GFFs do not perfectly equalize the gain response of the optical amplifiers perfectly due to physical and manufacturing constraints. Every optical amplifier has some residual gain ripple due to this imperfect gain response flattening. This residual gain ripple is unique to each optical amplifier, is largely deterministic, and varies little over the life of the optical amplifier.
Referring to
The above approach allows for the compensation of other effects as well. Preferably, this compensation is combined with the above. Referring to
The following example is instructive. Given a plurality of optical amplifiers, AMP 1, AMP 2, AMP 3, . . . , AMP M, . . . , AMP N−2, AMP N−1, and AMP N, and considering a source channel (Ck) added at AMP 1 (source) and dropped at AMP N (sink), where the source and sink can be end nodes, add/drops (ADs), or select terminals, each optical amplifier has a gain ripple spectrum GM(C) and TM(C). GM(C) is the gain ripple as a function of channel number and is defined by and stored in AMP M. TM(C) is calculated for each amplifier based on the ambient temperature read-back for AMP M. The correction applied to the VOA for channel Ck is given by:
Wk=−1/(N−1)*{(G1(Ck)+T1(Ck)*(N−1)+(G2(Ck))*(N−2)+ . . . (GM(Ck)+TM(Ck))*(N−M)+ . . . (GN−1(Ck)+TN−1(Ck))*(1)+(GN(Ck)+TN(Ck))*(N−N)},
where Wk is the attenuation applied to channel Ck at the MUX VOA (source) at AMP 1 in order to equalize channel Ck at AMP N. Wk can be positive or negative.
The following example is also instructive. Channel Ck passes through multiple SOADM/DGE (equalizing) nodes that are used to equalize channel Ck. If there are I equalizing nodes in the system located at n1, n2, n3, . . . , nI nodes, respectively, where n1=1 and I<N, I different weights can be calculated for channel Ck corresponding to each node. These weights are given by:
The weights satisfy the relationship:
WT=1/(N−1)*{(N−n1)*Wn1+(N−n2)*Wn2+(N−n3Wn3+ . . . (N−nI)*WnI }.
This is the ideal distribution of the equalizing weights for the VOAs and the RBFs. However, these equalizing weights may have to be redistributed if there is not enough dynamic range for the VOAs or the RBFs. The gain ripple values G1 . . . Gn are based on amplifier type (ILA-2, ILA-2 with Raman amplifier, etc.) and correspond to the gain ripple that channel Ck experiences through the optical amplifier chain due to optical amplifier ripple and temperature variations. The gain ripple that channel Ck experiences is normalized with respect to the mean gain of the optical amplifier. A positive value corresponds to a higher gain at the wavelength as compared to the mean gain, and a negative value corresponds to a lower gain at the wavelength as compared to the mean gain. Optionally, in a working system, this calculation is performed only time, three times in a modeling case: G1 . . . Gn=0 (corresponding to the mean value), G1 . . . Gn=maximum (corresponding to the high value), and G1 . . . Gn=minimum (corresponding to the low value). The pre-emphasis weights Wn1 . . . WnI are zero for the mean case, negative (channels are attenuated) for the high case, and positive (channels are amplified) for the low case. Any tilt in the amplifier due to high span loss will apply equally in all three cases.
The pre-emphasis and equalization algorithm is a single pass algorithm. The pre-emphasis and equalization weights are calculated from the launch side for each sub-span. Any under-equalization at an upstream sub-span is optimized at a downstream sub-span, if RBF margin is available.
The above approach allows for the compensation of other effects as well, such as fiber type, gain setting, GFF error, etc. Preferably, this compensation is combined with the above. For example, the ripple contribution from a Raman amplifier is a function of fiber type and gain setting, in addition to the residual ripple from the GFF. The effects are deterministic and the optical amplifier ripple can be compensated using the above techniques. Referring to
Advantageously, the optical amplifier pre-emphasis and equalization method of the present invention does not require any hardware components that are not already part of the optical communications system and provides a 30 to 50% improvement in the reach of long haul optical communications systems.
Although the present invention has been illustrated and described herein with reference to preferred embodiments and specific examples thereof, it will be readily apparent to those of ordinary skill in the art that other embodiments and examples may perform similar functions and/or achieve like results. All such equivalent embodiments and examples are within the spirit and scope of the present invention and are intended to be covered by the following claims.
Number | Name | Date | Kind |
---|---|---|---|
5225922 | Chraplyvy et al. | Jul 1993 | A |
6359726 | Onaka et al. | Mar 2002 | B1 |
6731424 | Wu | May 2004 | B1 |
6738183 | Shigeharra | May 2004 | B2 |
6839160 | Tsuda et al. | Jan 2005 | B2 |
6956687 | Moon et al. | Oct 2005 | B2 |
7006278 | Sakamoto et al. | Feb 2006 | B2 |
20020154359 | Tsuda et al. | Oct 2002 | A1 |
20030184843 | Moon et al. | Oct 2003 | A1 |
20050100271 | Frankel | May 2005 | A1 |
Number | Date | Country | |
---|---|---|---|
20080018990 A1 | Jan 2008 | US |