CN1310872A - Gain tilt control with mid-stage attenuators in erbium-doped fiber amplifiers - Google Patents

Abstract

Gain tilt control in erbium-doped fiber amplifiers is realized by adjusting the attenuation of mid-stage variable attenuators situated between multiple stages of the erbium-doped fiber amplifier. Positive power tilt in input signals in compensated by increasing attenuator loss while negative power tilt is compensated by decreasing the attenuator loss.

Description

Gain Tilt Control with Intermediate Stage Attenuators in Erbium-Doped Fiber Amplifiers

The present invention generally relates to the field of optical communication, and specifically relates to an erbium-doped optical fiber amplifier in which there is a gain tilt control of an intermediate-stage attenuator.

With the development of voice and data networks, higher capacity transmission systems are required, leading to the development of multiple wavelength wavelength division multiplexing (WDM) optical communication systems with a large number of individual channels. (see, for example, Y. Sun, J.B. Judkins, A.K. Srivastava, L. Garret, J.L. Zyskind, J.W. Sulhoff, C. Wolf, R.M. Derosier, A.H. Gnauck, R.W. Tkach, J. Zhou, R.P. Espindola, A.M. Vengsarkar, and A.R. Chraplyvy , "Transmission over 640 km of 32-WDM 10-Gb/s Channels Using Broadband Flat-Gain Erbium-Doped Silica Fiber Amplifiers", IEEE Photon.Tech.Lett., Vol.9, No.12, pp.1652- 1654, December 1997; A.K.Srivastava, Y.Sun, J.L.Zyskind, J.W.Sulhoff, C.Wolf, J.B.Judkins, J.Zhou, M.Zirngibl, R.P.Espindola, A.M.Vengsarkar, Y.P.Li, and A.R.Chraplyvy, "True Wave by 520km Error-free transmission of 64 WDM 10-Gb/s channels over optical fiber", NEED PUBLICATION DATE, and A.K.Srivastava, Y.Sun, J.W.Sulhoff, C.Wolf, M.Zirngibl, R.Monnard, A.R.Chraplyvy, A.A.Abramov, R.P. Espindola, T.A. Strasser, J.R. Pedrazzini, A.M. Vengarkar, J.L. Zyskind, J. Zhou, D.A. Ferrand, P.F. Wysocki, J.B. Judkins, and Y.P. Li, “100 WDM 10-Gb/s Channels Over 400 km of TRUEWAVE Fiber 1 Tb/s Transmission", OFC Technical Digest, Postdeadline Paper, PD 10-1-10-4, San Josa, CA. February 22-27, 1998).

In increasing the capacity of these WDM optical communication systems and networks, it has been shown that it is generally desirable to have as many wavelength division multiplexed (WDM) optical channels as possible within a given WDM system. Understandably, broadband optical amplifiers are required to implement these "dense" WDM (DWDM) optical systems and networks.

Proper use of rare earth doped fiber amplifiers to amplify optical signals in communication systems and networks. We have found that these rare earth-doped fiber amplifiers are cost-effective, exhibit low noise, provide relatively large bandwidths that are polarization-independent, exhibit greatly reduced crosstalk, and low insertion loss in relevant operating wavelengths. Due to these superior properties, rare earth-doped fiber amplifiers, such as Erbium-doped fiber amplifiers (EDFA), are gradually replacing optoelectronic regenerators currently used in a large number of lightwave communication systems and networks, especially in wavelength division multiplexing (WDM) optical communication systems and networks.

In order to support the growth of channel number in WDM transmission system and network, broadband mere amplifier is needed. Therefore, the gain of the amplifier should be uniform throughout the WDM bandwidth, so the channel can be transmitted without impairment.

As is known in the art, by using gain-equalizing filters, such as long-period gratings (see, for example, A.M. Vengsarkar, P.J. Jemaire, J.B. Judkins, V. Bhatia, T. Erdogan, and J.E. Filter", J.Lightwave Tech., Vol.14, No.1, pp.58-65, January, 1996), can achieve the required gain characteristics. Unfortunately, when such a broadband optical amplifier is used in a practical system, the power "flatness" of the system may be affected by several factors, such as spectral loss in transmission or dispersion compensating fiber, passive components in the Spectral loss of the input signal, changes in the power spectrum of the input signal, and Raman effects in optical fibers (see, for example, A.R. Chraplyvy and P.S. Henry, "Optical Power Limitation Due to Stimulated Raman Scattering in Multichannel Wavelength Division Multiplexing Systems", Electron. Lett., Vol.20, No.2, pp 58-59, January 1984).

As a first approximation, deviations from "ideal flatness" in broadband optical amplifiers can be approximated as linear slopes in the signal power spectrum. Therefore, there is a need for methods and apparatus for controlling linear tilt to produce broadband amplifiers with desired operating characteristics.

By placing an intermediate stage attenuator within the amplifier, we have invented a method of controlling the gain tilt of an optical amplifier. By changing the loss of the attenuator, the average inversion level of the erbium-doped fiber can be adjusted, and the gain tilt in the EDFA gain spectrum can be further changed.

Further features and advantages of the present invention, as well as the structure and operation of various embodiments of the invention, are described in detail with reference to the accompanying drawings.

Fig. 1 (a) is a schematic diagram of a two-stage optical amplifier, wherein there is an intermediate stage variable optical attenuator according to the present invention;

Fig. 1 (b) is a curve diagram of the relationship between the gain and the wavelength of the optical amplifier in Fig. 1;

Fig. 2 is a schematic diagram of an experimental setup for gain tilt control according to the present invention;

Fig. 3(a) is the input power spectrum curve diagram of 18 WDM channels, which are +4dB and -2 dB slopes respectively;

Fig. 3 (b) shows according to the amplifier of the present invention after amplifying, the output spectrum graph of tilt correction;

Figure 4 shows the attenuator loss curves necessary to obtain a flat output spectrum for different signal slopes from -4 dB to 4 dB; and

Figure 5 shows the attenuator loss curve necessary to operate at constant gain for different signal slopes from -4 dB to 4 dB.

Figure 1(a) illustrates the fundamentals of our amplifier and innovative approach. The amplifier 100 shown in the figure is mainly divided into two stages, including: optical isolator (OI) 101, erbium-doped fiber (EDF) part 103, wavelength selective coupler (WSC) 105, gain equalization filter (GEF) 107, variable Attenuator (VA) 109, and optical pumps 111 and 113 at 980 nm and 1480 nm respectively. The amplifier exhibits broadband, large dynamic range and high power characteristics required for wavelength division multiplexed optical signals.

Referring again to FIG. 1( a ), an optical signal (not shown) enters optical amplifier 100 through input port 110 and exits output port 120 , which is "downstream" of input port 110 . In general, the optical isolator 101, the attenuator 109, the GEF 107, and the WSC 105 are well known to those skilled in the art, some of which are commercially available. In addition, professionals know that the conventional practice is to place optical isolators in the upstream and downstream directions of the EDFA, but this is not mandatory.

Fig. 1(b) is a graph showing the relationship between gain and wavelength of the optical amplifier shown in Fig. 1(a). As shown, the optical amplifier exhibits a uniform gain characteristic over a bandwidth of 35 nm (1526 nm-1561 nm). By adjusting the variable attenuator 109, the gain spectrum can be kept flat over a range of input power levels. At an input power of -4 dBm and the attenuator set to minimum, the gain is 24 dB, with 12 dB of gain compression and a noise figure of about 5 dB.

Fig. 2 is a schematic diagram of an experimental setup for gain tilt control according to the present invention. As shown in the figure, two optical amplifiers are used. Specifically, the first erbium-doped fiber amplifier 210 prepares a simulated power ramped input signal spectrum for the second erbium-doped fiber amplifier 220 . The waveguide grating router 230 is used to multiplex 18 WDM signals (λ ₁ -λ ₁₈ ), which originate from external lasers (not shown). For the purposes of our demonstration, the signal channels range from 1531.4 nm to 1558.6 nm with a channel spacing of 200 GHz, resulting in a total bandwidth of approximately 27 nm.

The signal power of these channels is then routed through an attenuator/power meter 240 which controls the input power into a first (preparative) amplifier 210, which is similar in structure to the amplifier 100 shown in FIG. . Referring again to FIG. 1, the attenuator 109 in the intermediate stage of the amplifier 100 can be tuned to obtain a total power tilt of -4 dB to 4 dB between the shortest and longest wavelength channels. For the purposes of our evaluation, the slope of the power spectrum of the signal input to the second (test) amplifier 220 was monitored by a spectrum analyzer 260 and a second attenuator/power meter 250 was used to adjust the total input power into the test amplifier 220 .

As explained herein, a positive slope is a power slope with low power on the short wavelength side and high power on the long wavelength side. So, negative slope is the opposite case.

The input spectra of 18 WDM channels with +4 dB and -2 dB slopes are drawn in Fig. 3(a). By properly adjusting the intermediate stage variable optical attenuator of the experimental amplifier 220, the power spectrum tilt can be compensated. The output spectrum after the test amplifier 220 is tilt corrected for 4 dB and -2 dB tilt in the input spectrum is plotted in FIG. 3( b ). As shown, the dip in the input spectrum can be fully moderated by changing the mid-stage attenuator loss in both cases.

When the total input power is fixed at 0.4 dBm, Figure 4 shows the attenuator loss required to obtain a flat output spectrum for different input signal slopes ranging from -4 dB to +4 dB. At this input power level, set the attenuator to 4.5dB to produce a flat output spectrum with a flat input spectrum.

Compensation can be accomplished by adjusting the attenuator between 0 dB and 17 dB when the power spectrum slope is between -2 dB and 4 dB. However, this comes at the cost of reduced output power as attenuator losses increase. With a -4 dB slope, the minimum loss in the attenuator is not enough to flatten the output power spectrum.

A similar evaluation is performed for constant gain operation, as shown in Figure 5. In this evaluation, the input power is adjusted to maintain a constant 21.4 dB gain at each power ramp state. To produce a flat output spectrum, set the attenuator to 2 dB with a flat input spectrum. When the attenuator is set to its minimum value, full compensation is achieved at positive slopes and partial compensation is achieved at input slopes of -2 dB. Similar to the constant total input power case, the total output power decreases with increasing attenuation.

Various other modifications can be made to the present invention by those skilled in the art. However, all specific practices that are different from this technical description, and these practices rely on the principles of this technical development and related content, should be considered within the scope of the above-mentioned invention and claims.

Claims (7)

1. A method for controlling gain tilt in an optical amplifier having an input signal and an output signal, comprising the steps of: determining a power ramp characteristic of the input signal; and Adjust the loss of the attenuator so that the gain tilt is properly controlled. 2. 2. The method of claim 1, wherein the attenuator loss is increased if the input signal is determined to be characterized by a positive power slope. 3. 2. The method of claim 1, wherein if the input signal is determined to be characterized by a negative power slope, reducing the attenuator loss. 4. The method of claim 1, wherein the optical amplifier is a multistage erbium-doped fiber amplifier and the attenuator is a variable attenuator placed in an intermediate stage of the multistage amplifier. 5. A method of controlling gain tilt characteristics in an erbium-doped fiber amplifier comprising: The input port is used to receive the optical signal to be amplified; a first amplification stage in optical communication with the input port, the first amplification stage includes an erbium-doped optical amplification fiber; The second amplification stage, the second amplification stage includes an erbium-doped light amplification fiber; an output port in optical communication with the second amplification stage for outputting a signal of the second amplification stage; and a variable attenuator placed between the two stages for optical communication with the first amplification stage and the second amplification stage; The method includes the steps of: Introduce the optical signal to the input port; determining a power ramp characteristic of the input optical signal; and Adjust the loss of the variable attenuator to obtain the desired gain slope. 6. 6. The method of claim 5, wherein the attenuator loss is increased if the input signal is determined to be characterized by a positive power slope. 7. 6. The method of claim 5, wherein the attenuator loss is reduced if the input signal is determined to be characterized by a negative power slope.

Images