Optical: systems and elements – Optical amplifier – Particular active medium
Reexamination Certificate
2001-12-04
2004-12-28
Hellner, Mark (Department: 3663)
Optical: systems and elements
Optical amplifier
Particular active medium
Reexamination Certificate
active
06836357
ABSTRACT:
FIELD
This patent specification relates to optical amplifiers. More specifically, it relates to a semiconductor optical amplifier that amplifies an optical signal using energy from one or more nearby and/or intersecting laser cavities.
BACKGROUND
As the world's need for communication capacity continues to increase, the use of optical signals to transfer large amounts of information has become increasingly favored over other schemes such as those using twisted copper wires, coaxial cables, or microwave links. Optical communication systems use optical signals to carry information at high speeds over an optical path such as an optical fiber. Optical fiber communication systems are generally immune to electromagnetic interference effects, unlike the other schemes listed above. Furthermore, the silica glass fibers used in fiber optic communication systems are lightweight, comparatively low cost, and are capable of very high-bandwidth operation.
Optical amplifiers are important components of optical communications links. Optical amplifiers are commonly used as (i) power amplifiers at the source end of an optical communications link, (ii) line amplifiers along the optical signal transmission path, and (iii) preamplifiers at the receiving end of the optical communications link, and have other uses as well.
In general, the two primary types of optical amplifiers are optical fiber based amplifiers, such as erbium doped fiber amplifiers (EDFAs) and Raman amplifiers, and semiconductor optical amplifiers (SOAs). EDFAs are widely used in line amplifiers and other applications requiring high output power, high data rates, and low noise. However, EDFAs are quite bulky, having a typical fiber length of about 30 feet, and require the presence of a separate pumping laser to operate. Accordingly, EDFAs are difficult to incorporate into confined spaces, and are certainly not amenable to circuit-board-level or chip-level integration.
SOAs, on the other hand, are small in size and conveniently integrated into small devices. An SOA generally resembles a semiconductor laser structure, except that the end mirrors are replaced by antireflection coatings. In such devices the product of the gain and the reflectivity is less than one so that the device does not oscillate. Rather, the device is used to amplify an incoming optical signal as it passes through the device. Such devices are often called traveling wave amplifiers, which highlights the fact that the optical signal does not pass back and forth within the device, but merely passes through it essentially only once. SOAs generally yield lower output power and higher noise levels as compared to EDFAs, and/or are restricted to lower data rates at high output power due to pattern dependent amplitude variations. Research continues toward improving the performance of SOAs, including making SOAs with higher power and lower noise characteristics, and/or that are capable of operating at higher data rates at high input power without suffering significant pattern dependent amplitude fluctuations.
Crosstalk is one of the primary troublesome noise sources in conventional SOAs. Crosstalk, or cross-channel modulation, involves data-dependent gain fluctuations at high output levels from the SOA, and can occur for either time-multiplexed or wavelength-multiplexed data. Crosstalk arises from gain saturation effects in an SOA. These effects can be understood by recalling that SOA devices rely on the phenomenon of stimulated emission to provide the necessary amplification. Stimulated emission, in turn, requires the establishment of a population inversion. In typical SOAs or lasers a population inversion is evidenced by the presence of a specified carrier density. When a sufficiently large optical signal is passed through the amplifier, the population inversion is substantially reduced or depleted, i.e. the gain of the SOA is saturated, and is reestablished only over some finite period of time. Consequently, the gain of the SOA will be reduced for some period of time following the passage of the signal through the amplifier, a time period commonly denoted as the amplifier gain recovery time.
When the gain medium becomes saturated due to a high signal level on a first channel, changes are induced in the signal level of a second channel because the saturated gain medium cannot properly amplify both channels. Since gain is modulated by the first signal, this modulated gain is impressed on the second signal. Thus, for wavelength division multiplexed (WDM) systems in which a plurality of channels at &lgr;
1
,&lgr;
2
, . . . , &lgr;
N
are present in a common optical signal, gain saturation induced by a first channel at &lgr;
1
can produce unwanted level changes (i.e., errors) in a second data channel at &lgr;
2
, and vice versa.
Crosstalk can be reduced by keeping the SOA out of gain saturation for the data rates, signal levels, and number of channels on the optical signal of interest. If the SOA is operated near gain saturation levels, crosstalk may be reduced by making the “off” portion of the signal duty cycle longer in comparison to the gain recovery time, i.e., by slowing down the data rate. In general, an SOA will have reduced crosstalk effects if (i) its saturation power P
SAT
, i.e., the input optical power level for which the SOA gain is reduced to a predetermined percentage of its nominal value, is increased, and/or (ii) its gain recovery time is decreased. As used herein, an SOA has increased gain stability if (i) its saturation power P
SAT
is increased without a concomitant increase in gain recovery time, (ii) its gain recovery time is decreased without a concomitant decrease in saturation power P
SAT
, or (iii) both (i) and (ii) occur.
Several methods for dealing with crosstalk problems are discussed in U.S. Pat. No. 5,436,759, which is incorporated by reference herein. One particularly appealing strategy is to place a transverse laser across the SOA such that the laser's gain medium and the SOA's signal gain medium share an overlapping region. The lasing cavity is operated above threshold and the gain of the laser is clamped to overcome the losses of the cavity. As used herein, a laser cavity is gain-clamped and lasing when it is excited by a bias current greater than a threshold current. When the transverse laser is gain-clamped, gain along the SOA signal path is stabilized. The transverse lasing enhances the establishment and maintenance of a population inversion in the overlapping region, resulting in both increased saturation power and a decreased gain recovery time due to increased photon density in the laser cavity. Advantageously, independent lasing only builds up in the transverse direction and does not corrupt the quality of the amplified signal. Other methods in which the gain media of SOAs is coextensive with the gain media of transverse lasers are discussed in Francis, et. al., “A Single Chip Linear Optical Amplifier,” IEEE Optical Fiber Communication Conference, Anaheim, Calif. (2001), which is incorporated by reference herein, and in U.S. Ser. No. 09/972,146, supra.
Amplified spontaneous emission (ASE) is another primary troublesome noise source in conventional SOAs. ASE arises from random, spontaneous energy state drops in a small fraction of the excited carriers of the gain medium. Light emitted as a result of these energy state drops is generally random in direction and wavelength. Some of this light will be emitted in the direction of signal propagation and will therefore be amplified as it propagates, resulting in output background noise signal similar to white noise. For good performance, it is generally desirable to keep the amplified signal levels at least 10 dB higher than ASE levels.
One potential disadvantage of an SOA having a gain medium coextensive with the gain medium of a transverse laser cavity is that ASE levels in the output can be relatively high. This is because the population inversion of the gain medium, and therefore the number of carriers experiencing spontaneous energy state drops, is maintained at high l
Islam Saif M.
Wang Shih-Yuan
Wei Haiqing
Zhu Miao
Zhu Zuhua
Cooper & Dunham LLP
Gazillion Bits Inc.
Hellner Mark
LandOfFree
Semiconductor optical amplifier using laser cavity energy to... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Semiconductor optical amplifier using laser cavity energy to..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Semiconductor optical amplifier using laser cavity energy to... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3293791