Optical communications based on optical polarization multiplexing and demultiplexing

[0001] This application claims the benefit of U.S. Provisional Patent Application No. 60/575,127 entitled “Method and apparatus for polarization demultiplexing” and filed May 28, 2004 . The entire disclosure of the above referenced application is incorporated herein by reference as part of the specification of this application.

[0002] This application relates to optical polarization and use of optical polarization in optical communications.

[0003] Optical polarization is an important parameter of an optical signal can be used in various optical devices and systems. In some applications, optical polarization may cause technical difficulties. For example, in fiber optic communication systems, polarization-dependent effects in fibers and other devices, such as the polarization-dependent loss (PDL), polarization-mode dispersion (PMD) and differential group delay (DGD), can adversely affect the performance and proper operations of optical devices or systems.

[0004] Optical polarization may also be used to benefit various applications. For example, two orthogonal polarized beams can be easily separated as in optical disk drives and many other optical devices and systems. In optical communication systems, two optical beams with mutually perpendicular linear polarizations may be used to carry two different optical channels and hence can be used to double the capacity of the information carried at a particular optical wavelength. This mechanism is essentially a polarization multiplexing technique and can be used in combination with the wavelength-division multiplexing (WDM) to double the capacity of a given WDM system. Implementations of such polarization multiplexing can be used to improve the capacity of existing WDM systems at a low cost in comparison with building new WDM systems to meet the demand for increased communication capacity.

[0005] The state of polarization (SOP) of light, however, often fluctuates when transmitting through an optical transmission medium due to, e.g., inherent optical birefringence in the medium or induced optical birefringence caused by the surrounding conditions of the media (such as stresses). Many commercial optical fibers are known to be optically birefringent and exhibit different refractive indices for light with different polarizations. Often, fiber birefringence is characterized along two orthogonal principal directions. Causes for such birefringence in fibers include, among others, imperfect circular core and unbalanced stresses in a fiber along different transverse directions. Notably, the axis of birefringence of the optical fiber may change randomly over time. This random nature in the SOP of light can render implementations of polarization multiplexing difficult, if not possible.

[0006] This application includes implementations and examples of devices and techniques for demultiplexing polarization multiplexed optical signals in a simple hardware environment, based on simple signal processing with a feedback control, and based on generally optical processing and control of the optical signal at the receiver to provide reliable demultiplexing at a high speed and a relative low cost.

[0007] In one implementation, a device for demultiplexing polarization multiplexed channels in an optical signal can include a polarization beam splitter to receive an optical signal and split the received optical signal into a first beam in a first polarization and a second beam in a polarization orthogonal to the first polarization. The received optical signal carries first and second signal channels via polarization multiplexing and is originally created by combining two optical beams with mutually orthogonal polarizations and unequal amplitudes which are modulated to carry the first and second signal channels, respectively. A first optical detector is provided to convert the first beam into a first electrical signal and a second optical detector is provided to convert the second beam into a second electrical signal. This device uses a control circuit to obtain a differential signal representative of a difference between the first and second beams and being a function of a relative angle between a polarization direction of the polarization beam splitter and a polarization in the received optical signal. This device further includes a polarization controller placed in an input optical path of the optical polarization beam splitter. The polarization controller is used to adjust a polarization of the received optical signal in response to the differential signal to set the difference to be either at a positive maximum or a negative maximum to make the first and second electrical signals to represent the two signal channels.

[0008] In another implementation, a demultiplexing method is described as follows. A polarization multiplexed optical signal is split into a first beam and a second beam with mutually orthogonal polarizations. The polarization multiplexed optical signal carries first and second signal channels via polarization multiplexing and is originally created by combining two optical beams with mutually orthogonal polarizations and unequal amplitudes which are modulated to carry the first and second signal channels, respectively. The first beam and the second beam are converted into first and second electrical signals, respectively. A difference between the first and second electrical signals is obtained as a function of an angle of a state of polarization of one of the two orthogonal polarizations in the received optical signal. The state of polarization of the received optical signal is adjusted prior to the splitting to make the difference between the first and second electrical signals at either a positive maximum or a negative minimum. After the difference is set, one of the first and second electrical signals is used as one signal channel and another is used as another signal channel, thus demultiplexing the two signals channels in the received optical signal.

[0009] These and other implementations and applications are described in greater detail in the attached drawings, the detailed description, and the claims.

[0013] The demultiplexing techniques described in this application can be used in a polarization multiplexing optical communication system. FIG. 1 shows one example 100 which includes a polarization multiplexed optical transmitter 110, an optical transmission medium, network, or system 115 which may be a fiber link, and a polarization demultiplexing optical receiver 120. The transmitter 110 is used to produce a multiplexed optical signal 114 which carries first and second signal channels (channels 1 and 2) via polarization multiplexing. The optical transmitter 110 as illustrated includes a first device such as a first optical transmitter TX1 (111) to generate a first initial optical beam in a first polarization to carry the first signal channel (channel 1), a second device such as a second optical transmitter TX2 (112) to generate a second initial optical beam in a first polarization orthogonal to the first polarization to carry the second signal channel, and a third device 113 such as a polarization beam splitting cube (PBC) to combine the first and second initial optical beams to form the multiplexed optical signal 114. One example for two mutually orthogonal polarizations is two mutually perpendicular linear polarizations, which are illustrated in FIG. 1. Two mutually orthogonal polarizations may be in polarizations other than linear polarizations, e.g., circular polarizations or elliptical polarizations.

[0014] In FIG. 1, the amplitudes of the two optical beams in orthogonal polarizations that are produced by the two different transmitters 111 and 112 and combined at the PBC 113 are different and should be kept different which is generally true in practical devices unless they are equalized on purpose. In addition, the optical phases of the two beams should be independent from each other without a fixed phase relationship and can often vary with time. If the two beams are extracted from the same beam or from a common laser, then the phase of one beam should be scrambled relative to the other beam. Alternatively, one of the beams should be delayed with respect to the other one with the relative delay larger than the coherence length of the laser. Polarization maintaining (PM) fibers may be used to carry the beams in the transmitter 110. Other optical transmitters may also be used using other methods to produce the polarization multiplexed signal 114.

[0015] After transmitting through the optical link 115, the polarization multiplexed signal 114 becomes a polarization multiplexed signal 116 which still carries the two different signal channels (i.e., channels 1 and 2) but the polarization states of the two channels are unknown due to propagation in the link 115. To a certain extent, the polarization states in the signal 116 may appear random and hence the two signal channels may seemingly be scrambled due to the randomness or fluctuations of the polarization states. The optical receiver 120 is designed to recover or demultiplex the two polarization multiplexed signal channels from the signal 115 in an efficient and simple way.

[0016] In the illustrated example, optical receiver 120 includes a polarization controller 121 to control the polarization of the signal 116 at entry of a polarization beam splitter 122. Such a polarization controller may be implemented in various configurations and may include multiple adjustable polarization rotators. The polarization beam splitter 122 receives the signal 115 from the controller 121 and splits the received optical signal into a first beam in a first received polarization and a second beam in a polarization orthogonal to the first received polarization. Two optional optical polarizers 122-1 and 122-2 may be used to respectively filter the two beams to ensure their mutual orthogonal nature of the polarizations. Next, the receiver 120 uses a first optical detector RX1 (120-1) to convert the first beam into a first electrical signal (V1) and a second optical detector RX2 (120-2) to convert the second beam into a second electrical signal (V2). The signals V1 and V2 are then used to control the polarization controller 121 to set the difference between V1 and V2 at a positive maximum or a negative minimum. After the difference is set, the signals V1 and V2 represent the two signal channels, respectively, and the polarization demultiplexing is completed.

[0017] A control circuit 125 is used in the receiver 120 to obtain a differential signal representative of the difference between the first and second beams (of the signals V1 and V2) and the receiver 200 is designed to make the differential signal as a function of a relative angle between a polarization direction of the polarization beam splitter 122 and a polarization in the received optical signal from the polarization controller 121. The polarization controller 121, placed in an input optical path of the optical polarization beam splitter 122 and to adjust a polarization of the received optical signal in response to the differential signal to set the difference to be either at a positive minimum or a negative maximum to make the first and second electrical signals V1 and V2 to represent to two signal channels.

[0018] The circuit 125 can produce the differential signal in two different ways. The first way is illustrated in FIG. 1 where two optical tap splitters 123-1 and 123-2 are used in the paths of the two beams, respectively, to split a fraction of the two beams to generate first and second monitor beams. A first monitor detector 124-1 is used to convert the first monitor beam into a first electrical monitor signal and a second monitor detector 124-2 is used to convert the second monitor beam into a second electrical monitor signal. The two monitor signals are basically representative of the signals V1 and V2 and are fed into the control circuit 125 to produce the differential signal. Electrical amplifiers may be used to amplify the signals received by the circuit 125.

[0019] Alternatively, the signals V1 and V2 respectively produced by the detectors 120-1 and 120-2 may be split to produce the two monitor signals for processing by the circuit 125. In this design, the optical tap splitters 123-1 and 123-2 and the monitor optical detectors 124-1 and 124-2 are no longer needed. Two electrical signal splitters may be used to produce the monitor signals. Referring to FIG. 3, the control circuit 125 processes the monitor signals from the two electrical signal splitters and adjusts the polarization controller 121 to set the difference between the signals V1 and V2 to be at either the positive maximum or the negative maximum. In one implementation, two bias T splitters may be used to split the monitor signals from the signals V1 and V2. Because only the average optical power level of the differential signal (V1-V2) is used for controlling the polarization controller 121 for the polarization demultiplexing, the bias T may be used to split out the DC or low frequency part of the signal V1 or V2 as the monitor signal while the remaining signal from the bias T has high frequency components which carry the signal channel and are used for data extraction or other operation on the signal channel.

[0020] Notably, the above demultiplexing does not need to monitor the polarization state of the received optical signal at the receiver 120 and provides a simple method to demultiplex two polarization channels by complete optical means, which results in high speed operation and low cost. The average optical power levels from TX1 and TX2 are different and the average power levels in the output two arms of the PBS 122 at the receiving end are monitored to achieve nearly perfect demultiplexing without ambiguity.

[0021] FIG. 2 illustrates one example of the operational flow of the receiver 120 in FIG. 1. The details of the operation are described in the following sections.

[0022] The electrical fields emitted by the TX1 and TX2 can be expressed as:
{right arrow over (E)}₁=A₁e^iφ1(t){circumflex over (x)}
{right arrow over (E)}₂=A₂e^iφ2(t)ŷ
where A₁ and A₂ are the timing varying amplitudes and are different from each other; φ₁ and φ₂ are the time varying phases of the two signals, respectively and are independent of each other. This can be done, e.g., by using two different lasers as the TX1 111 and TX2 112 so there is no phase relationship between them. The output polarization multiplexed signal out of the transmitter 110, which propagates through the link 115 (e.g., a single mode fiber), can be expressed as the sum of the two orthogonally polarized components carrying two different signal channels:
{right arrow over (E)}=A₁e^iφ1(1){circumflex over (x)}+A₂e^iφ2(t)ŷ

[0023] The transmission functions of the polarization beam splitter (PBS) 122 at the receiver 120 for the two channels can be expressed as:
{right arrow over (T)}₁=cosθ{circumflex over (x)}+sinθŷ
{right arrow over (T)}₂=sinθ{circumflex over (x)}−cosθŷ
wherein the angle θ is the relative angle between the polarization of light and the polarization direction of the PBS 122 in the receiver 120. The optical power at two output arms of the PBS 122 can be calculated as
P_i=|{right arrow over (E)}·{right arrow over (T)}_i|²
P₂=|(A₁e^iφ1{circumflex over (x)}+A₂e^iφ2ŷ)·(sin θ{circumflex over (x)}−cos θŷ)|²=A₁² sin² θ+A₂² cos² θ−A₁A₂ sin 2θ cos[φ₁(t)−φ₂(t)]
P₂=|(A₁e^iφ1{circumflex over (x)}+A₂e^iφ2ŷ)·(sin θ{circumflex over (x)}−cos θŷ)|²=A₁² sin² θ+A₂² cos² θ+A₁A₂sin 2θ cos[φ₁(t)−φ₂(t)]
where subscript i is 1 and 2 to denote the two output arms of the PBS 122, respectively. Because φ₁ and φ₂ have no phase relationship, the term cos(φ₁−φ₂) varies rapidly and averages to zero in a slow photodetector. Consequently, the powers in the two arms are:
P₁=A₁² cos²θ+A₂² sin²θ
P₂=A₁² sin²θ+A₂² cos²θ

[0024] The photovoltages generated in the corresponding tap detectors 124-1 and 124-2 are:
V₁=α₁(A₁² cos²θ+A₂² sin²θ)
V₂=α₂ (A₁² sin²θ+A₂² cos²θ)
where α₁ and α₂ are proportional constants taking into account of splitting ratios of the power taps, the responsivity of the photodetectors, and electronic amplifier gains of the corresponding channels. The difference of the two voltages yields:
V₁−V₂=(α₁A₁²−α₂A₂²)cos²θ−(α₂A₁²−α₁A₂²)sin²θ
The gain of the amplifiers for amplifying the signals V1 and V2 may be adjusted such that α₁ and α₂ are equal. In this case, the voltage difference or the differential signal (V1-V2) becomes:
ΔV=V₁−V₂=α·(A₁²−A₂²)(cos²θ−sin²θ)=2α(A₁²−A₂²)cos2θ
which is a function of the angle θ.

[0025] Therefore, as long as the powers from TX1 and TX2 are not equal, A1≠A2, the voltage difference between the two power monitors strongly depends on the relative orientation angle θ between the PBS 122 and the SOP of the two signals. By feeding the differential signal ΔV back to the polarization controller 121 to adjust the relative orientation angle to achieve either the positive maximum (θ=0) or negative maximum (θ=180°), the two channels can be effectively separated. It is interesting to note that, in the example shown in FIG. 1, when ΔV is maximum, signal from TX1 goes to RX1 while signal from TX2 goes to RX2. When ΔV is negative maximum, signal from TX1 goes to RX2 while signal from TX2 goes to RX1.

[0026] When α₁ and α₂ are equal, then
V₁+V₂=α(A₁²+A₂²)=cons tan t
Therefore, by measuring V₁+V₂, one can ensure that the gains of the two channels are properly set.

[0027] The above polarization multiplexing may be used in combination with wavelength-division multiplexing (WDM) to further expand the capacity of a WDM system. Each WDM channel at a WDM wavelength can be used to carry two different channels based on the present polarization multiplexing. Hence, two or more multiplexed optical signals at different WDM wavelengths each carrying two polarization multiplexed signal channels can be generated and sent through a common fiber link to a destination. At the destination, the multiplexed optical signals at different wavelengths can be separate into separate multiplexed optical signals. Each of separate multiplexed optical signals is then processed based on the above demultiplexing process to demultiplex two signals channels therein.

[0028] The above techniques for implementing polarization multiplexing and demultiplexing are simple in their hardware designs and in operation. A simple feedback is used to adjust the polarization of the received light to set the differential signal at a positive maximum or negative maximum. The optical receiver 120 does not need to perform complex signal processing tasks. Notably, the present techniques do not require any special features in optical transmitters or receivers. As such, a wide range of optical transmitters and receivers, including commercial transmitters and receivers for WDM systems, may be used, without any change or modification, to implement the present polarization multiplexing and demultiplexing. As an example, existing optical transmitters and receivers in the existing optical communication systems can be used without change so that a carrier can deploy polarization multiplexing and demultiplexing described in this application to double its communication capacity with a significantly reduced effort, cost, and time in comparison with deployment of other polarization multiplexing and demultiplexing systems which may require specialized and proprietary optical transmitters and receivers and complex demultiplexing systems.

[0029] The present polarization demultiplexing scheme is essentially performed in the optical domain by controlling the polarization of the received polarization multiplexed signal. Hence, the implementation of this demultiplexing scheme does not require any changes in electronic hardware of an existing system.

[0030] Only a few examples and implementations are described. However, other implementations, variations, modifications, and enhancements are possible.