Polarization demultiplexing optical receiver using polarization oversampling and electronic polarization tracking

#An optical network is a system for communicating information over optical fiber using optical transmitters (e.g., lasers or light-emitting diodes (LEDs)) and optical receivers. Some optical networks use a digital modulation scheme that conveys data by changing, or modulating, a phase of a reference signal (e.g., a carrier wave or carrier signal). The digital modulation scheme may use a finite number of distinct signals to represent digital data. For example, a phase-shift keying (PSK) modulation scheme may capture modulation formats in which a phase of a carrier wave is modulated. One technique for transmitting a pair of data bits with a four-level code is quadrature phase-shift keying (QPSK) modulation, where each pair of bits is encoded during each symbol period as one of four possible phases of a transmitted carrier signal. Differential phase-shift keying (DPSK) is touted as a promising modulation format for optical communication systems requiring high spectral efficiency. The four-level version of DPSK (i.e., differential quadrature phase-shift keying (DQPSK)) transmits two bits for every symbol. In DQPSK modulation, each pair of data bits is encoded as one of four possible phase changes of a transmitted carrier signal. DQPSK has a narrower optical spectrum than conventional binary DPSK, which tolerates more dispersion (both chromatic and polarization-mode), allows for stronger optical filtering, and enables closer channel spacing. As a result, DQPSK may be the simplest modulation format which allows processing of forty (40) gigabits per second (G/s or Gbps) data-rate in a fifty (50) gigahertz (GHz) channel spacing system.

#Some optical networks may employ a DQPSK digital modulation scheme by incorporating a DQPSK modulator in an optical transmitter and a DQPSK demodulator in an optical receiver. In order to recover independent data signals, without degradation from crosstalk, the carrier signal phase needs to be determined and demultiplexed with the DQPSK based optical receiver. This places stringent requirements on the alignment between the DQPSK based transmitter and the DQPSK demodulation filter (e.g., typically on the order of one-hundred megahertz (MHz) when twenty (20) G/s traffic is present). When this alignment is not met, in-phase and quadrature signal paths are not sufficiently isolated and each path suffers eye closure (e.g., distortion). Typical DQPSK demodulator filters are polarization sensitive, which compounds the difficulty of aligning the filters with carrier waves. Conventional DQPSK demodulators must be carefully designed in order to prevent such polarization sensitivity.

#A polarization multiplexed intensity and phase modulated signal can be generally reconstructed by a class of cross polarization interference canceller (XPIC) circuitry either in the optical domain or the electrical domain. For example, polarization multiplexed QPSK information may be electronically recovered at an optical receiver, without polarization control hardware, using a combination of coherent detection (e.g., mixing a received signal with a local optical light source) and an electrical XPIC either in analog or digital format. However, coherent detection has many consequences, such as stringent phase requirements on transmit and receive optical light source stability and additional optical components. Furthermore, the electronics used in digital signal processing to recover a polarization-multiplexed QPSK signal consumes considerable power and are highly complex. Coherent detectors can also rely on analog carrier recovery circuits, but locking a local carrier can be a difficult task. Coherent detectors always carry an additional burden in terms of the optical generation and mixing of a local light source to the received light source.

#The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate one or more implementations described herein and, together with the description, explain these implementations. In the drawings:

#FIG. 1 is a diagram of an exemplary network in which systems and methods described herein may be implemented;

#FIG. 2 illustrates exemplary components of an optical transmitter of the network depicted in FIG. 1;

#The following detailed description refers to the accompanying drawings. The same reference numbers in different drawings may identify the same or similar elements. Also, the following detailed description does not limit the invention.

#As high-bandwidth applications for networks (e.g., optical networks) have grown, service providers responsible for providing such applications have sought higher data capacity within their optical networks. This has led to increased bitrates on individual optical channels, and an increased number of wavelength-division multiplexing based channels provided on optical fibers. However, this growth is becoming more difficult to handle because data bandwidth of individual optical channels cannot continue to increase without data bands overlapping. Thus, service providers are faced with finding ways to increase a spectral efficiency (e.g., a number of bits per second packed within a given optical bandwidth) of their optical networks.

#One solution is polarization multiplexing, which takes advantage of the orthogonality of two polarization states provided in an optical fiber. By transmitting into orthogonal states, the data capacity of the optical fiber can be doubled. However, a main difficulty with polarization multiplexing is successfully recovering the orthogonal data streams at a receive end of an optical fiber. While splitting the polarization may be straightforward, the polarization must first be correctly aligned into an optical splitter. Since the optical fiber may rapidly rotate the polarization, an optical polarization controller is typically provided before the polarization splitting to guarantee that polarizations are aligned into an optical receiver. However, an optical polarization controller is complex, expensive, and requires complex algorithms.