Training sequences with enhanced IQ imbalance tolerances for training-aided frequency domain equalization

• US 10,129,054 B2
• Filed: 02/10/2017
• Issued: 11/13/2018
• Est. Priority Date: 02/10/2017
• Status: Active Grant

• Current Assignee: Futurewei Technologies Incorporated (Huawei Investment & Holding Co., Ltd.)
• Original Assignee: Futurewei Technologies Incorporated (Huawei Investment & Holding Co., Ltd.)
• Inventors: Guo, Qing, Peng, Wei-Ren, Cui, Yan, Bai, Yu Sheng
• Primary Examiner(s): Tadese, Berhanu
• Application Number:
  US15/429,447
• Publication Number:
  US 20180234273A1
• Time in Patent Office: 641 Days
• CPC Class Codes: H04B 10/532 Polarisation modulation
• CPC Class Codes: H04B 10/6151 comprising a polarization controller at the receiver's input stage
• CPC Class Codes: H04L 2025/03522 Frequency domain
• CPC Class Codes: H04L 25/03159 operating in the frequency domain H04L25/03165, H04L25/03178 take precedence

Abstract

Systems and methods for creating and using a first and second group of training sequences in a training-aided single-carrier frequency domain equalization system. The first group of training sequences are conventional training sequences and the second group of training sequences are 90°-rotated versions of the first group of training sequences.

Claims

1. A method for a digital signal processor (DSP) to create a data stream for transmission over fiber optic media, the method comprising:

receiving a plurality of data segments;

writing a first frame to a first buffer, wherein the first frame comprises a first training sequence and a first data segment of the plurality of data segments,

writing a second frame to the first buffer, wherein the second frame comprises a second training sequence and a second data segment of the plurality of data segments;

writing a third frame to the first buffer, wherein the third frame comprises a third training sequence and a third data segment of the plurality of data segments; and

writing a fourth frame to the first buffer, wherein the fourth frame comprises a fourth training sequence and a fourth data segment of the plurality of data segments,

wherein the first training sequence and the second training sequences are mathematically-related to the third training sequence and the fourth training sequence,

wherein the third training sequence and the fourth training sequence are derived from the first training sequence and the second training sequence according to the relationships;

2. The method of claim 1, wherein the third training sequence is derived from the first training sequence by rotating the first training sequence by ninety degrees, and wherein the fourth training sequence is derived from the second training sequence by rotating the second training sequence by ninety degrees.

3. The method of claim 1 further comprising:

writing a first frame to a second buffer, wherein the first frame comprises a fifth training sequence and a fifth data segment of the plurality of data segments;

writing a second frame to the second buffer, wherein the second frame comprises a sixth training sequence and a sixth data segment of the plurality of data segments;

writing a third frame to the second buffer, wherein the third frame comprises a seventh training sequence and a seventh data segment of the plurality of data segments; and

writing a fourth frame to the second buffer, wherein the fourth frame comprises an eighth training sequence and an eighth data segment of the plurality of data segments,

wherein the seventh training sequence and the eighth training sequence are mathematically-related to the fifth training sequence and the sixth training sequence.

4. The method of claim 3, wherein the seventh training sequence is derived from the fifth training sequence by rotating the fifth training sequence by ninety degrees, and wherein the eighth training sequence is derived from the sixth training sequence by rotating the sixth training sequence by ninety degrees.

5. The method of claim 3, wherein the seventh training sequence and the eighth training sequence are derived from the fifth training sequence and the sixth training sequence according to the relationships:

6. The method of claim 3, wherein the data stream comprises an X-polarization data stream portion and a Y-polarization data stream portion, wherein the first buffer corresponds to an X-polarization data stream portion, and wherein the second buffer corresponds to a Y-polarization data stream portion.

7. The method of claim 1 further comprising:

precoding the first buffer into an in-phase high-speed data signal and a quadrature high-speed data signal; and

sending the in-phase high-speed data signal and the quadrature high-speed data signal to a modulator.

8. The method of claim 6 further comprising:

precoding the X-polarization data stream portion into a first in-phase high-speed data signal and a first quadrature high-speed data signal;

precoding the Y-polarization data stream portion into a second first in-phase high-speed data signal and a second quadrature high-speed data signal; and

sending the first in-phase high-speed data signal, the first quadrature high-speed data signal, the second in-phase high-speed data signal, and the second quadrature high-speed data signal to a dual-polarization modulator.

9. A method for a receiver to interpret a dual-polarization optical signal, the method comprising:

receiving the optical signal, wherein the optical signal comprises four frames of an X-polarization data stream and four frames of a Y-polarization data stream, wherein each frame comprises a unique training sequence, wherein the training sequence of the first frame of the X-polarization data stream is mathematically-related to the training sequence of the third frame of the X-polarization data stream, wherein the training sequence of the second frame of the X-polarization data stream is mathematically-related to the training sequence of the fourth frame of the X-polarization data stream, wherein the training sequence of the first frame of the Y-polarization data stream is mathematically-related to the training sequence of the third frame of the Y-polarization data stream, and wherein the training sequence of the second frame of the Y-polarization data stream is mathematically-related to the training sequence of the fourth frame of the Y-polarization data stream;

estimating the first and second frames of the X-polarization data stream and the first and second frames of the Y-polarization data stream according to a first transfer function;

calculating a second transfer function according to the first transfer function, the training sequence of the first and second frames of the X-polarization data stream, and the training sequences of the first and second frames of the Y-polarization data stream;

estimating the third and fourth frames of the X-polarization data stream and the third and fourth frames of the Y-polarization data stream according to the second transfer function; and

calculating a third transfer function according to the second transfer function, the training sequences of the third and fourth frames of the X-polarization data stream, and the training sequences of the third and fourth frames of the Y-polarization data stream,

wherein the mathematical relationship between the training sequences is;

10. The method of claim 9, wherein the training sequence of the third frame of the X-polarization data stream is rotated by ninety degrees with respect to the training sequence of the first frame of the X-polarization data stream, wherein the training sequence of the fourth frame of the X-polarization data stream is rotated by ninety degrees with respect to the training sequence of the second frame of the X-polarization data stream, wherein the training sequence of the third frame of the Y-polarization data stream is rotated by ninety degrees with respect to the training sequence of the first frame of the Y-polarization data stream, and wherein the training sequence of the fourth frame of the Y-polarization data stream is rotated by ninety degrees with respect to the training sequence of the second frame of the Y-polarization data stream.

11. The method of claim 9, wherein the second transfer function is calculated according to the relationship:

12. The method of claim 9, wherein the third transfer function is calculated according to the relationship:

13. The method of claim 9, wherein estimating the first and second frames of the X-polarization data stream and the first and second frames of the Y-polarization data stream comprises:

transforming the optical signal from an analog representation to a first time-domain digital representation;

transforming the first time-domain digital representation to a first frequency-domain digital representation;

creating a second frequency-domain digital representation from the first frequency-domain digital representation according to the first transfer function;

transforming the second frequency-domain digital representation to a second time-domain digital representation; and

decoding the second time-domain digital representation according to a modulation scheme.

14. The method of claim 13, wherein creating a second frequency-domain digital representation comprises applying the first transfer function to the first frequency-domain digital representation according to the relationship:

15. A transceiver for sending and receiving data streams over fiber optic media, the transceiver comprising:

a transmitter comprising a first digital signal processor (DSP) and a dual-polarity optical modulator; and

a receiver comprising a second DSP and an integrated coherent receiver (ICR);

wherein the first DSP is configured to;

insert a plurality of training sequences into a first X-polarization data stream, wherein a first training sequence of the first X-polarization data stream is mathematically-related to a second training sequence of the first X-polarization data stream;

insert a plurality of training sequences into a first Y-polarization data stream, wherein a first training sequence of the first Y-polarization data stream is mathematically-related to a second training sequence of the first Y-polarization data stream;

precode the first X-polarization data stream and the first Y-polarization data stream into a first plurality of high-speed data signals for modulation by the dual-polarity optical modulator, and

wherein the second DSP is configured to;

receive an analog electrical signal from the ICR, wherein the signal comprise four frames of a second X-polarization data stream and four frames of a second Y-polarization data stream, wherein each frame comprises a unique training sequence, wherein a first training sequence of the second X-polarization data stream is mathematically-related to a second training sequence of the second X-polarization data stream, and wherein a first training sequence of the second Y-polarization data stream is mathematically-related to a second training sequence of the second Y-polarization data stream;

estimate the first and second frames of the X-polarization data stream and the first and second frames of the Y-polarization data stream according to a first transfer function; and

calculate a second transfer function according to the first transfer function, the training sequences of the first and second frames of the X-polarization data stream, and the training sequences of the first and second frames of the Y-polarization data stream,

wherein the first training sequence of the first X-polarization data stream is mathematically-related to the second training sequence of the first X-polarization data stream according to the relationship;

16. The transceiver of claim 15, wherein a third training sequence of the first X-polarization data stream is mathematically-related to a fourth training sequence of the first X-polarization data stream, wherein a third training sequence of the first Y-polarization data stream is mathematically-related to a fourth training sequence of the first Y-polarization data stream, wherein a third training sequence of the second X-polarization data stream is mathematically-related to a fourth training sequence of the second X-polarization data stream, and wherein a third training sequence of the second Y-polarization data stream is mathematically-related to a fourth training sequence of the second Y-polarization data stream.

17. The transceiver of claim 15, wherein the first DSP and the second DSP are physically a single DSP.