CN105553891A

CN105553891A - Channel recovery method of coherent optical orthogonal frequency division multiplexing system of hidden training sequence

Info

Publication number: CN105553891A
Application number: CN201510933232.7A
Authority: CN
Inventors: 张旭; 李响; 王元祥; 杨奇; 余少华
Original assignee: Wuhan Research Institute of Posts and Telecommunications Co Ltd
Current assignee: Wuhan Research Institute of Posts and Telecommunications Co Ltd
Priority date: 2015-12-15
Filing date: 2015-12-15
Publication date: 2016-05-04
Anticipated expiration: 2035-12-15
Also published as: CN105553891B

Abstract

The present invention relates to the field of optical fiber communication, and particularly relates to a channel recovery method of a coherent optical orthogonal frequency division multiplexing system of a hidden training sequence. Through adding a training sequence on a transmission signal, the training sequence is extracted at a receiving end and the channel is estimated. Then the interference of the training sequence is eliminated at the receiving end, the estimated channel is used to balance the signal, and thus the recovery of the signal is completed. According to the method, the spectral efficiency of the coherent optical orthogonal frequency division multiplexing system can be improved.

Description

The channel recovery method of the coherent light ofdm system of recessive training sequence

Technical field

The present invention relates to fiber optic communication field, specifically relate to a kind of channel recovery method of coherent light ofdm system of recessive training sequence.

Background technology

The channel estimating of current coherent light ofdm system is all based on the method for training sequence.But long training sequence can bring the decline of spectrum efficiency.Therefore, in coherent light ofdm system, need new technology to realize the decline that channel estimating does not bring spectrum efficiency simultaneously.

Summary of the invention

For the deficiencies in the prior art, the present invention proposes in coherent light ofdm system, adopt the method for recessive training sequence to realize channel estimating and signal recuperation first.The method mainly coherent light ofdm system.Here with the dual-polarization coherent light OFDM (OrthogonalFrequencyDivisionMultiplexing of standard, OFDM) system is example, adopt 2 × 2MIMO (MultipleInputMultipleOutput, enter to have more more) method, Digital Signal Processing is carried out to the dual-polarization signal received.

Technical scheme of the present invention is: a kind of channel recovery method of coherent light ofdm system of recessive training sequence, is characterized in that: it comprises the following steps:

Steps A. will at transmitting terminal training sequence is loaded into and sends on signal, and for same polarization state, the OFDM training symbol that odd number moment, even number moment send is identical;

Step B., at receiving terminal, gets sufficiently long recovery sequence, obtains the channel information on the n-th subcarrier

Then step C. recovers the information in two polarization states on the n-th subcarrier wherein be respectively the data-signal of x-polarisation, y-polarisation recovery.

According to the channel recovery method of the coherent light ofdm system of recessive training sequence as above, it is characterized in that: in described step B, the length recovering sequence is not less than 100.

Compared with prior art, advantage of the present invention is as follows:

The present invention adopts the method for recessive training sequence, and namely training sequence is directly loaded on the signal of transmission, therefore on the spectrum efficiency of signal without any impact.The method when the spectrum efficiency of inhibit signal, can realize channel estimating and signal recuperation.

Accompanying drawing explanation

Fig. 1 is embodiment of the present invention transmitting terminal schematic diagram;

Fig. 2 is the transmitting terminal schematic diagram of traditional scheme;

Embodiment

Below in conjunction with drawings and the specific embodiments, the present invention is described in further detail.

The embodiment of the present invention does not adopt extra time or frequency spectrum to send training sequence, but is directly loaded on signal by training sequence, as shown in Figure 1.Here we are for the orthogonal frequency-division multiplex singal of dual-polarization (x-polarisation and y-polarisation).

We define the mathematic sign that will describe.Definition represent the signal that in transmitting terminal x-polarisation 2m OFDM symbol, the n-th subcarrier sends; Definition represent the signal that in transmitting terminal y-polarisation 2m OFDM symbol, the n-th subcarrier sends; Definition represent the signal of the n-th subcarrier reception in receiving terminal x-polarisation 2m OFDM symbol; Definition represent the signal of the n-th subcarrier reception in receiving terminal x-polarisation 2m OFDM symbol; represent the channel information of the n-th subcarrier respectively.So in double polarizing light ofdm system, for the n-th subcarrier, the channel model on its frequency domain can be expressed as:

[\begin{matrix} r_{x}^{n} (2 m) \\ r_{y}^{n} (2 m) \end{matrix}] = [\begin{matrix} H_{x x}^{n} & H_{x y}^{n} \\ H_{y x}^{n} & H_{y y}^{n} \end{matrix}] [\begin{matrix} s_{x}^{n} (2 m) \\ s_{y}^{n} (2 m) \end{matrix}] - - - (1)

Wherein transmit by data-signal with training sequence signal composition, namely

s_{x}^{n} (2 m) = t_{x}^{n} (2 m) + d_{x}^{n} (2 m) - - - (2)

In like manner, transmit by data-signal with training sequence signal composition, namely

s_{y}^{n} (2 m) = t_{y}^{n} (2 m) + d_{y}^{n} (2 m) - - - (3)

So, we can obtain

[\begin{matrix} r_{x}^{n} (2 m) \\ r_{y}^{n} (2 m) \end{matrix}] = [\begin{matrix} H_{x x}^{n} & H_{x y}^{n} \\ H_{y x}^{n} & H_{y y}^{n} \end{matrix}] [\begin{matrix} t_{x}^{n} (2 m) + d_{x}^{n} (2 m) \\ t_{y}^{n} (2 m) + d_{y}^{n} (2 m) \end{matrix}] - - - (4)

Namely

[\begin{matrix} r_{x}^{n} (2 m) \\ r_{y}^{n} (2 m) \end{matrix}] = [\begin{matrix} H_{x x}^{n} & H_{x y}^{n} \\ H_{y x}^{n} & H_{y y}^{n} \end{matrix}] [\begin{matrix} d_{x}^{n} (2 m) \\ d_{y}^{n} (2 m) \end{matrix}] + [\begin{matrix} H_{x x}^{n} & H_{x y}^{n} \\ H_{y x}^{n} & H_{y y}^{n} \end{matrix}] [\begin{matrix} t_{x}^{n} (2 m) \\ t_{y}^{n} (2 m) \end{matrix}] - - - (4)

Below, we are to training sequence signal with design.We design:

\begin{matrix} t_{x}^{n} (1) = t_{x}^{n} (3) = t_{x}^{n} (5) = ... = t_{x}^{n} (2 m - 3) = t_{x}^{n} (2 m - 1) = T_{x}^{n}; \\ t_{x}^{n} (2) = t_{x}^{n} (4) = t_{x}^{n} (6) = ... = t_{x}^{n} (2 m - 2) = t_{x}^{n} (2 m) = 0; t_{y}^{n} (1) = \\ t_{y}^{n} (3) = t_{y}^{n} (5) = ... = t_{y}^{n} (2 m - 3) = t_{y}^{n} (2 m - 1) = 0; t_{y}^{n} (2) = \\ t_{y}^{n} (4) = t_{y}^{n} (6) = ... = t_{y}^{n} (2 m - 2) = t_{y}^{n} (2 n m) = T_{y}^{n}; \end{matrix}

Have randomness owing to sending signal, its average is 0.So we get sufficiently long one section of recovery sequence (such as 1000 OFDM symbol, i.e. 2m=1000, m=500), and in general, the length recovering sequence should be not less than 100, and we have even number OFDM symbol

\begin{matrix} [\begin{matrix} Σ_{m = 1}^{500} r_{x}^{n} (2 m) \\ Σ_{m = 1}^{500} r_{y}^{n} (2 m) \end{matrix}] \\ = [\begin{matrix} H_{x x}^{n} & H_{x y}^{n} \\ H_{y x}^{n} & H_{y y}^{n} \end{matrix}] [\begin{matrix} Σ_{m = 1}^{500} d_{x}^{n} (2 m) \\ Σ_{m = 1}^{500} d_{y}^{n} (2 m) \end{matrix}] \\ + [\begin{matrix} H_{x x}^{n} & H_{x y}^{n} \\ H_{y x}^{n} & H_{y y}^{n} \end{matrix}] [\begin{matrix} Σ_{m = 1}^{500} t_{x}^{n} (2 m) \\ Σ_{m = 1}^{500} t_{y}^{n} (2 m) \end{matrix}] \end{matrix} - - - (5)

Due to with can be similar to and think 0.And

Σ_{m = 1}^{500} t_{x}^{n} (2 m) = 0.

Therefore

H_{x y}^{n} = Σ_{m = 1}^{500} r_{x}^{n} (2 m) / Σ_{m = 1}^{500} t_{y}^{n} (2 m) - - - (6)

H_{y y}^{n} = Σ_{m = 1}^{500} r_{y}^{n} (2 m) / Σ_{m = 1}^{500} t_{y}^{n} (2 m) - - - (7)

In like manner, we have odd number OFDM symbol

\begin{matrix} [\begin{matrix} Σ_{m = 1}^{500} r_{x}^{n} (2 m - 1) \\ Σ_{m = 1}^{500} r_{y}^{n} (2 m - 1) \end{matrix}] \\ = [\begin{matrix} H_{x x}^{n} & H_{x y}^{n} \\ H_{y x}^{n} & H_{y y}^{n} \end{matrix}] [\begin{matrix} Σ_{m = 1}^{500} d_{x}^{n} (2 m - 1) \\ Σ_{m = 1}^{500} d_{y}^{n} (2 m - 1) \end{matrix}] \\ + [\begin{matrix} H_{x x}^{n} & H_{x y}^{n} \\ H_{y x}^{n} & H_{y y}^{n} \end{matrix}] [\begin{matrix} Σ_{m = 1}^{500} t_{x}^{n} (2 m - 1) \\ Σ_{m = 1}^{500} t_{y}^{n} (2 m - 1) \end{matrix}] \end{matrix} - - - (8)

Due to

Σ_{m = 1}^{500} d_{x}^{n} (2 m - 1)

With

Σ_{m = 1}^{500} d_{x}^{n} (2 m - 1)

Can be similar to and think 0.And

Σ_{m = 1}^{500} t_{y}^{n} (2 m - 1) = 0.

Therefore

H_{x x}^{n} = Σ_{m = 1}^{500} r_{x}^{n} (2 m - 1) / Σ_{m = 1}^{500} t_{x}^{n} (2 m - 1) - - - (9)

H_{y x}^{n} = Σ_{m = 1}^{500} r_{y}^{n} (2 m - 1) / Σ_{m = 1}^{500} t_{x}^{n} (2 m - 1) - - - (10)

Therefore, we can obtain for the channel information on the n-th subcarrier

Based on formula (4), the information that we just can recover in two polarization states on the n-th subcarrier is:

[\begin{matrix} d_{x 1}^{n} (2 m) \\ d_{y 1}^{n} (2 m) \end{matrix}] = {[\begin{matrix} H_{x x}^{n} & H_{x y}^{n} \\ H_{y x}^{n} & H_{y y}^{n} \end{matrix}]}^{- 1} [\begin{matrix} r_{x}^{n} (2 m) \\ r_{y}^{n} (2 m) \end{matrix}] - [\begin{matrix} t_{x}^{n} (2 m) \\ t_{y}^{n} (2 m) \end{matrix}] - - - (11)

Wherein be respectively the data-signal of x-polarisation, y-polarisation recovery.

As a comparison, the method for designing of traditional training sequence for channel estimating is we illustrated.In the derivation of formula, have ignored phase noise, the impact of white Gaussian noise and frequency shift (FS).This is because this kind of problem can process separately by traditional method, and the channel estimation process in this programme can not be affected

The present invention is not limited to above-mentioned execution mode, and for those skilled in the art, under the premise without departing from the principles of the invention, can also make some improvements and modifications, these improvements and modifications are also considered as within protection scope of the present invention.The content be not described in detail in this specification belongs to the known prior art of professional and technical personnel in the field.

Claims

1. a channel recovery method for the coherent light ofdm system of recessive training sequence, is characterized in that: it comprises the following steps:

2. the channel recovery method of the coherent light ofdm system of recessive training sequence according to claim 1, is characterized in that: in described step B, and the length recovering sequence is not less than 100.

3. the channel recovery method of the coherent light ofdm system of recessive training sequence according to claim 1, is characterized in that: in described step B,

H_{x y}^{n} = Σ_{m = 1}^{500} r_{x}^{n} (2 m) / Σ_{m = 1}^{500} t_{y}^{n} (2 m)

H_{y y}^{n} = Σ_{m = 1}^{500} r_{y}^{n} (2 m) / Σ_{m = 1}^{500} t_{y}^{n} (2 m)

H_{x x}^{n} = Σ_{m = 1}^{500} r_{x}^{n} (2 m - 1) / Σ_{m = 1}^{500} t_{x}^{n} (2 m - 1)

H_{y x}^{n} = Σ_{m = 1}^{500} r_{y}^{n} (2 m - 1) / Σ_{m = 1}^{500} t_{x}^{n} (2 m - 1) .

4. the channel recovery method of the coherent light ofdm system of recessive training sequence according to claim 1, is characterized in that: in described step C,

[\begin{matrix} d_{x 1}^{n} (2 m) \\ d_{y 1}^{n} (2 m) \end{matrix}] = {[\begin{matrix} H_{x x}^{n} & H_{x y}^{n} \\ H_{y x}^{n} & H_{y y}^{n} \end{matrix}]}^{- 1} [\begin{matrix} r_{x}^{n} (2 m) \\ r_{y}^{n} (2 m) \end{matrix}] - [\begin{matrix} t_{x}^{n} (2 m) \\ t_{y}^{x} (2 m) \end{matrix}] .