CN101834825A - Channel compensation method based on Jones matrix resisting ASE noise in optical OFDM system - Google Patents

Abstract

The invention relates to a channel compensation method based on Jones matrix resisting ASE (Amplified Spontaneous Emission Noise) noise in an optical OFDM (Orthogonal Frequency Division Multiplexing) system, belongs to the optical communication field and is mainly applied in (Polarization Division Multiplexed) PDM optical OFDM system. The method comprises the steps of: inserting a specific training sequence in a transmitter of a PDM-OFDM system, estimating a Jones matrix of the light path by using the sent training sequence in a receiver part of the system through the channel compensation method, and multiplying the received data by a fake inverse matrix of the Jones matrix so as to estimate the initially sent signal. The technical scheme of the invention can efficiently lighten the impact of the ASE noise in the light path on the system property.

Description

In the optical OFDM system based on the channel compensating method of Jones matrix resisting ASE noise

Technical field

The present invention's design relates to the light positive of a kind of anti-ASE based on Jones matrix (Amplified Spontaneous Emission noise) noise and hands over frequency division multiplexing (Orthogonal Frequency Division Multiplexing, OFDM) the channel estimating compensation method of system, belong to optical communication field, (Polarization DivisionMultiplexed-Orthogonal Frequency Division Multiplex PDM-OFDM) carries out channel compensation in the system to be applied in the OFDM of the multiplexing light of polarization mode.

Background technology

As a kind of multi-carrier transmission pattern, light positive hands over frequency division multiplexing (OFDM) technology to be converted to one group of low-speed parallel data flow transmitted by the data flow with one group of high-speed transfer, the optical communication system of feasible employing OFDM technology reduces greatly to the susceptibility of chromatic dispersion, polarization mode dispersion etc., and Cyclic Prefix (Cycle Prefix in the OFDM symbol, CP) introducing has further strengthened the resisting chromatic dispersion ability of ofdm communication system again.In addition, the bandwidth availability ratio height of ofdm communication system realizes that characteristics such as simple make the OFDM technology at optical communication field, and especially the above field of high-speed optical communications of 100Gbit/s has very fine application prospect.At present for the high speed ofdm system, generally adopt the mode of the multiplexing ofdm system of polarization mode (PDM-OFDM) to realize, spectrum efficiency is doubled, adopt polarization classification coherent reception and electric territory channel estimating at receiving terminal, can well eliminate polarization mode dispersion (Polarization Mode Dispersion, influence PMD).

For optical OFDM system, a very big advantage is can insert training sequence at transmitting terminal to estimate channel response, can obtain channel information more accurately like this, and then can handle the optical channel transmission impairment very easily, as chromatic dispersion (Chromatic Dispersion, CD) and polarization mode dispersion (PMD) etc.But for the optical transmission ofdm system, the accuracy of channel estimating often is subjected to the influence of light path noise (the especially ASE noise of amplifier).In order to improve the accuracy of channel estimating, often be averaged then and realize, but insert the efficiency of transmission that too many training sequence can reduce system's useful information by once sending a plurality of training sequences.Therefore have contradiction between the accuracy of channel estimating and the information transmission efficiency, how obtaining accurately with the least possible training sequence, channel estimating is a difficult point of the research of optical OFDM system.

Summary of the invention

The present invention is directed to the PDM-OFDM system and propose a kind of new channel estimating compensation method, at first insert specific training sequence,, adopt this method to carry out channel compensation then in the electric territory of PDM-OFDM system receiver part in the electric territory of PDM-OFDM system transmitter.Adopting same number of training sequence, same Optical Signal To Noise Ratio (Optical Signal Noise Ratio, OSNR) under, compare with the channel estimating that adopts traditional time division multiplexing algorithm to carry out, this method can obviously improve the noise immunity of channel estimating, improves the overall performance of PDM-OFDM system.

The present invention is as follows:

The PDM-OFDM system comprises transmitter, light path and receiver section.Transmitter section in the PDM-OFDM system, after the mapping of the bit sequence that sends through M-PSK (M-Phase Shift Keying) or M-QAM modulation formats such as (M-Quadrature AmplitudeModulation), be split two sequences of X, Y, pass through zero padding respectively, IFFT (InverseFast Fourier Transform), add CP after, adopting the IQ modulator to be modulated to light carrier respectively gets on, (Polarization Beam Splitter PBS) is multiplexed on the polarization state of two quadratures of light carrier to adopt the polarization spectro sheet then.The receiver section of system, the polarization state that at first received signal is divided into any two quadratures with PBS, these two parts are mixing at random of initialization signal, for amplitude and the phase information that can recover this two parts light signal, adopt coherent optical heterodyne communicatio respectively, adopt similar wireless MIMO (Multiple Input Multiple Output) processing method then, polarization state is rotated and be divided into the two-way received signal.

As if the sequence of representing with a vector to send

$t=\left[\begin{array}{c}{t}_x\\ t_y\end{array}\right]$

T wherein _x, t _yRepresent X respectively, two sequence datas of Y, the X and the Y polarization state that are loaded into light carrier respectively get on.

For ofdm system, when the number of sub carrier wave that adopts was very big, very narrow of the frequency band that is occupied at each subcarrier of frequency domain was to such an extent as to that the domain channel response of each subcarrier can be thought is smooth, the time is constant.CD in channel then, PMD, under the combined effect of PDL (Polarization Dependent Loss) etc., channel model can be expressed as

r(k)＝H(k)t(k)+n(k)

Promptly

$\left[\begin{array}{c}{r}_x\left(k\right)\\ r_y\left(k\right)\end{array}\right]=\left[\begin{array}{cc}{h}_{11}\left(k\right)& h_{12}\left(k\right)\\ h_{21}\left(k\right)& h_{22}\left(k\right)\end{array}\right]\left[\begin{array}{c}{t}_x\left(k\right)\\ t_y\left(k\right)\end{array}\right]+\left[\begin{array}{c}{n}_x\left(k\right)\\ n_y\left(k\right)\end{array}\right]$

In the formula, vector r (k), t (k) represent respectively that k subcarrier uploaded and send and receive ofdm signal, comprise two orthogonal polarisation state information r respectively _x(k) and r _y(k), t _x(k) and t _y(k).2*2 matrix H (k) is the channel response matrix (Jones matrix) of k subcarrier, the linear response of expression channel, h _Ij(k) be matrix component.Frequency domain noise on two polarization states of k subcarrier of vector n (k) expression comprises two orthogonal polarisation state noise n _x(k) and n _y(k).By following formula as seen, the noise that is produced by image intensifer or other noise sources also can cause H (k) misjudgment.

Therefore, in order to estimate channel Jones matrix H (k) accurately in noisy channel, the present invention proposes a kind of new channel estimating compensation method.At first, the transmitter terminal in the PDM-OFDM system, a pair of training sequence that contains two polarization state data is one in front and one in back inserted in the front of transmission data, is written as t here ₁, t ₂, promptly

$t_1=\left[\begin{array}{c}{t}_x\\ t_y\end{array}\right]{\mathrm{<figure-callout\; id="2"\; label="t"\; filenames="HSA00000093585000022.png"\; state="\{\{state\}\}">t</figure-callout>}}_2=\left[\begin{array}{c}{t}_x\\ -t_y\end{array}\right]$

T wherein _x, t _yBe two known sequences, have very low peak-to-average power ratio (Peak-to-Average Power Ratio, PAPR).By following formula as seen, the training sequence that is applicable to this method has following characteristics: t ₁, t ₂The X polarization state data of two training sequences are identical, and Y polarization state data are opposite each other.

Through after the optical channel transmission, be with symbol on k the subcarrier of two corresponding receiving sequences of training sequence

$\left[\begin{array}{c}{r}_{1x}\left(k\right)\\ r_{1y}\left(k\right)\end{array}\right]=\left[\begin{array}{cc}{h}_{11}\left(k\right)& h_{12}\left(k\right)\\ h_{21}\left(k\right)& h_{22}\left(k\right)\end{array}\right]\left[\begin{array}{c}{t}_{1x}\left(k\right)\\ t_{1y}\left(k\right)\end{array}\right]\left[\begin{array}{c}{h}_{11}\left(k\right)\&CenterDot;t_{1x}\left(k\right)+h_{12}\left(k\right)\&CenterDot;t_{1y}\left(k\right)\\ h_{21}\left(k\right)\&CenterDot;t_{1x}\left(k\right)+h_{22}\left(k\right)\&CenterDot;t_{1y}\left(k\right)\end{array}\right]$

$\left[\begin{array}{c}{r}_{2x}\left(k\right)\\ r_{2y}\left(k\right)\end{array}\right]=\left[\begin{array}{cc}{h}_{11}\left(k\right)& h_{12}\left(k\right)\\ h_{21}\left(k\right)& h_{22}\left(k\right)\end{array}\right]\left[\begin{array}{c}{t}_{1x}\left(k\right)\\ -t_{1y}\left(k\right)\end{array}\right]\left[\begin{array}{c}{h}_{11}\left(k\right)\&CenterDot;t_{1x}\left(k\right)-h_{12}\left(k\right)\&CenterDot;t_{1y}\left(k\right)\\ h_{21}\left(k\right)\&CenterDot;t_{1x}\left(k\right)-h_{22}\left(k\right)\&CenterDot;t_{1y}\left(k\right)\end{array}\right]$

T wherein _1x(k), t _1y(k) in first training sequence that sends be modulated to k on the subcarrier X and the data on the Y orthogonal polarisation state, t _2x(k), t _2y(k) in second training sequence that sends be modulated to k on the subcarrier X and the data on the Y orthogonal polarisation state, r _1x(k), r _1y(k) for receive corresponding in first training sequence be modulated to k on the subcarrier X and the data on the Y orthogonal polarisation state, r _2x(k), r _2y(k) for receive corresponding in second training sequence be modulated to k on the subcarrier X and the data on the Y orthogonal polarisation state, h _Ij(k) be the channel response Jones matrix component of k subcarrier.

By above two formulas, the Jones matrix that obtains channel is

$\tilde{H}\left(k\right)=\left[\begin{array}{cc}{h}_{11}\left(k\right)& h_{12}\left(k\right)\\ h_{21}\left(k\right)& h_{22}\left(k\right)\end{array}\right]=\left[\begin{array}{cc}(r_{1x}\left(k\right)+r_{2x}\left(k\right))/t_{1x}\left(k\right)/2& (r_{1x}\left(k\right)-r_{2x}\left(k\right))/t_{1y}\left(k\right)/2\\ (r_{1y}\left(k\right)+r_{2y}\left(k\right))/t_{1x}\left(k\right)/2& (r_{1y}\left(k\right)-r_{2y}\left(k\right))/t_{1y}\left(k\right)/2\end{array}\right]$

Following formula is exactly the channel response matrix that the inventive method is estimated k subcarrier Realization.

Obtain Jones matrix

After, channel compensation adopts zero forcing algorithm, and it is that the signal r (k) that contains noise that will receive multiply by the pseudo inverse matrix (Pseudo-Inverse Matrix) of estimating the channel matrix that obtains.For the k subcarrier, adopt

Approximate representation channel matrix H (k) estimates that then the transmission vector that obtains is

$\tilde{t}\left(k\right)={\tilde{H}}^{*}\left(k\right)r\left(k\right)$

Superscript symbol * represents pseudo-inverse operation, is defined as

H ^*＝(H ^HH) ^-1H ^H

Wherein, operator H and-1 represents conjugate transpose and matrix inversion respectively, supposes it is desirable channel estimating, then

Therefore estimate Can be expressed as

$\tilde{t}\left(k\right)=t\left(k\right)+H^{*}\left(k\right)n\left(k\right)$

Carry out pseudo-inverse operation then, ask

${\tilde{H}}^{*}\left(k\right)={\left({\tilde{H}}^H\left(k\right)\tilde{H}\left(k\right)\right)}^{-1}{\tilde{H}}^H\left(k\right)$

Estimate that at last the initialize signal that obtains transmitting is

$\tilde{t}\left(k\right)={\tilde{H}}^{*}\left(k\right)r\left(k\right)$

Wherein r (k) is received data on k the subcarrier,

Be the primary data that sends on k the subcarrier that uses this method to estimate to obtain.

Description of drawings

Fig. 1: the PDM-CO-OFDM overall system framework schematic diagram of using this method;

Fig. 2: a pair of training sequence schematic diagram that this method is used;

Fig. 3: the training sequence schematic diagram that traditional time division multiplexing algorithm inserts;

Fig. 4: adopting under this method and the traditional time division multiplexing algorithm condition, the relation curve of systematic function and OSNR (back-to-back, BtoB);

Fig. 5: adopting under this method and the traditional time division multiplexing algorithm condition relation curve of systematic function and OSNR (800km);

Embodiment

Below in conjunction with the accompanying drawing in the embodiment of the invention, the technical scheme in the embodiment of the invention is clearly and completely described, obviously, described embodiment also only is a part of embodiment of the present invention, rather than whole embodiment.Based on the embodiment among the present invention, those of ordinary skills belong to the scope of protection of the invention not making the every other embodiment that is obtained under the creative work prerequisite.

The embodiment of the invention is applicable to 100Gbit/s polarisation of light mode multiplexing OFDM (PMD-CO-OFDM) modulation demodulation system that is concerned with.

Fig. 1 is the structural representation of PMD-CO-OFDM system.

The binary data sequence of 100Gbit/s is at first through the QPSK modulation; be split two sequences of X, Y (baud rate is 25Gbaud/s); carry out data padding (Zero-Padding) then and form the frequency domain protection at interval; pass through inverse-Fourier transform (IFFT) then and add Cyclic Prefix (CP) computing and obtain electric territory ofdm signal; insert a pair of training sequence of this algorithm in the front of data sequence; through being modulated to the light territory behind digital-to-analog conversion (D/A), the IQ modulator (IQ Mod.), obtain the coherent light ofdm signal.The coherent light signal that transmitter sends is transferred to optical fiber link, by common standard optical fiber (Single ModeFiber, SMF) and image intensifer (Erbium-doped fiber amplifier, EDFA) form, do not contain the optical path dispersion compensation, after being transferred to receiver, after polarization classification optical mixer unit and analog/digital conversion (A/D), through synchronously, remove CP, Fourier transform (FFT) is removed zero padding value (zero-padding), preliminary dispersion compensation algorithm (Electrical DispersionCompensation, EDC) after, carry out the channel estimating of this algorithm, then through laser make an uproar mutually backoff algorithm (PilotAssistedCommon Phase Error Compensation, PA-CPEC) after, obtain X, the Y transfer sequence, process QPSK demodulation obtains the binary sequence of initial transmission again.

Entire method comprises two parts:

One is the insertion of the training sequence of transmitter, is positioned at system transmitter.After this module is positioned at and adds the CP module, before forming OFDM time domain data stream, insert a pair of training sequence that this method is used.The a pair of training sequence schematic diagram that Fig. 2 adopts for this method, as seen from the figure, training sequence is a pair of sequence that comprises two orthogonal polarisation state data respectively, and length is the sequence length of an OFDM symbol, correlation is that the X polarization data is identical, and the Y polarization data is opposite each other.Two sequences are as follows

$t_1=\left[\begin{array}{c}{t}_x\\ t_y\end{array}\right],$ $t_2=\left[\begin{array}{c}{t}_x\\ -t_y\end{array}\right]$

Fig. 3 is the training sequence schematic diagram that traditional time division multiplexing algorithm uses.Two training sequences are respectively

$t_1=\left[\begin{array}{c}{t}_x\\ 0\end{array}\right],$ ${\mathrm{<figure-callout\; id="2"\; label="t"\; filenames="HSA00000093585000022.png"\; state="\{\{state\}\}">t</figure-callout>}}_2=\left[\begin{array}{c}0\\ t_y\end{array}\right]$

Another part is positioned at the receiver of system, utilizes the training sequence that inserts to carry out the channel estimating compensation.Algoritic module is positioned at after the electric territory EDC algorithm of receiver, and by this a pair of known training sequence, the frequency domain response (Jones matrix) that obtains each subcarrier correspondence is

$\tilde{H}\left(k\right)=\left[\begin{array}{cc}{h}_{11}\left(k\right)& h_{12}\left(k\right)\\ h_{21}\left(k\right)& h_{22}\left(k\right)\end{array}\right]=\left[\begin{array}{cc}(r_{1x}\left(k\right)+r_{2x}\left(k\right))/t_{1x}\left(k\right)/2& (r_{1x}\left(k\right)-r_{2x}\left(k\right))/t_{1y}\left(k\right)/2\\ (r_{1y}\left(k\right)+r_{2y}\left(k\right))/t_{1x}\left(k\right)/2& (r_{1y}\left(k\right)-r_{2y}\left(k\right))/t_{1y}\left(k\right)/2\end{array}\right]$

Try to achieve pseudo inverse matrix then

${\tilde{H}}^{*}\left(k\right)={\left({\tilde{H}}^H\left(k\right)\tilde{H}\left(k\right)\right)}^{-1}{\tilde{H}}^H\left(k\right)$

Estimate that at last obtaining sending signal is

$\tilde{t}\left(k\right)={\tilde{H}}^{*}\left(k\right)r\left(k\right)$

In order to verify the effect of this method, simulation study PMD-CO-OFDM system (Fig. 1) adopt traditional time division multiplexing algorithm and this method to carry out the result of channel estimating respectively.Basic PMD-CO-OFDM system parameter setting is as follows: bit rate is 100Gbit/s, and modulation format is QPSK, and the number of sub carrier wave of employing is 2048, the frequency domain zero insertion is 50%, CP is 25%, and the laser transmitting power of transmitter and receiver is 10mW, and live width is 300kHz.

Fig. 4 is under the condition of (BtoB) back-to-back, adopts the relation curve of this method and traditional time division multiplexing algorithmic system performance and OSNR.As seen from the figure, under the condition of identical Optical Signal To Noise Ratio (OSNR), this method can improve the performance of system preferably.At native system, under the BtoB condition, adopt the Q of this method system approximately can improve 2dB.

After Fig. 5 is system transmissions 800km, under first section different optical fiber launched powers, adopt the simulation result of the system of this method and contrast algorithm.Link amounts to 10 sections, and every section optical fiber comprises SMF (abbe number 16ps/nm/km, the non linear coefficient 2.6 * 10 of 80km ^-20m ²/ W) and EDFA (amplification coefficient is 16dB, NF=6dB), has considered chromatic dispersion, non-linear and ASE noise.As seen from Figure 5, launched power hour (＜-2dB), system for restricting mainly be the ASE noise, adopt this algorithm can obviously improve systematic function, Q has approximately improved 2dB (close with the value that improves) under the BtoB condition; When launched power increases (＞0dB), this moment, system was restricted by the nonlinear impairments of optical fiber, and this algorithm and traditional algorithm performance are approximate, prove that this algorithm has identical stability with traditional algorithm when nonlinear impairments is big.

More than be the detailed introduction and the emulation of specific embodiment of adopting a kind of PDM-OFDM system of the inventive method, the explanation of above embodiment just is used for helping to understand method of the present invention and core concept thereof; Simultaneously, for one of ordinary skill in the art, according to thought of the present invention, the part that all can change in specific embodiments and applications, in sum, this description should not be construed as limitation of the present invention.

The major technique advantage

OFDM (OFDM) has been widely used in wireless and wired (copper cash) communications field at present. Light positive hands over frequency division multiplexing (OFDM) system to combine OFDM technology and optical transport technology, has wide development prospect at the high-speed light transmission field, especially the multiplexing PDM-OFDM of polarization pattern system is under the ability with strong resisting chromatic dispersion, the ability that also has stronger anti-polarization mode dispersion, the spectrum efficiency of simultaneity factor doubles, and is one of main research approach of the big capacity of at present research realization, long haul optical transmission.

Compare with other method, propose this new channel estimation compensation process for the PDM-OFDM system among the present invention and have following features: only need to adopt the frequency domain response that specific training sequence just can well the estimating system channel, and can effectively alleviate in the optical channel ASE noise to the impact of systematic function, algorithm is realized simple, estimates accurately.

Claims (7)

1. the light positive of an anti-ASE (Amplified Spontaneous Emission noise) noise is handed over frequency division multiplexing (OrthogonalFrequency Division Multiplexing, OFDM) channel compensating method of system, it is characterized in that described method may further comprise the steps:

Send specific training sequence;

Utilize training sequence to estimate the Jones matrix of channel;

The Jones matrix that utilization obtains carries out channel compensation;

2. method according to claim 1 is characterized in that, the specific implementation of the training sequence that described transmission is specific:

Specific training sequence is that a pair of length is the sequence of a symbol lengths of OFDM, and each sequence contains the training sequence data of two polarization states.

3. method according to claim 2 is characterized in that, the specific implementation of a pair of training sequence of described transmission: this a pair of t that is written as respectively ₁, t ₂, promptly

$t_1=\left[\begin{array}{c}{t}_x\\ t_y\end{array}\right]$ $t_2=\left[\begin{array}{c}{t}_x\\ -t_y\end{array}\right]$

T wherein _x, t _yBe two known sequences, have very low peak-to-average power ratio (Peak-to-Average Power Ratio, PAPR), by following formula as seen, t ₁, t ₂X polarization state data in two training sequences are identical, and Y polarization state data are opposite each other.

4. method according to claim 1 is characterized in that, the described specific implementation of utilizing training sequence to estimate the Jones matrix of channel:

Use the channel Jones matrix on each subcarrier of this a pair of training sequence estimating OFDM system It is frequency domain response.

5. method according to claim 4 is characterized in that, the specific implementation of the Jones matrix on each subcarrier of described estimating OFDM system:

$\tilde{H}\left(k\right)=\left[\begin{array}{cc}{h}_{11}\left(k\right)& h_{12}\left(k\right)\\ h_{21}\left(k\right)& h_{22}\left(k\right)\end{array}\right]=\left[\begin{array}{cc}(r_{1x}\left(k\right)+r_{2x}\left(k\right))/t_{1x}\left(k\right)/2& (r_{1x}\left(k\right)-r_{2x}\left(k\right))/t_{1y}\left(k\right)/2\\ (r_{1y}\left(k\right)+r_{2y}\left(k\right))/t_{1x}\left(k\right)/2& (r_{1y}\left(k\right)-r_{2y}\left(k\right))/t_{1y}\left(k\right)/2\end{array}\right]$

T wherein _1x(k), t _1y(k) in first training sequence that sends be modulated to k on the subcarrier X and the data on the Y orthogonal polarisation state, t _2x(k), t _2y(k) in second training sequence that sends be modulated to k on the subcarrier X and the data on the Y orthogonal polarisation state, r _1x(k), r _1y(k) for receive corresponding in first training sequence be modulated to k on the subcarrier X and the data on the Y orthogonal polarisation state, r _2x(k), r _2y(k) for receive corresponding in second training sequence be modulated to k on the subcarrier X and the data on the Y orthogonal polarisation state, the channel response of k subcarrier is

h _Ij(k) be matrix component.

6. method according to claim 1 is characterized in that, the Jones matrix that described utilization obtains carries out the specific implementation of channel compensation:

Try to achieve channel response After, carry out pseudo-inverse operation, try to achieve pseudo inverse matrix

${\tilde{H}}^{*}\left(k\right)={\left({\tilde{H}}^H\left(k\right)\tilde{H}\left(k\right)\right)}^{-1}{\tilde{H}}^H\left(k\right)$

7. method according to claim 1 is characterized in that, the Jones matrix that described utilization obtains carries out the specific implementation of channel compensation:

By pseudo inverse matrix

Estimating to obtain k subcarrier uploads defeated initialize signal and is

$\tilde{t}\left(k\right)={\tilde{H}}^{*}\left(k\right)r\left(k\right)$

Wherein r (k) is the received data of k subcarrier,

Be the primary data that sends on k the subcarrier that uses this method to estimate to obtain.

Images