CN114785650A

CN114785650A - Novel blind phase search algorithm structure and implementation method

Info

Publication number: CN114785650A
Application number: CN202210505811.1A
Authority: CN
Inventors: 乔耀军; 孙中良
Original assignee: Beijing University of Posts and Telecommunications
Current assignee: Beijing University of Posts and Telecommunications
Priority date: 2022-05-10
Filing date: 2022-05-10
Publication date: 2022-07-22
Anticipated expiration: 2042-05-10
Also published as: CN114785650B

Abstract

The invention discloses a novel blind phase search algorithm structure and an implementation method thereof, which can effectively improve the performance damage of the existing blind phase search algorithm in a low signal-to-noise ratio optical fiber communication system, obviously reduce the length of a required filter and improve the accuracy of phase estimation. The invention is realized by the following steps: firstly, rotating an input signal according to a test angle; the rotated signal is sent to a decision circuit, and a standard constellation point closest to the Euclidean distance of the rotated signal is output; calculating the difference value between the rotated signal and the standard constellation point and taking the square, thereby obtaining the square distance between the rotated signal and the standard constellation point; initializing a Gaussian fuzzy filter; convolving a Gaussian filter with square distances with the same test angle; performing summation operation on the updated square distance; sending the summed square distance sum to a minimum comparator, and outputting a test angle corresponding to the minimum sum as phase noise estimated by an algorithm; and when all the signals are output, finishing the whole algorithm.

Description

Novel blind phase search algorithm structure and implementation method

Technical Field

The invention relates to the technical field of communication, in particular to a novel blind phase search algorithm structure and an implementation method

Background

With the advent of big data, cloud computing, and various emerging intelligent applications, global communication data traffic is increasing, and the demand for data transmission rates is on an explosive growth trend. Probability shaping techniques have received extensive attention in both academia and industry due to their flexibility in rate adjustment and their ability to more closely approach shannon limits. On the other hand, the probability shaping technology enables each constellation point of the signal not to be distributed in equal probability, especially the probability of occurrence of high-amplitude signal points is lower, which is equivalent to reducing the modulation order of the signal, and therefore the requirement of the signal-to-noise ratio required by the system is reduced. However, the traditional digital signal processing algorithms are originally designed based on the premise of equal probability distribution of signal constellation points, so that the direct application of the algorithms to a probability shaping system may have incompatibility problems. It has been reported that the performance of the carrier phase recovery algorithm, especially the Blind phase search algorithm (BPS), in the probability shaping system is severely degraded, mainly by the reduction of the mutual information of the system. This degradation is more pronounced, especially in lower signal-to-noise ratios.

How to deal with the degradation of the performance of the blind phase search algorithm in the low signal-to-noise ratio optical fiber communication system has attracted extensive attention in research. In general, in order to cope with such performance deterioration, the conventional blind phase search algorithm reduces the performance loss by increasing the filter length, but an excessively long filter length also reduces the ability of tracking the phase while increasing the complexity of algorithm processing and processing delay.

Recently, one has adopted the addition of a forgetting factor in the blind phase search algorithm to improve its performance. However, the selection of the forgetting factor needs to be obtained through a complicated optimization process, and the forgetting factor is also related to the signal-to-noise ratio of the system. Different signal-to-noise ratios require different forgetting factors to be changed, which both increase system cost.

The invention comprehensively considers the characteristics of a blind phase search algorithm of a low signal-to-noise ratio optical fiber communication system, and provides a novel blind phase search algorithm structure and an implementation method.

Disclosure of Invention

The invention aims to improve the performance damage of the existing blind phase search algorithm in a low signal-to-noise ratio optical fiber communication system, greatly reduce the length of a filter and introduce an unequal-weight filter, thereby providing a blind phase search algorithm (GBA-BPS) based on Gaussian blur.

The GBA-BPS algorithm of the invention comprises the following concrete implementation steps:

step 1: to-be-processed signal r containing phase noise_kWith B test phasesBit multiplication, i.e. rotation of the signal in the plane of the constellation,

wherein the test phase expression is characterized by (1)

Wherein the content of the first and second substances,

denotes a test phase, B denotes the number of test phases, and B denotes an index of the test phase.

It can be seen that the parameter B directly affects the accuracy of the algorithm, and a larger B makes the phase noise estimation more accurate but at the same time results in a larger amount of computation and complexity of the algorithm.

Step 2: the signal obtained in the step 1 enters a decision circuit and outputs a standard constellation point which is closest to the signal obtained in the step 1,

and 3, step 3: subtracting the constellation points obtained in the step (1) and the step (2) to obtain a difference, and squaring to obtain the square distance between the rotated signal and the ideal constellation points;

and 4, step 4: initializing a Gaussian fuzzy filter; the gaussian blur filter formula is characterized by (2):

the length of the Gaussian filter is N, the length setting of the Gaussian fuzzy filter is consistent with the filter length of the existing blind phase search algorithm, and specific parameters are adjusted according to actual conditions;

wherein, the variance of the Gaussian blur filter is the expectation of the square of the signal at the transmitting end, and the calculation formula is characterized by (3):

σ²＝E(|A_k|²) (3)

wherein σ²Representing the variance of the gaussian blur filter, E representing the desired result, | | | representing the result of modulo，A_kSignals representing the transmitting end

And 5: updating the squared distance obtained in step 3 with a Gaussian blur filter, and the convolution formula is characterized by (4):

the square distances with the same test angle obtained in the step 3 are respectively convolved with the Gaussian blur filter obtained in the step 4 to obtain updated square distances;

wherein N is_sDenotes the number of symbols to be processed, G_λB represents the index of the test angle;

step 6: the sum of squared distances from step 4 is summed at the same test angle, and the calculation formula is characterized by (5):

wherein, | d_m-n,B|²Represents the squared distance after updating by the gaussian filter, N represents the length of the filter;

and 7: sending the sum of squared distances obtained in the step 6 to a minimum comparator to obtain a minimum sum of squared distances, wherein a corresponding test angle is the phase noise estimated by the final algorithm;

so far, from step 1 to step 7, the whole algorithm operation is finished.

Advantageous effects

The novel blind phase search algorithm structure and the implementation method provided by the invention have the effect of effectively reducing the length of the filter and the complexity of the algorithm due to the consideration of the correlation of the front symbol and the rear symbol. Meanwhile, because different weights are given to the taps of the filter through Gaussian blur, compared with the existing blind phase search algorithm, the blind phase search algorithm has a better effect of eliminating white noise, and has an effect of assisting the blind phase search algorithm in accurately estimating the phase. Therefore, the novel blind phase search algorithm provided by the invention can simultaneously reduce the length of the filter and improve the accuracy of phase estimation.

Drawings

FIG. 1 is a schematic diagram of a novel blind phase search algorithm proposed by the present invention;

FIG. 2 is a diagram of a simulation system of the present invention in embodiment 1;

FIG. 3 is a graph of the shaping factor-mutual information of the new blind phase search algorithm and the existing blind phase search algorithm employed in the present invention in example 1;

Detailed Description

Example 1

In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention are clearly and completely described below with reference to the drawings in the embodiments of the present invention.

Fig. 1 is a diagram of a simulation system of this embodiment 2, and it can be seen that the modulation formats are two (PS-16QAM and PS-64 QAM). At a transmitting end, firstly, a pseudo-random sequence is mapped into an amplitude sequence of probability distribution through a distribution matcher, and then symbol mapping of 16QAM and 64QAM is carried out. And then transmitted in an additive white gaussian noise channel. In this example, the laser linewidth is set to 100kHz, the shaping factor is set to 0 to 0.15, and the adjustment step size is 0.01. In the off-line processing process of the receiving end, signals enter a carrier phase recovery algorithm module through data, and then mutual information calculation is carried out after cycle skipping is eliminated.

Fig. 3 shows the relationship between the mutual information of the receiving end and the shaping factor of the transmitting end under the condition of using the Conventional blind phase search algorithm Conventional BPS and the algorithm GBA-BPS proposed in the present invention, where the abscissa is the shaping factor of the transmitting end and the ordinate is the mutual information of the receiving end. It can be found that under the same filter length, the mutual information which can be achieved by the blind phase search algorithm based on the Gaussian blur is obviously higher than that of the existing blind phase search algorithm.

While the foregoing is directed to the preferred embodiment of the present invention, it is not intended that the invention be limited to the embodiment and the drawings disclosed herein. Equivalents and modifications may be made without departing from the spirit of the disclosure, which is to be considered as within the scope of the invention.

Claims

1. A novel blind phase search algorithm structure and a realization method are characterized in that the algorithm comprises the following steps:

step 1: rotating the input signal according to the test angle;

step 2: the signal obtained in the step 1 enters a decision circuit, and a standard constellation point which is closest to the Euclidean distance of the signal obtained in the step 1 is output;

and step 3: subtracting the standard constellation point obtained in the step (2) from the signal obtained in the step (1), and then taking the square, thereby obtaining the square distance between the rotated signal and the standard constellation point;

and 4, step 4: initializing a Gaussian fuzzy filter;

wherein, the length of the Gaussian filter is N;

wherein, the variance of the Gaussian blur filter is the expectation of the square of the signal at the transmitting end, and the calculation formula is characterized by (1):

σ²＝E(|A_k|²) (1)

wherein σ²Representing the variance of the gaussian blur filter, E representing the desired result, | | | representing the modulo result, a_kA signal indicating a transmitting end;

and 5: updating the square distance obtained in the step 3 by using the Gaussian blur filter obtained in the step 4;

and 6: summing the sum of squared distances obtained in step 5 under the same test angle;

and 7: sending the sum of squared distances obtained in the step 6 to a minimum comparator to obtain a minimum sum of squared distances, wherein the corresponding test angle is the phase noise estimated by the final algorithm;

from step 1 to step 7, the whole algorithm operation is finished.

2. A novel blind phase search algorithm structure as claimed in claim 1, characterized in that the described square distance updating filter can be a gaussian blur filter or other non-equal weight filter, and the specific filter type should be adjusted according to the actual situation.

3. The structure of claim 1, wherein in step 4, the initialization of the gaussian blur filter: the variance of the gaussian blur filter is set to the expectation of the square of the signal at the transmit end.

4. The structure of a novel blind phase search algorithm as claimed in claim 1, wherein in step 4, the initialization of the gaussian blur filter: the length setting of the Gaussian blur filter is consistent with the filter length of the existing blind phase search algorithm, and specific parameters are adjusted according to actual conditions.

5. The structure of a novel blind phase search algorithm as claimed in claim 1, wherein in step 5, the gaussian blur filter updates the squared distance: and the square distances with the same test angle are respectively convolved with the Gaussian blur filter, and new square distances are updated and established.