CN-121984825-A - Multidimensional constellation probability shaping method based on dimensional energy statistics and entropy constraint

Abstract

The invention discloses a multidimensional constellation probability shaping method based on dimensional energy statistics and entropy constraint, which belongs to the technical field of communication and comprises the steps of obtaining coordinates of constellation points, carrying out statistical analysis on energy distribution characteristics of each dimension to obtain energy statistics of each dimension to construct dimensional weights, constructing dimensional weighted energy metrics of the constellation points based on the dimensional weights and coordinate components of the constellation points on each dimension of a multidimensional modulation constellation, providing a probability mapping function for constructing the constellation points, mapping the dimensional weighted energy metrics into probability weights of the constellation points, carrying out normalization to obtain initial constellation point probability distribution, determining a search interval of probability shaping intensity parameters under constraint conditions of target information entropy, calculating information entropy according to the initial constellation point probability distribution, updating the search interval, searching shaping intensity parameters meeting entropy constraint, and calculating the probability distribution of the constellation points.

Inventors

• WANG CHEN
• REN JIANXIN

Assignees

• 盐城工业职业技术学院

Dates

Publication Date

20260505

Application Date

20260211

Claims (10)

1. 1. The multidimensional constellation probability shaping method based on the dimensional energy statistics and entropy constraint is characterized by comprising the following steps of: Acquiring coordinates of all constellation points in a multidimensional modulation constellation; Based on the coordinates of all constellation points, carrying out statistical analysis on the energy distribution characteristics of each dimension of the multidimensional modulation constellation to obtain energy statistics of each dimension; Constructing a dimension weight reflecting the energy distribution difference of different dimensions through a monotonic mapping function according to the energy statistics of each dimension; based on the dimensional weights and the coordinate components of the constellation points in each dimension of the multidimensional modulation constellation, constructing a dimensional weighted energy measurement of the constellation points by utilizing linear weighting; constructing a probability mapping function of constellation points based on the vector weighting energy measurement and the probability shaping strength parameter, and mapping the vector weighting energy measurement into probability weights of the constellation points; Normalizing the probability of the constellation points based on the probability weights of the constellation points to obtain initial constellation point probability distribution; Determining a search interval of the probability shaping intensity parameter under the constraint condition of the target information entropy; calculating information entropy according to the initial constellation point probability distribution, and updating a search interval based on an error between the information entropy and the target information entropy; Taking the actual information entropy as entropy constraint, and searching shaping strength parameters meeting the entropy constraint in the updated search interval by utilizing bipartite iteration; And calculating the probability distribution of constellation points according to the shaping intensity parameters meeting the entropy constraint.
2. 2. The multidimensional constellation probability shaping method based on the dimensional energy statistics and entropy constraints of claim 1, wherein coordinates of the constellation points are expressed as: ; In the formula, Represent the first The coordinates of the individual constellation points, Represent the first The coordinate components of the individual constellation points in the 1 st dimension of the multi-dimensional modulation constellation, Represent the first The coordinate components of the individual constellation points in the 2 nd dimension of the multi-dimensional modulation constellation, Represent the first Constellation point in multidimensional modulation constellation The component of the coordinates in the dimension, Representing the total number of constellation points, And representing the total number of dimensions of the multi-dimensional modulation constellation, wherein the constellation point coordinates are real numbers, and the coordinate components on each dimension of the multi-dimensional modulation constellation correspond to different signal dimensions.
3. 3. The multidimensional constellation probability shaping method based on the dimensional energy statistics and entropy constraints of claim 2, wherein the energy statistics are expressed as: ; In the formula, Representing a multidimensional modulation constellation Energy statistics of dimensions.
4. 4. A multidimensional constellation probability shaping method based on a vinylon energy statistic and entropy constraint according to claim 3, wherein the power function form of the vinylon weights is expressed as: ; In the formula, Representing a multidimensional modulation constellation The dimension corresponds to the dimension of the dimension, The adjusting parameters are used for controlling the action degree of the current dimension energy distribution difference in the probability shaping process, and the value range is larger than 0; The exponential function form of the maintenance weight is expressed as: ; In the formula, Representing an exponential function.
5. 5. The multidimensional constellation probability shaping method based on the vinylic energy statistics and entropy constraints of claim 4, wherein said vinylic weighted energy metric is expressed as: ; In the formula, Represent the first The vinylic weighted energy metric for each constellation point.
6. 6. The multidimensional constellation probability shaping method based on the dimensional energy statistics and entropy constraints of claim 5, wherein the probability mapping function of the constellation points is expressed as: ; In the formula, Represent the first The probability map function values corresponding to the individual constellation points, Represents a probability shaping intensity parameter for adjusting the concentration of the probability distribution.
7. 7. The multidimensional constellation probability shaping method based on the dimensional energy statistics and entropy constraints of claim 6, wherein said initial constellation point probability distribution is represented as: ; In the formula, Represent the first Initial constellation point probabilities for the individual constellation points, Represent the first Probability mapping function values corresponding to the constellation points.
8. 8. The multidimensional constellation probability shaping method based on the dimensional energy statistics and entropy constraints of claim 7, wherein said information entropy is expressed as ; In the formula, The entropy of the information is represented and, Represents a logarithmic function based on a natural number 2, A minute constant for avoiding numerical computation abnormality is represented.
9. 9. The multidimensional constellation probability shaping method based on the dimensional energy statistics and the entropy constraints according to claim 8, wherein the shaping strength parameter satisfying the entropy constraints is expressed as: ; In the formula, Represents shaping intensity parameters that satisfy the entropy constraint, The lower and upper boundaries of the search interval are respectively represented by the probability shaping intensity parameters.
10. 10. The multidimensional constellation probability shaping method based on the dimensional energy statistics and entropy constraints of claim 9, wherein calculating the constellation point probability distribution based on shaping strength parameters satisfying the entropy constraints comprises: Reconstructing a probability mapping function of constellation points based on the vector weighting energy measure and the shaping intensity parameter meeting entropy constraint, and remapping the vector weighting energy measure into probability weights of the constellation points; And normalizing the probability of the constellation points based on the probability weights of the constellation points to obtain the probability distribution of the constellation points.

Description

Multidimensional constellation probability shaping method based on dimensional energy statistics and entropy constraint Technical Field The invention relates to a multidimensional constellation probability shaping method based on dimensional energy statistics and entropy constraint, and belongs to the technical field of communication. Background As optical and wireless communication systems evolve towards higher transmission rates and higher spectral efficiency, higher order modulation techniques play a key role in improving system throughput. However, as the modulation order increases, the minimum euclidean distance between constellation points decreases gradually, and the average transmit power increases accordingly, resulting in a system with limited noise immunity and power efficiency. To alleviate the above problems, multidimensional modulation and probability shaping techniques are introduced into communication system designs. The probability shaping is to allocate non-uniform transmission probabilities to different constellation points, so that constellation points with lower energy appear with higher probability, and average transmission power is reduced on the premise of meeting the target information transmission rate, and controllable compromise between information entropy and power efficiency is realized. The technology has been extended to multidimensional modulation scenarios, becoming an important research direction for current efficient communication system optimization. Although probability shaping is widely applied in multidimensional modulation, the existing scheme is generally based on the total energy of constellation points only when a probability model is constructed, and cannot consider the energy distribution difference of different dimensions in a multidimensional constellation. In practical high-dimensional constellation structures, especially in constellations that are irregular or made up of multiple subspaces superimposed, each dimension often has an unbalanced energy characteristic and affects system performance to a different extent. The prior method implicitly assumes that the contribution of each dimension to the performance is approximately equivalent, resulting in significant limitations on the effect of probability shaping in scenes where the energy distribution between dimensions is unbalanced. The method has the advantages that under the condition of the same target information entropy, the average transmitting power is limited in reducing amplitude, the shaping parameters are sensitive to constellation structure change and insufficient in adaptability, and stable and consistent shaping performance is difficult to maintain among different multidimensional constellations, so that the complexity of system debugging and engineering realization is increased. Disclosure of Invention The invention aims to provide a multidimensional constellation probability shaping method based on dimensional energy statistics and entropy constraint, which realizes reasonable configuration of multidimensional constellation emission probability by adaptively regulating and controlling constellation probability distribution under the constraint condition of target information entropy so as to solve the problem of limited probability shaping effect caused by unbalanced energy distribution of different dimensions in the existing multidimensional modulation constellation. In order to solve the technical problems, the invention is realized by adopting the following technical scheme. The invention provides a multidimensional constellation probability shaping method based on dimensional energy statistics and entropy constraint, which comprises the following steps: Acquiring coordinates of all constellation points in a multidimensional modulation constellation; Based on the coordinates of all constellation points, carrying out statistical analysis on the energy distribution characteristics of each dimension of the multidimensional modulation constellation to obtain energy statistics of each dimension; Constructing a dimension weight reflecting the energy distribution difference of different dimensions through a monotonic mapping function according to the energy statistics of each dimension; based on the dimensional weights and the coordinate components of the constellation points in each dimension of the multidimensional modulation constellation, constructing a dimensional weighted energy measurement of the constellation points by utilizing linear weighting; constructing a probability mapping function of constellation points based on the vector weighting energy measurement and the probability shaping strength parameter, and mapping the vector weighting energy measurement into probability weights of the constellation points; Normalizing the probability of the constellation points based on the probability weights of the constellation points to obtain initial constellation point probability distribution