CN111162886A - Pilot pattern distribution optimization method in digital amplitude modulation broadcast channel estimation - Google Patents

Abstract

A pilot pattern allocation optimization method in digital amplitude modulation broadcast channel estimation. The assignment of pilot patterns is modeled as a combinatorial optimization problem by minimizing the overall correlation of the pilot patterns to the perceptual matrix, and the problem is solved using an improved quantum genetic algorithm. And encoding the individuals by using quantum bits to obtain an initial pilot frequency pattern set, calculating the fitness value of each individual in the initial pilot frequency pattern set by taking the overall correlation as a fitness function, and finally updating the population individuals through a quantum revolving gate. Repeating the above steps for multiple times, and selecting the pilot pattern with the maximum fitness value as the optimal deterministic pilot pattern. The invention solves the problems of small search space and low efficiency of the pilot frequency pattern in the search process by utilizing the improved quantum genetic algorithm, and avoids the defect of trapping in local optimum in the search process. The optimized deterministic pilot pattern can obtain accurate channel information with less pilot overhead, and the frequency spectrum utilization rate is effectively improved.

Description

Pilot pattern distribution optimization method in digital amplitude modulation broadcast channel estimation

Technical Field

The invention relates to the technical field of digital amplitude modulation broadcasting, in particular to a pilot frequency pattern distribution optimization method in digital amplitude modulation broadcasting channel estimation.

Technical Field

Drm (digital Radio monitor) is a digital amplitude modulation broadcasting technique for frequencies below 30 MHz. Because the channel in the transmission process is a multipath fading channel with limited bandwidth, the DRM adopts the orthogonal frequency division multiplexing technology to resist multipath interference, and the spectrum utilization efficiency is effectively improved. In order to obtain better system performance, both coherent demodulation and channel equalization at the receiving end need to obtain accurate channel state information. The conventional DRM channel estimation method detects channel-selective fading using equally spaced pilot patterns. The method uses more pilot signals, so that effective transmission data is reduced, the frequency spectrum utilization rate is low, and the development of digital amplitude modulation broadcast multimedia data service is greatly limited.

In recent years, compressed sensing-based sparse channel estimation formed using wireless channel sparsity has received extensive attention and research. Theoretical studies show that sparse channel estimation based on compressed sensing obtains good estimation performance using random pilot patterns. However, for DRM channel estimation, it is not practical in practical systems to randomly select the pilot pattern each time. Therefore, a more effective pilot pattern allocation optimization method needs to be researched to provide a deterministic pilot pattern for a DRM communication system, so that the spectrum utilization efficiency is improved while the estimation performance is ensured, which has an important practical value for developing multimedia data service application for digital amplitude modulation broadcasting.

In the prior pilot pattern optimization method based on compressed sensing, the minimum cross-correlation of the sensing matrix is taken as a judgment criterion, and a deterministic pilot pattern is obtained through an optimization algorithm. However, the cross-correlation minimization criterion reflects the most extreme case of two-column correlation of the sensing matrix and cannot describe the average recovery capability of the sensing matrix. In addition, the search algorithm adopted in the optimization process is easy to cause precocity, so that the optimal solution falls into local optimization.

Disclosure of Invention

In order to solve the above technical problems, the present invention provides a method for optimizing the allocation of pilot patterns in digital amplitude modulation broadcast channel estimation, which solves the problems of small search space and low efficiency of pilot patterns in the optimization process, and avoids the defect of partial optimization in the optimization process. The optimized deterministic pilot pattern can obtain accurate channel information with less pilot overhead, and the frequency spectrum utilization rate is effectively improved.

The technical scheme for solving the technical problems is as follows:

a pilot pattern distribution optimization method in digital amplitude modulation broadcast channel estimation specifically comprises the following steps:

step (A), the minimum overall correlation of the corresponding perception matrix of the pilot frequency pattern is used as a judgment criterion in the optimization process, and the optimal distribution of the pilot frequency pattern is converted into the minimum value solution of the combined optimization problem;

step (B), setting parameters of an improved quantum genetic algorithm and initializing a population; the number of individuals of the initialization population is n, namely the number of pilot frequency patterns; each individual adopts quantum bit coding, and the bit width is m; the population evolution algebra is t, and the maximum evolution algebra is Gs; initializing a population

When t is 0, the jth individual

The qubit coding of (a) is shown in equation (1):

in the formula (1), the reaction mixture is,

is a probability amplitude constant;

step (C), observing the state of Q (0) to obtain a binary solution P (0); randomly generating a number r between 0 and 1, if

Ith position of

Then the bit takes 1, otherwise 0; subsequently, storing the converted Q (0) in P (t) & gtY_t＝0In which

A binary number string of m bits;

step (D), calculating the fitness value of each individual in P (0), and setting an optimal population B (t), the initialized optimal population

Same as P (0);

step (E), updating an evolution algebra t as t + 1;

observing Q (t-1), measuring to obtain P (t), and calculating the fitness value of P (t); comparing B (t-1) with P (t), selecting excellent individuals and storing in B (t);

and (G) updating the population Q (t) to Q (t +1) by adopting a quantum rotary gate, wherein the quantum bit of each individual is updated as follows:

wherein U theta_iBeing quantum revolving doors, theta_iIs the angle of rotation, U theta, of the revolving door_iCan be expressed as:

step (H), catastrophe operation; checking whether the individuals in the population converge and do not reach the preset precision, if so, entering the step (I), otherwise, executing catastrophe operation;

and (I) checking whether the iteration condition is met, if so, ending the searching process, and outputting the optimal solution b in the step (B) and the step (t) as the optimal pilot frequency pattern, otherwise, entering the step (E).

Preferably, in the step (a):

assuming that the total number of sub-carriers transmitting each OFDM symbol in the DRM system is M, wherein the number of sub-carriers used by the pilot signal is C C < M; in a DRM system, the pilot pattern p ═ p₁,p₂,…,p_C(1≤p₁≤p₂≤…≤p_CM) is determined by the position number of the pilot frequency subcarrier; the received and transmitted pilot signals may be represented as:

r＝SWh+z (4)

wherein, S ═ diag Sp₁,Sp₂,…,Sp_CIs a transmitted pilot signal, and a receiver receives the pilot signal r ═ rp₁,rp₂,…,rp_C ^TWhere T represents transpose. z1, z2, …, zC is white gaussian noise in the frequency domain, and the channel impulse response h (1), h (2), …, h (m)^TIs a sparse vector. W is the partial fourier matrix that constitutes C × M after decimation according to the pilot pattern p:

wherein ω ═ e^-j2π/MAnd j is an imaginary unit.

Further defining D-SW as a perception matrix corresponding to the pilot frequency pattern, carrying out column normalization on D, and carrying out integral correlation mu of the matrix D_wD is defined as:

μ_wD＝||G-I||₂(6)

wherein G ═ D^HD is called Gram matrix of matrix D, H is conjugate transpose of matrix, I is unit matrix; the assignment optimization of the pilot patterns is modeled as solving a combinatorial optimization problem according to the overall correlation minimization criterion, as shown in the following equation:

where Δ is the set of all pilot patterns.

Preferably in step (D):

global correlation mu with pilot pattern to perceptual matrix D_wAnd D is used as an objective function, and the reciprocal of the objective function is mapped into a population fitness function to be used for calculating the fitness value of the individual.

Preferably, the step (H) performs a disaster operation by the following specific process:

step (H1): calculating the average fitness value f of the population individuals in the iteration_avgThe calculation is shown as follows:

wherein n is the total number of individuals in the population, f_iThe fitness value of the ith individual in the population;

step (H2): calculating the population individual fitness precocity degree f in the iteration_dIs calculated as follows:

step (H3): if not, f_dThe smaller the size, the more premature the population tends to be, i.e., the diversity of the population is destroyed, at which time the catastrophe operation is performed. In the iteration, the excellent individuals with the largest population fitness value are reserved, and the rest individuals are randomly generated again.

Compared with the prior art, the invention has the following advantages:

1. the new pilot pattern optimization criterion is adopted to describe the correlation among the column vectors of the sensing matrix, namely the overall correlation is used as the judgment basis of pilot pattern searching, the correlation of the column vectors of the sensing matrix is reflected on the whole, and the optimization guarantee is provided for the pilot pattern with better channel estimation design effect.

2. The pilot frequency pattern adopts an improved quantum genetic algorithm as a searching method in the searching process, and the understanding space range is expanded by utilizing the parallelism of quantum bit coding. In addition, the improved quantum genetic algorithm adopts the catastrophe operation to prevent the algorithm from falling into precocity, thereby effectively avoiding the defect that the pilot frequency pattern falls into local optimum in the searching process.

Drawings

Fig. 1 is a flowchart of a method for optimizing the allocation of pilot patterns in DRM channel estimation.

Fig. 2 is a diagram illustrating a comparison of mean square errors of pilot patterns obtained by different methods in DRM channel estimation.

Fig. 3 is a schematic diagram showing the comparison of bit error rates of pilot patterns obtained by different methods in DRM channel estimation.

Detailed Description

The invention is further explained below with reference to the drawings and the embodiments. The invention discloses a pilot pattern distribution optimization method in digital amplitude modulation broadcasting (DRM) channel estimation, which converts the optimization of the pilot pattern in the DRM channel estimation into the solution of the combined optimization problem by using the minimum integral correlation of a sensing matrix in compressed sensing, and solves the optimization problem by using an improved quantum genetic algorithm to obtain an optimal deterministic pilot pattern;

the specific optimization method comprises the following steps:

step (A), the minimum overall correlation of the corresponding perception matrix of the pilot frequency pattern is used as a judgment criterion in the optimization process, and the optimal distribution of the pilot frequency pattern is converted into the minimum value solution of the combined optimization problem;

step (B), setting parameters of an improved quantum genetic algorithm and initializing a population; the number of individuals of the initialization population is n, namely the number of pilot frequency patterns; each individual adopts quantum bit coding, and the bit width is m; the population evolution algebra is t, and the maximum evolution algebra is Gs; initializing a population

When t is 0, the jth individual

The qubit coding of (a) is shown in equation (1):

in the formula (1), the reaction mixture is,

is a probability amplitude constant;

step (C), observing the state of Q (0) to obtain a binary solution P (0); randomly generating a number r between 0 and 1, if

Ith position of

Then the bit takes 1, otherwise 0; subsequently, storing the converted Q (0) in P (t) & gtY_t＝0In which

A binary number string of m bits;

step (D), calculating the fitness value of each individual in P (0), and setting an optimal population B (t), the initialized optimal population

Same as P (0);

step (E), updating an evolution algebra t as t + 1;

observing Q (t-1), measuring to obtain P (t), and calculating the fitness value of P (t); comparing B (t1) with P (t), selecting excellent individuals and storing in B (t);

and (G) updating the population Q (t) to Q (t +1) by adopting a quantum rotary gate, wherein the quantum bit of each individual is updated as follows:

wherein U theta_iBeing quantum revolving doors, theta_iIs the angle of rotation, U theta, of the revolving door_iCan be expressed as:

step (H), catastrophe operation; checking whether the individuals in the population converge and do not reach the preset precision, if so, entering the step (I), otherwise, executing catastrophe operation;

and (I) checking whether the iteration condition is met, if so, ending the searching process, and outputting the optimal solution b in the step (B) and the step (t) as the optimal pilot frequency pattern, otherwise, entering the step (E).

In the step (A), the concrete steps are as follows:

the total number of sub-carriers for transmitting each OFDM symbol in the DRM system is assumed to be M, wherein the number of the sub-carriers used by the pilot signal is CC < M; in a DRM system, the pilot pattern p ═ p₁,p₂,…,p_C(1≤p₁≤p₂≤…≤p_CM) is determined by the position number of the pilot frequency subcarrier; the received and transmitted pilot signals may be represented as:

r＝SWh+z (4)

wherein, S ═ diag Sp₁,Sp₂,…,Sp_CIs a transmitted pilot signal, and a receiver receives the pilot signal r ═ rp₁,rp₂,…,rp_C ^TWhere T represents transpose. z1, z2, …, zC is white gaussian noise in the frequency domain, and the channel impulse response h (1), h (2), …, h (m)^TIs a sparse vector. W is the partial fourier matrix that constitutes C × M after decimation according to the pilot pattern p:

wherein ω ═ e^-j2π/MAnd j is an imaginary unit.

Further defining D-SW as a perception matrix corresponding to the pilot frequency pattern, carrying out column normalization on D, and carrying out integral correlation mu of the matrix D_wD is defined as:

μ_wD＝||G-I||₂(6)

wherein G ═ D^HD is called Gram matrix of matrix D, H is conjugate transpose of matrix, I is unit matrix; the assignment optimization of the pilot patterns is modeled as solving a combinatorial optimization problem according to the overall correlation minimization criterion, as shown in the following equation:

where Δ is the set of all pilot patterns.

In the step (D), the concrete steps are as follows:

global correlation mu with pilot pattern to perceptual matrix D_wAnd D is used as an objective function, and the reciprocal of the objective function is mapped into a population fitness function to be used for calculating the fitness value of the individual.

In the step (H), the concrete steps are as follows:

the specific process of performing the catastrophic operation is as follows, step (H1): calculating the average fitness value f of the population individuals in the iteration_avgThe calculation is shown as follows:

wherein n is the total number of individuals in the population, f_iThe fitness value of the ith individual in the population;

step (H2): calculating the population individual fitness precocity degree f in the iteration_dIs calculated as follows:

step (H3): if not, f_dThe smaller the size, the more premature the population tends to be, i.e., the diversity of the population is destroyed, at which time the catastrophe operation is performed. In the iteration, the excellent individuals with the largest population fitness value are reserved, and the rest individuals are randomly generated again.

As mentioned above, a detailed description of the preferred embodiments of the invention has been given to enable those skilled in the art to make and practice the invention. Although the present invention has been described with reference to exemplary embodiments, those skilled in the art will appreciate that various modifications and changes can be made in the present invention without departing from the spirit or scope of the invention described in the appended claims. Thus, the present invention is not intended to be limited to the particular embodiments shown and described herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims (4)

1. A method for optimizing the distribution of pilot patterns in digital amplitude modulation (DRM) channel estimation is characterized in that: modeling the pilot pattern optimization in DRM channel estimation by using the minimum correlation of the whole perception matrix as a combined optimization problem, solving the problem by using an improved quantum genetic algorithm to further obtain an optimal deterministic pilot pattern, wherein the specific optimization method comprises the following steps of:

step (A), the minimum overall correlation of the corresponding perception matrix of the pilot frequency pattern is used as a judgment criterion in the optimization process, and the optimal distribution of the pilot frequency pattern is converted into the minimum value solution of the combined optimization problem;

step (B), setting parameters of an improved quantum genetic algorithm and initializing a population; the number of individuals of the initialization population is n, namely the number of pilot frequency patterns; each individual adopts quantum bit coding, and the bit width is m; the population evolution algebra is t, and the maximum evolution algebra is Gs; initializing a population

When t is 0, the jth individual

The qubit coding of (a) is shown in equation (1):

in the formula (1), the reaction mixture is,

is a probability amplitude constant;

step (C), observing the state of Q (0) to obtain a binary solution P (0); randomly generating a number r between 0 and 1, if

Ith position of

Then the bit takes 1, otherwise 0; subsequently, storing the converted Q (0) in P (t) & gtY_t＝0In which

A binary number string of m bits;

step (D), calculating the fitness value of each individual in P (0), and setting an optimal population B (t), the initialized optimal population

Same as P (0);

step (E), updating an evolution algebra t as t + 1;

observing Q (t-1), measuring to obtain P (t), and calculating the fitness value of P (t); comparing B (t-1) with P (t), selecting excellent individuals and storing in B (t);

and (G) updating the population Q (t) to Q (t +1) by adopting a quantum rotary gate, wherein the quantum bit of each individual is updated as follows:

wherein U theta_iBeing quantum revolving doors, theta_iIs the angle of rotation, U theta, of the revolving door_iCan be expressed as:

step (H), catastrophe operation; checking whether the individuals in the population converge and do not reach the preset precision, if so, entering the step (I), otherwise, executing catastrophe operation;

and (I) checking whether the iteration condition is met, if so, ending the searching process, and outputting the optimal solution b in the step (B) and the step (t) as the optimal pilot frequency pattern, otherwise, entering the step (E).

2. The method of claim 1 wherein the step (a) comprises:

the total number of sub-carriers for transmitting each OFDM symbol in the DRM system is assumed to be M, wherein the number of the sub-carriers used by the pilot signal is CC < M; in a DRM system, the pilot pattern p ═ p₁，p₂，…，p_C(1≤p₁≤p₂≤…≤p_CM) is determined by the position number of the pilot frequency subcarrier; the received and transmitted pilot signals may be represented as:

r＝SWh+z (4)

wherein, S ═ diag Sp₁，Sp₂，…，Sp_CIs a transmitted pilot signal, and a receiver receives the pilot signal r ═ rp₁，rp₂，…，rp_C ^TWhere T represents transpose. z1, z2, …, zC is white gaussian noise in the frequency domain, and the channel impulse response h (1), h (2), …, h (m)^TIs a sparse vector. W is the partial fourier matrix that constitutes C × M after decimation according to the pilot pattern p:

wherein w ═ e^-j2π/MAnd j is an imaginary unit.

Further defining D-SW as a perception matrix corresponding to the pilot frequency pattern, carrying out column normalization on D, and carrying out integral correlation mu of the matrix D_wD is defined as:

μ_wD＝||G-I||₂(6)

wherein G ═ D^HD is called Gram matrix of matrix D, H is conjugate transpose of matrix, I is unit matrix; the assignment optimization of the pilot patterns is modeled as solving a combinatorial optimization problem according to the overall correlation minimization criterion, as shown in the following equation:

where Δ is the set of all pilot patterns.

3. The method of claim 1 wherein the step (D) comprises:

global correlation mu with pilot pattern to perceptual matrix D_wAnd D is used as an objective function, and the reciprocal of the objective function is mapped into a population fitness function to be used for calculating the fitness value of the individual.

4. The method of claim 1, wherein the step (H) performs the catastrophe operation by the following steps:

step (H1): calculating the average fitness value f of the population individuals in the iteration_avgThe calculation is shown as follows:

wherein n is the total number of individuals in the population, f_iThe fitness value of the ith individual in the population;

step (H2): calculating the population individual fitness precocity degree f in the iteration_dIs calculated as follows:

step (H3): if not, f_dThe smaller the size, the more premature the population tends to be, i.e., the diversity of the population is destroyed, at which time the catastrophe operation is performed. In the iteration, the excellent individuals with the largest population fitness value are reserved, and the rest individuals are randomly generated again.

Images