A frequency reuse planning method for low-orbit satellite systems based on genetic algorithm

Through the frequency multiplexing planning method based on genetic algorithm, the frequency allocation of wave bits in low-orbit satellite systems is optimized, the problem of homofrequency interference is solved, the distance between homofrequency wave bits is maximized, and the spectrum utilization of the system is improved.

CN115941023BActive Publication Date: 2024-11-19THE 54TH RESEARCH INSTITUTE OF CHINA ELECTRONICS TECHNOLOGY GROUP CORPORATION
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202211354499.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-11-01
Publication Date
2024-11-19
Estimated Expiration
2042-11-01

AI Technical Summary

Technical Problem

The problem of syn-frequency interference caused by frequency multiplexing in low-orbit satellite systems, especially the interference is serious when the distance between the same frequency waves is too close, and the prior art is difficult to effectively suppress.

Method used

Using a frequency multiplexing planning method based on genetic algorithm, the frequency multiplexing scheme is optimized to maximize the minimum distance between the same frequency wave bits and reduce the same frequency interference by dividing wave bits. The specific steps include dividing the outer and inner wave positions, setting optimization parameters, initializing the population, performing cross-mutation operations and selection operations, and iteratively calculating the optimal frequency multiplexing scheme.

Benefits of technology

It realizes the optimal multicolor multiplexed wave position arrangement scheme in a given coverage area, significantly increasing the minimum distance between the same frequency wave positions, effectively suppressing the same frequency interference, and improving the spectrum utilization of the system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115941023B_ABST
    Figure CN115941023B_ABST
Patent Text Reader

Abstract

The present invention belongs to the field of satellite communication technology, and discloses a frequency reuse planning method for a low-orbit satellite system based on a genetic algorithm, comprising the following steps: dividing the wave position of the coverage area; setting the frequency allocation scheme of the inner wave position; performing genetic algorithm iteration on the frequency allocation of the outer wave position, including setting optimization parameters, initializing the population, performing cross-mutation operations, performing selection operations, and updating the global optimum. The frequency reuse planning method for a low-orbit satellite system based on a genetic algorithm provided by the present invention can iteratively calculate the optimal frequency reuse scheme under a given coverage area wave position division scheme, and increase the minimum distance between the same-frequency wave positions as much as possible to reduce the same-frequency interference. It is used in the field of satellite communication technology.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of satellite communications, and in particular relates to a frequency reuse planning method for a low-orbit satellite system based on a genetic algorithm. Background Art

[0002] Low-orbit satellite systems have the advantages of low transmission delay, high spectrum utilization, large transmission capacity, good coverage continuity, and low terminal cost. They are considered to be an effective technical approach to building a new generation of broadband satellite communication systems. The ground coverage area of ​​low-orbit satellites is often divided into multiple wave positions. Through beam hopping, frequency reuse, and other methods, the capacity of satellite communication systems can be greatly improved. However, frequency reuse will bring about the problem of co-channel interference. If the distance between two co-frequency wave positions is too close, the co-channel interference problem will become serious. In the system design stage, with the help of optimization algorithms, the frequency reuse times and wave position arrangement schemes are reasonably designed to maximize the minimum distance between co-frequency wave positions to suppress the co-channel interference problem.

[0003] In recent years, swarm intelligence algorithms have received widespread attention from the academic community and have become an effective tool for solving optimization problems. Swarm intelligence algorithms belong to the category of computational intelligence and can be regarded as the application of bionics in the field of optimization. Swarm intelligence algorithms can be classified into two categories:

[0004] The first type is based on Darwin's idea of ​​natural selection and survival of the fittest, imitating the process of biological population evolution, and allowing the entire population to evolve to a higher level by retaining good individuals and eliminating inferior individuals. This type of algorithm is also called evolutionary algorithm, and typical examples include genetic algorithm, genetic programming, evolutionary strategy, evolutionary programming and differential evolution algorithm, etc.

[0005] The second type is to complete the optimization search process by imitating the individual behavior of social animals and the division of labor and cooperation between individuals. In nature, many social animals can show very complex and intelligent group behaviors through the division of labor and cooperation between individuals, while from the perspective of individuals, individual behaviors are relatively simple. Regarding the above phenomenon, typical examples include bird flocks foraging, bees collecting honey, and ants finding their way. Scholars have conducted extensive research on the behavior of these organisms and proposed particle swarm optimization algorithms, artificial bee colony algorithms, and artificial ant colony algorithms.

[0006] In view of the characteristics of the frequency multi-color multiplexing wave position arrangement problem, a genetic algorithm can be used. The genetic algorithm was first proposed by Professor Holland of the University of Michigan in the early 1970s, and then DeJong extended it to the application of optimizing numerical function calculation. The genetic algorithm follows the principle of natural selection of survival of the fittest, adopts the genetic mechanism of crossover and mutation, simulates the evolution process of organisms, and forms an efficient parallel global search algorithm. During the search process, it can automatically obtain and accumulate knowledge about the search space, and adaptively control the search process to obtain the optimal solution. In recent years, genetic algorithms have been widely used in problem solving, optimization and search, machine learning, system control, pattern recognition and other fields, and have achieved great success. Genetic algorithms have universal adaptability, are not limited by the nature and specific form of the problem, and have good compatibility with the solution of the frequency multi-color multiplexing wave position arrangement problem. Summary of the invention

[0007] In response to the co-channel interference problem caused by multi-wavelength frequency reuse of low-orbit satellites, the present invention proposes a frequency reuse planning method for low-orbit satellite systems based on genetic algorithm, which iteratively calculates the optimal frequency reuse scheme and increases the minimum distance between co-frequency waves as much as possible to reduce co-channel interference.

[0008] A frequency reuse planning method for a low-orbit satellite system based on a genetic algorithm comprises the following steps:

[0009] Step 1: Divide the coverage area of ​​the low-orbit satellite system into several layers of wave positions, of which the outermost layer of wave positions is called the outer layer of wave positions, and the remaining layers of wave positions are collectively called the inner layer of wave positions;

[0010] Step 2: Set the frequency allocation scheme for the inner wave positions according to the principle that the same frequency wave positions are the farthest apart;

[0011] Step 3: Perform genetic algorithm iteration on the outer wave position frequency allocation, including setting optimization parameters, initializing the population, performing crossover-mutation operations, performing selection operations, and updating the global optimum.

[0012] Furthermore, the step 3 comprises:

[0013] Step 3.1: Set the optimization parameters, including population size N, maximum number of iterations G, and number of mutant individuals P;

[0014] Step 3.2: Initialize the population. Each individual in the population represents a scheme for the outer wave position arrangement. The gene of each individual is a full arrangement of the following set A:

[0015]

[0016] Among them, Q kis the number of times the kth frequency is reused in the outer beam, J is the number of outer wave positions, which is also the number of genes of each individual. K is the number of multiplexed frequencies. The j-th digit k indicates that the j-th wave position fills the k-th frequency, j = 1, 2, ..., J, k = 1, 2, ..., K;

[0017] Generate multiple random permutations of set A as individual genes, and select N fitness values ​​less than the threshold N t The individuals are used as the initial population, and the individual with the lowest fitness value in the initial population is set as the global optimum;

[0018] Step 3.3: Perform crossover-mutation operation. For the nth individual in the initial population, n = 1, 2, ..., N, denoted as individual a, randomly select an individual from the other N-1 individuals, denoted as individual b, retain the same part of the genes of individual a and individual b, denoted as set Same; the part of the genes of individual a that is different from that of individual b is denoted as set Diff0, randomly select two genes in set Diff0 and swap their order to obtain set Diff1, and combine set Same and set Diff1 to obtain mutant individuals; if the genes of individual a and individual b are the same, randomly select two genes and swap their order to generate mutant individuals; calculate the fitness value of the mutant individuals; repeat this step until P mutant individuals with fitness values ​​no greater than that of the original individual a are generated;

[0019] Step 3.4: Perform a selection operation. Select the individual with the smallest fitness value from the P mutant individuals obtained in step 3.3 to replace the original individual a. If the fitness values ​​are the same, randomly select an individual to replace the original individual a.

[0020] Step 3.5: Repeat steps 3.3-3.4 until all N individuals have completed the operation, and set the individual with the lowest fitness value as the new global optimum;

[0021] Step 3.6: Repeat steps 3.3 to 3.5 until the number of iterations reaches the maximum, and output the global optimum as the result.

[0022] Furthermore, the fitness value is calculated as follows:

[0023] An individual is represented by:

[0024] A′=[a1,a2,...,a J ]

[0025] In the formula, a j ∈{1,2,...,K}, and A′ is a full permutation of set A;

[0026] The fitness value is expressed as:

[0027]

[0028] Wherein, the method for determining the size of W is as follows: the initial value of W is set to 0, and W increases by 1000 for each pair of adjacent or one-position-separated same-frequency wave positions on the outer circle wave positions; W increases by 1 for each pair of two-position-separated same-frequency wave positions on the outer circle wave positions;

[0029] Indicates that the jth wave position fills the ath j The fitness value contributed by the jth frequency is determined according to the frequency allocation scheme of the inner wave position. j After the frequency, the inner wave position is a j The frequencies are adjacent, If the jth wave position fills the ath j After the frequency, the inner wave position is a j The frequency interval is 1 wave position, then If the jth wave position fills the ath j After the frequency, the inner wave position is a j The frequency interval is 2 wave positions, then If the jth wave position fills the ath j After the frequency, the inner wave position is a j If the frequency interval is 3 or more than 3 wave positions, then All The value of is expressed as a J×K matrix, called the outer wave position fitness value matrix.

[0030] The beneficial effects of the present invention are:

[0031] 1. The frequency reuse planning method for the low-orbit satellite system provided by the present invention can calculate the optimal multi-color multiplexing wave position arrangement scheme under a given coverage area wave position division scheme, and increase the minimum distance between the same-frequency waves as much as possible to suppress the same-frequency interference.

[0032] 2. The present invention encodes population individuals based on multi-color wave position sorting, converts the multi-color multiplexing wave position arrangement problem into a form that can be processed by genetic algorithms, thereby achieving efficient iterative calculation of the optimal solution. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is a flow chart of the frequency reuse planning method of the present invention.

[0034] Figure 2 This is a schematic diagram of the wave position division of a low-orbit satellite's coverage area on the ground.

[0035] Figure 3 This is a schematic diagram of the frequency allocation plan for the inner wave position.

[0036] Figure 4 It is a schematic diagram of the decrease of fitness value during the iteration process of the genetic algorithm of the present invention.

[0037] Figure 5 It is a schematic diagram of the iterative calculation result (frequency reuse scheme) of the genetic algorithm of the present invention. DETAILED DESCRIPTION

[0038] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0039] The present invention describes a frequency reuse planning method for a low-orbit satellite system based on a genetic algorithm. The process is as follows: Figure 1 As shown, the specific steps include:

[0040] Step 1: If Figure 2 As shown in the figure, the coverage area of ​​a low-orbit satellite is divided into 52 positions, which are divided into three layers. The first layer contains 4 positions, the second layer contains 16 positions, and the third layer contains 32 positions. The third layer of positions is called the outer layer of positions, and the first and second layers of positions are collectively called the inner layer of positions.

[0041] Step 2: Set the frequency allocation plan for the inner wave position.

[0042] The working frequency band is divided into 12 color frequencies. Taking into account the large difference in path loss between the edge of the coverage area and the sub-satellite point in the coverage range of low-orbit satellites, the transmission power required for the edge wave positions is relatively large. In order to avoid large co-channel interference from the edge wave positions to the first layer wave positions, separate 4-color frequencies are set for the 4 wave positions in the first layer, that is, the frequencies of the 4 wave positions in the first layer are different from the frequencies of the second and third layer wave positions, and the frequencies of the 4 wave positions in the first layer are also different. The second and third layers reuse the remaining 8 color frequencies, and the number of wave positions occupied by each frequency is (16+32) / 8=6. According to the principle of average distribution, each frequency occupies 2 wave positions in the second layer and 4 wave positions in the third layer. At the same time, in order to avoid large co-channel interference, the 2 wave positions occupied by each frequency in the second layer are the farthest apart on the circle, that is, at the two ends of the diameter of the circle. The frequency allocation scheme for the inner wave positions is as follows: Figure 3 As shown, the frequencies of the first 4 wave positions are different from the frequencies of other wave positions.

[0043] Step 3: Perform genetic algorithm iteration on the outer wave position frequency allocation.

[0044] The number of outer wave positions is 32, and the number of multiplexed frequencies is 8. The fitness value matrix of the outer wave positions can be expressed as:

[0045]

[0046] where f jk(j=1,2,...,32,k=1,2,...,8) represents the fitness value contributed by the jth wave position after filling the kth frequency. jk The value of depends on the second-layer frequency allocation scheme set in step 2. If the jth wave position is filled with the kth frequency and is adjacent to the kth frequency of the inner wave position, then f jk = 1000; if the jth wave position is filled with the kth frequency, and the kth frequency of the inner wave position is separated by 1 wave position, then f jk =2; if the jth wave position is filled with the kth frequency, it is separated from the kth frequency of the inner wave position by 2 wave positions, then f jk =1; if the jth wave position is filled with the kth frequency, and is separated from the kth frequency of the inner wave position by 3 or more wave positions, then f jk = 0. According to the second-layer frequency allocation scheme set in step 2, the fitness value matrix is ​​set as:

[0047]

[0048] Now compare Figure 3 The above formula is explained. Figure 3 The second layer number is the frequency number, and the third layer number is the wave position number. 11 =1000 means that wave position 1 fills frequency 1, which will make the same frequency wave positions of frequency 1 closely adjacent; f 13 =2, indicating that the wave position 1 is filled with frequency 3, which will make the wave position with the same frequency 3 be 1 wave position away; f 14 =0, indicating that wave position 1 fills frequency 4, which will make the distance between the wave positions with the same frequency of frequency 4 greater than or equal to 3 wave positions.

[0049] Genetic algorithm iteration is performed on the outer wave position frequency allocation. The iteration process is shown in Figure 1 In the dashed box, the specific steps are as follows:

[0050] Step 3.1: Set the optimization parameters, set the population size N = 50, the maximum number of iterations G = 500, and the number of mutant individuals P = 40;

[0051] Step 3.2: Initialize the population. Each individual represents a scheme for the arrangement of the outer wave position, containing 32 genes, represented by a 32-bit number, where the j-th digit k represents the j-th wave position filling the k-th frequency. It has been determined in step 2 that the number of multiplexing of each frequency in the outer beam is 4. The genes of each individual are a full permutation of the following set A:

[0052] A=[1,1,1,1,2,2,2,2,...,8,8,8,8]

[0053] An entity can be represented as:

[0054] A′=[a1,a2,...,a 32 ]

[0055] where a j ∈{1,2,...,8}. Then the fitness value is expressed as:

[0056]

[0057] The method for determining the size of W is as follows: the initial value of W is set to 0. When there is a pair of wave positions with one wave position interval or adjacent wave positions with the same frequency on the outer circle, W increases by 1000. When there is a pair of wave positions with the same frequency interval of 2 wave positions on the outer circle, W increases by 2. Indicates that the jth wave position fills the ath j The fitness value contributed by the frequency is determined according to the above fitness value matrix. Initializing the population is to continuously generate random full permutations of set A as the genes of the initial individuals and calculate the fitness value. If the fitness value is less than the threshold N t =1000, then keep the individual. Repeat this step until 50 individuals with fitness values ​​less than 1000 are generated as the initial population. The individual with the lowest fitness value in the initial population is set as the initial global optimum;

[0058] Step 3.3: Perform crossover-mutation operations. Since the problem to be optimized is a sorting problem, in order to adapt to this problem, the crossover and mutation operations in the traditional genetic algorithm are integrated into one step. The specific method is: for the nth individual (denoted as individual a, n = 1, 2, ..., 50), randomly select an individual (denoted as individual b) from the other 49 individuals, and retain the same part of the genes of individual a and individual b (denoted as set Same); for the part of the genes of individual a that is different from that of individual b (denoted as set Diff0), randomly select two genes in set Diff0 and swap their order to obtain set Diff1; mutant individuals are obtained by combining set Same and set Diff1. In particular, if the genes of individual a and individual b are the same, randomly select two genes and swap their order to generate mutant individuals; calculate the fitness value of the mutant individual; repeat this step until 40 mutant individuals with fitness values ​​no greater than the original individual are generated;

[0059] Step 3.4: Perform selection operation. Among the 40 mutant individuals obtained in step 3.3, select the individual with the smallest fitness value to replace the original individual. If their fitness values ​​are the same, then randomly select one to replace the original individual;

[0060] Step 3.5: Repeat steps 3.3-3.4 until all 50 individuals have completed the operation. At this point, one iteration is completed and the global optimum is updated.

[0061] Step 3.6: Repeat 3.3-3.5 until the number of iterations reaches 500, and output the global optimal result. During the iteration process, the fitness value of the global optimal individual decreases as follows: Figure 4 As shown, the frequency reuse scheme represented by the iteration result is as follows Figure 5 shown.

Claims

1. A frequency reuse planning method for a low-orbit satellite system based on a genetic algorithm, characterized in that: The following steps are involved: Step 1: Divide the coverage area of ​​the low-orbit satellite system into several layers of wave positions, of which the outermost layer of wave positions is called the outer layer of wave positions, and the remaining layers of wave positions are collectively called the inner layer of wave positions; Step 2: Set the frequency allocation scheme for the inner wave positions according to the principle that the same frequency wave positions are the farthest apart; Step 3: Perform genetic algorithm iteration on the outer wave position frequency allocation, including setting optimization parameters, initializing the population, performing crossover-mutation operations, performing selection operations, and updating the global optimum; Wherein, the step 3 comprises: Step 3.1: Set the optimization parameters, including population size N, maximum number of iterations G, and number of mutant individuals P; Step 3.2: Initialize the population. Each individual in the population represents a scheme for the outer wave position arrangement. The gene of each individual is a full arrangement of the following set A: Among them, Q k is the number of times the kth frequency is reused in the outer beam, J is the number of outer wave positions, which is also the number of genes of each individual. K is the number of multiplexed frequencies. The j-th digit k means that the j-th wave position fills the k-th frequency, j = 1, 2, ..., J, k = 1, 2, ..., K; Generate multiple random permutations of sets A as individual genes, and select N fitness values ​​less than the threshold N t The individuals are used as the initial population, and the individual with the lowest fitness value in the initial population is set as the global optimum; Step 3.3: Perform crossover-mutation operation. For the nth individual in the initial population, n = 1, 2, ..., N, denoted as individual a, randomly select an individual from the other N-1 individuals, denoted as individual b, retain the same part of the genes of individual a and individual b, denoted as set Same; the part of the genes of individual a that is different from that of individual b is denoted as set Diff0, randomly select two genes in set Diff0 and swap their order to obtain set Diff1, and combine set Same and set Diff1 to obtain mutant individuals; if the genes of individual a and individual b are the same, randomly select two genes and swap their order to generate mutant individuals; calculate the fitness value of the mutant individuals; repeat this step until P mutant individuals with fitness values ​​no greater than that of the original individual a are generated; Step 3.4: Perform a selection operation. Select the individual with the smallest fitness value from the P mutant individuals obtained in step 3.3 to replace the original individual a. If the fitness values ​​are the same, randomly select an individual to replace the original individual a. Step 3.5: Repeat steps 3.3-3.4 until all N individuals have completed the operation, and set the individual with the lowest fitness value as the new global optimum; Step 3.6: Repeat steps 3.3 to 3.5 until the number of iterations reaches the maximum, and output the global optimum as the result.

2. The method for frequency reuse planning of a low-orbit satellite system based on a genetic algorithm according to claim 1, characterized in that: The calculation method of the fitness value is: An individual is represented by: <h2 style=";text-align:left;direction:ltr">A′ = [a1,a2,...,a<h2 style=";text-align:left;direction:ltr"> J <h2 style=";text-align:left;direction:ltr"> ] In the formula, a j ∈{1,2,...,K}, and A′ is a full permutation of set A; The fitness value is expressed as: Wherein, the method for determining the size of W is as follows: the initial value of W is set to 0, and W increases by 1000 for each pair of adjacent or one-position-separated same-frequency wave positions on the outer circle wave positions; W increases by 1 for each pair of two-position-separated same-frequency wave positions on the outer circle wave positions; f j(aj) Indicates that the jth wave position fills the ath j The fitness value contributed by the jth frequency is determined according to the frequency allocation scheme of the inner wave position. j After the frequency, the inner wave position is a j If the frequencies are adjacent, then f j(aj) =1000; if the jth wave position fills the ath j After the frequency, the inner wave position is a j The frequency interval is 1 wave position, then f j(aj) =2; if the jth wave position fills the ath j After the frequency, the inner wave position is a j The frequency interval is 2 wave positions, then f j(aj) =1; if the jth wave position fills the ath j After the frequency, the inner wave position is a j If the frequency interval is 3 or more than 3 wave positions, then f j(aj) =0; set all f j(aj) The value of is expressed as a J×K matrix, called the outer wave position fitness value matrix.