Wireless multicast satellite network topology construction and genetic algorithm flow distribution method

Abstract

The invention relates to the technical field of satellite communication, in particular to a wireless multicast satellite network topology construction and genetic algorithm flow distribution method capable of remarkably improving the network capacity, a plurality of data streams on a satellite can be compressed through a coding process, so that the use of communication and storage resources is reduced to the greatest extent, and in addition, the network capacity is improved. In addition, a topological construction and flow distribution method based on a genetic algorithm is introduced to improve the performance of network coding, and a simulation result shows that optimization is added to the topological structure, so that the network capacity can be remarkably improved.

Description

Technical field:

[0001] The invention relates to the technical field of satellite communications, in particular to a method for constructing a wireless multicast satellite network topology and allocating genetic algorithm traffic, which can significantly improve network capacity. Background technology:

[0002] In recent years, satellite networks have attracted increasing attention due to their ability to provide greater coverage and throughput than traditional optical fiber and mobile networks, especially in remote areas such as mountains and oceans, without incurring additional deployment costs. Therefore, satellite networks have become an important component of next-generation communication networks. However, satellite networks have limited resources such as storage, power, and frequency band, while high-bandwidth transmission tasks such as streaming video, VR / AR, and large language models are basic services provided by current and future networks. Therefore, how to maximize network transmission capacity under limited communication resources is a current research hotspot in the field of satellite communications and is also the key to the practical application of satellite networks.

[0003] Traditional methods often improve network capacity by optimizing the allocation of communication resources, such as constructing a gain function to describe the characteristics of multiple services, and proposing a multi-beam joint resource allocation scheme to improve the capacity of satellite networks. Although these studies have achieved certain capacity optimization by studying the resource allocation problem, they lack consideration of the dynamic characteristics of satellite networks. In actual transmission scenarios, due to the constant changes in the environment and tasks, resource optimization strategies also need to be updated in real time. However, most of the optimization algorithms used in these studies take too long to calculate and occupy a lot of memory in larger constellations, making it difficult to maintain high real-time performance. Therefore, improving the characteristics of the network itself is a more practical research direction.

[0004] Network coding is an effective tool to improve network performance in multicast scenarios, and a large number of studies have demonstrated its ability to improve network capacity. According to current research, network topology has a significant impact on the effectiveness of network coding in improving network capacity, and it has been found that it may be beneficial to reduce the probability of failure. At the same time, optimizing the topology structure can also be an effective way to improve the performance of network coding systems. However, due to the characteristics of the satellite itself, its network topology inevitably has time-varying characteristics, and adding a time dimension to the network model will make it difficult to represent and optimize. Summary of the invention:

[0005] In view of the shortcomings and deficiencies in the prior art, the present invention proposes a wireless multicast satellite network topology construction and genetic algorithm traffic distribution method which can significantly improve network capacity.

[0006] The present invention is achieved by the following measures:

[0007] A wireless multicast satellite network topology construction and genetic algorithm traffic distribution method, the wireless multicast satellite network is composed of multiple communication satellites, each satellite establishes multiple satellite-to-satellite wireless links (ISLs) with other satellites to become a network node, each node can be used as a source node, relay node or destination node according to needs, in multicast transmission using network coding, multiple source satellites first encode information and transmit it to relay satellites connected through established links, the relay satellites then encode and forward these data packets, the destination satellite that needs to obtain data from one of the source satellites collects and decodes the data packets from the relay satellite to obtain the original information, the network in a short period of time is regarded as static, the total transmission time T is divided into N time slots, then the length of each time slot is τ=T / N, when τ is small enough, the change of ISL characteristics is negligible, wherein, is the vertex set, represents the i-th satellite node in the t-th time slot, E, where S represents the source satellite set, R is the relay satellite set, D is the destination satellite set, and E is the set of edges, which includes a communication edge between two satellites in a satellite slot and a storage edge between two satellites in a satellite slot. The capacity of the communication edge is defined as the throughput of the ISL between two satellites. The capacity is expressed as,

[0008]

[0009] Where B is the ISL bandwidth, is the transmission power of the ith satellite in the tth time slot, is the average noise power,

[0010] is the ISL gain between the i-th satellite and the j-th satellite, which can be calculated as:

[0011] n=kBΓ (3),

[0012] Where λ is the wavelength, G tr and G re are the transmit gain and receive gain respectively, is the distance between the i-th satellite and the j-th satellite, k is the Boltzmann constant, Γ is the noise temperature, and is expressed as The storage edge, whose capacity is the storage capacity STO of the i-th satellite i；

[0013] It is characterized in that the established time extension graph G is used to describe the topological structure of the satellite network, and the vertex s of the time extension G of the multicast transmission problem from the source satellite to the target satellite flow distribution problem is (1) The vertex set D between the source and the destination, the network topology is G, and a set of binary To describe whether should be established, considering that each satellite has a limited number of antennas, with the constraints:

[0014]

[0015] To indicate the source satellite via relay satellite and transmission, and eventually to the target satellite Collected information flow, D s ∈D is the receiver from The set of target satellites for data, defined The traffic balance constraint of each relay satellite is expressed as:

[0016]

[0017] In addition, all information is sent by the source satellite and collected by the target satellite within a given time, so there are flow constraints on the source satellite and the target satellite:

[0018]

[0019] According to the linear network coding theory, During the transmission process, the amount of data transmitted is equal to the maximum flow of the ISL to different target satellites and cannot exceed the capacity of the ISL. Therefore,

[0020]

[0021] Similarly, for the data that has been stored by satellite i but not transmitted Their number cannot exceed the storage capacity, so there is

[0022]

[0023] After finding the constraints, network coding is used to construct the optimization problem of the multicast satellite network. The optimization problem is to optimize the scheme of establishing ISLs. Traffic distribution Require Encoding capacity In line with linear network coding theory, specifically, the optimization problem is:

[0024]

[0025] The optimization problem is solved by an ISL construction method based on genetic algorithm GA.

[0026] The ISL construction method based on the genetic algorithm GA of the present invention to solve the optimization problem includes: individuals with different chromosomes are randomly generated to form an initial population, and the transmission performance of the network topology represented by each individual is evaluated as its fitness function to determine which individuals can be replicated, and the chromosomes of these individuals are crossed to simulate the sexual reproduction in nature, so as to find other better solutions under the premise of retaining some advantages of the current network topology structure. At the same time, the population is periodically randomly updated to simulate gene mutations, and searches in unknown areas of the solution space are encouraged to prevent the algorithm from falling into a local optimal solution. The above process is repeated iteratively. After the population characteristics tend to be stable or the maximum number of iterations M is reached, the individual with the highest fitness is selected, and the ISL establishment set A represented by it is the solution obtained by this algorithm, which specifically includes the following steps:

[0027] Step 1: Create an initial population, which is a set of randomly selected valid candidate solutions (individuals), each of which has a different randomly generated chromosome; build a complete topology to get the optimization variable A, and as long as two satellites are visible, build ISL and add ISL as an element to the set A, and encode A to chromosome g i If there are P individuals, there is a population G = {g i |i∈P};

[0028] Step 2: Calculate the fitness The fitness function value of each individual is calculated. This operation is performed once for the initial population and then repeated for each generation after applying the genetic operators of selection, crossover and mutation. Since the fitness of each individual is independent of other individuals, it can be calculated in parallel. In order to solve this problem, the optimization problem is used as the fitness function. The network topology represented by each individual chromosome is used to obtain the maximum multicast capacity that can be achieved by network coding and determine the value of each individual. Since the variable A is given by the chromosome, the fitness function is calculated as follows:

[0029]

[0030] Step 3: Selection, crossover and mutation Apply the genetic operators of selection, crossover and mutation to a population to produce a new generation based on the better individuals in the current generation. The purpose of selection is to select individuals that have advantages in the current population. The probability of an individual chromosome being selected for reproduction is related to its fitness f(g i ), the set p i is the probability that the chromosome of the i-th individual is selected,

[0031] Crossover is responsible for producing offspring from selected individuals. This process is usually accomplished by two selected individuals exchanging parts of their chromatids to produce two new chromosomes representing the offspring, with the intersection being k and the parental chromosomes being g. i and gj, then the offspring g new It can be expressed as:

[0032] g new =(g i1 ,g i2 ,…g ik ,g j(k+1) ,…,g jn ), the mutation operation randomly changes one or more chromosome values ​​(genes) of each newly created individual. The probability of mutation is usually very low. Uniform mutation is used. For chromosome g nk , the mutation operation is expressed as:

[0033] g' nk =g nk +δ, with probabilityμ.

[0034] In step 2 of the present invention, the following operations are performed to solve the maximum flow problem:

[0035] Step 2-1: First, solve the flow distribution problem, i.e., the maximum flow problem, for each source node, and accumulate the solved flows to find the overflow edge;

[0036] Step 2-2: Create a subgraph based on the maximum flow of different source nodes, and reduce the capacity of the corresponding edge in the subgraph according to the exceeded capacity;

[0037] Step 2-3: Solve the maximum flow problem on each of the new subgraphs.

[0038] Through the encoding process, the present invention can compress multiple data streams on the satellite, thereby minimizing the use of communication and storage resources. In addition, a topology construction and traffic distribution method based on a genetic algorithm is introduced to improve the performance of network coding. Simulation results show that adding optimization to the topology structure can significantly improve network capacity. Description of the drawings:

[0039] Attached Figure 1 It is a schematic diagram of the multicast satellite network model in the present invention.

[0040] Attached Figure 2 It is a flowchart of traversing the network topology into set a and then traversing into the chromosome in the present invention.

[0041] Attached Figure 3 It is a schematic diagram of the chromosome crossover operation process in the present invention.

[0042] Attached Figure 4 It is a schematic diagram of multi-source network coding capability and iteration in an embodiment of the present invention.

[0043] Attached Figure 5 It is a curve diagram of the relationship between the multi-source network coding capacity and the transmission power in the embodiment of the present invention.

[0044] Attached Figure 6 1 is a graph showing the multi-source network coding capacity and storage capacity in an embodiment of the present invention.

[0045] Attached Figure 7 It is the multi-source network coding capacity and definition number in the embodiment of the present invention.

[0046] Attached Figure 8 4 is a curve diagram showing the relationship between the multi-source network coding capacity and the number of sources in an embodiment of the present invention. Specific implementation method:

[0047] The present invention uses a time expansion diagram to represent the dynamic changes and network connections between satellites, and uses an improved genetic algorithm to perform topology optimization and traffic distribution on the satellite network. Its purpose is to improve the multicast performance of the satellite network by transmitting the content of a single source satellite to multiple target satellites under the network coding mechanism, thereby increasing the network capacity from the perspective of the basic network topology. The simulation results show that the network capacity is significantly improved by using the network coding scheme after topology optimization.

[0048] The present invention considers Figure 1 The wireless multicast satellite network shown in the figure is composed of multiple communication satellites. Source satellite 1 sends data to target satellite 1.1 and target satellite 1.2, and source satellite 2 sends data to target satellite 2.1 and target satellite 2.2. Each satellite can establish multiple inter-satellite wireless links (ISLs) with other satellites to become a network node. Each node has sufficient communication and computing functions and can be used as a source node, relay node or destination node as needed. In multicast transmission using network coding, multiple source satellites first encode the information and transmit it to the relay satellite connected through the established link. The relay satellite then encodes and forwards these data packets. The destination satellite that needs to obtain data from one of the source satellites collects and decodes the data packets from the relay satellite to obtain the original information.

[0049] Given the characteristics of satellites, the periodic motion of satellites will lead to periodic and predictable performance changes of ISL, even if it exists over time. For ease of analysis, we regard the network as static in a short period of time. Therefore, the total transmission time T is divided into N time slots, and the length of each time slot is τ = T / N. When τ is small enough, the change of ISL characteristics can be ignored. Among them, is the vertex set, Denote the i-th satellite node in the t-th time slot, E, where S denotes the source satellite set, R is the relay satellite set, and D is the destination satellite set. E is a set of edges, which includes communication edges between two satellites in a satellite slot and storage edges between two satellites in a satellite slot. The capacity of the communication edge is defined as the throughput of the ISL between two satellites. The capacity is expressed as

[0050]

[0051] Where B is the ISL bandwidth, is the transmission power of the ith satellite in the tth time slot, is the average noise power,

[0052] is the ISL gain between the i-th satellite and the j-th satellite, which can be calculated as

[0053] n=kBΓ(3), where λ is the wavelength, G tr and G re are the transmit gain and receive gain respectively, is the distance between the i-th satellite and the j-th satellite, k is the Boltzmann constant, Γ is the noise temperature, and is expressed as The storage edge of , whose capacity is the storage capacity STOi of the i-th satellite.

[0054] Problem model construction:

[0055] Using the established time-expanded graph G to describe the topology of the satellite network, the multicast transmission problem can be extended from the source satellite to the target satellite flow distribution problem. The vertex s(1) of G is the vertex set D between the source and the destination. The network topology is G, and a set of binary To describe whether should be established, considering that each satellite has a limited number of antennas, with the constraint

[0056]

[0057] Unlike traditional transmission methods, since the information packets are encoded by the source satellite relay satellite, the information packets are compressed and coordinated. Therefore, the general flow balance constraint that the inflow and outflow of each node are equal is invalid for our problem. But we can still see that the information collected by the same target satellite follows the flow balance constraint at each node. In order to represent the source satellite via relay satellite and transmission, and eventually to the target satellite The amount of information collected,

[0058] Ds ∈D is the receiver from The set of target satellites for data, we define Now the traffic balance constraint of each relay satellite can be expressed as

[0059]

[0060] In addition, all information is sent by the source satellite and collected by the target satellite within a given time, so there are flow constraints on the source and target satellites.

[0061]

[0062] According to the linear network coding theory, During the transmission process, the amount of data transmitted is equal to the maximum flow of the ISL to different target satellites and cannot exceed the capacity of the ISL.

[0063]

[0064] Similarly, for the data that has been stored by satellite i but not transmitted Their number cannot exceed the storage capacity.

[0065]

[0066] After finding the constraints, network coding can be used to formulate the optimization problem of multicast satellite networks. The optimization problem is to optimize the scheme of establishing ISLs. Traffic distribution Require Encoding capacity In line with linear network coding theory, the optimization problem is generally

[0067]

[0068] This problem is an integer linear programming problem, so there is usually no effective way to get the optimal solution. Therefore, an auxiliary algorithm is designed to obtain a suboptimal solution. The original optimization problem can be decomposed into a topology problem and several maximum flow problems. The maximum flow problem is easy to solve, so we need to find a way to construct an effective topology A.

[0069] In this paper, a method for constructing ISLs based on genetic algorithm (GA) is proposed, which is a search algorithm inspired by natural evolution theory and searches for high-quality solutions by imitating natural selection and reproduction processes.

[0070] Individuals with different chromosomes are randomly generated to form an initial population. By evaluating the transmission performance of the network topology represented by each individual as its fitness function, we determine which individuals can be replicated. The chromosomes of these individuals are crossed to simulate sexual reproduction in nature, so as to find other better solutions while retaining some advantages of the current network topology. At the same time, the population is periodically randomly updated to simulate gene mutations, encourage searches in unknown areas of the solution space, and prevent the algorithm from falling into local optimal solutions. The above process is repeated iteratively. After the population characteristics tend to stabilize or reach the maximum number of iterations M, the individual with the highest fitness is selected, and the ISL establishment set A represented by it is the solution obtained by this algorithm.

[0071] Specifically, GA includes the following steps:

[0072] (1) Create an initial population, which is a set of randomly selected valid candidate solutions (individuals). Each individual has a different randomly generated chromosome. A complete topology needs to be built to obtain the optimization variable A. As long as there are two satellites visible, ISL is built and added as an element to the set A, and A is encoded into the chromosome g. i In Figure 2 As shown. If there are P individuals, there is a population G = {g i |i∈P};

[0073] (2) Calculate the fitness The fitness function value of each individual is calculated. This operation is performed once for the initial population and then repeated for each generation after applying the genetic operators of selection, crossover, and mutation. Since the fitness of each individual is independent of other individuals, it can be calculated in parallel. To solve this problem, we formulate the optimization problem as a fitness function, using the network topology represented by each individual chromosome to obtain the maximum multicast capacity that can be achieved by network coding, and determine the value of each individual. Since the variable A is already given by the chromosome, the fitness function can be calculated as follows:

[0074]

[0075] This problem is difficult to solve directly, so we designed a multi-source, multi-target traffic allocation method. The main idea of ​​this method is to first solve the traffic allocation problem, i.e., the maximum flow problem, for each source node separately, and accumulate the solved flows to find the overflowed edges. Subgraphs are created according to the maximum flows of different source nodes, and the capacity of the corresponding edges in the subgraph is reduced according to the exceeded capacity. Finally, the maximum flow problem is solved separately on the new subgraphs. The process is summarized in Table 1.

[0076] Table 1

[0077]

[0078] Selection, crossover, and mutation Applying the genetic operators of selection, crossover, and mutation to a population produces a new generation based on the better individuals in the current generation. The purpose of selection is to select individuals that have advantages in the current population. The probability of an individual chromosome being selected for reproduction is related to its fitness f(g i ). The set p i is the probability that the chromosome of the i-th individual is selected,

[0079]

[0080] Crossing over is responsible for producing offspring from selected individuals, a process usually accomplished by two selected individuals exchanging parts of their chromatids to produce two new chromosomes representing the offspring, e.g. Figure 3 shown.

[0081] The intersection is k, and the parent chromosomes are g i and g j , then the offspring g new It can be expressed as:

[0082] g new =(g i1 ,g i2 ,…g ik ,g j(k+1) ,…,g jn )

[0083] The mutation operation randomly changes one or more chromosome values ​​(genes) of each newly created individual. The probability of mutation is usually very low. We use uniform mutation to replace the original chromosome at each site of the individual with a random number uniformly distributed within a certain range, with a small probability. nk , the mutation operation can be expressed as:

[0084] The whole algorithm is summarized in Table 2.

[0085]

[0086] Example:

[0087] To verify the performance of the topology construction method, a Walker-Delta constellation consisting of 3 orbits, each containing 20 LEO satellites, is built. We randomly select source and target satellites from all satellites and repeat 50 rounds to get the average result. Then, we will evaluate the performance based on the multi-source coding capacity, which is the sum of the coding capacities from different sources. The simulation parameters are shown in Table 3.

[0088] Table 3

[0089]

[0090] First, we focus on how the fitness of the population of the genetic algorithm (i.e., the encoding capacity of the topology it represents) changes with the number of iterations. We randomly select 2 source satellites and 2 target satellites for each source satellite, set the transmission power to 80W, the satellite storage capacity to 2000Mbits, and perform 200 iterations. Then we get Figure 4 It can be seen that as the number of iterations increases, the coding capacity gradually increases, stabilizes after 125 generations, and finally stabilizes at 2032.437Mbits after 150 generations. We can take the population after 125 generations as the solution to this problem and apply it to subsequent simulations. We can take the population after 125 generations as the solution to this problem and apply it to subsequent simulations.

[0091] The relationship between network coding capacity and satellite transmission power is as follows: Figure 5 As shown. Obviously, as the transmission power increases, the capacity of each ISL also increases, and the network coding capacity also increases. Compared with the existing fair contact plan (FCP) method, this method has obvious advantages.

[0092] Depend on Figure 6 It can be seen that when the storage capacity is low, the impact of storage capacity on network coding capacity is significant. We believe that the reason for this is that the storage capacity is too small, resulting in some data packets being cached in the relay satellites instead of waiting for more suitable transmission time slots. In addition, since our GVE models the storage behavior as an edge, this effect can also be explained as a direct reduction in the capacity of some edges, thereby reducing the maximum flow between the source and target satellites. In addition, since the storage capacity is close to the maximum network coding capacity at this time, the network coding capacity is almost independent of the storage capacity, which can be explained as the throughput of the ISL becoming the bottleneck for improving performance, rather than the storage capacity.

[0093] like Figure 7 As shown in the figure, as the number of target satellites increases, the network coding capacity decreases. According to the linear network coding theory, the coding capacity depends on the smallest capacity among the multiple target satellites. Therefore, as the number of target satellites increases, the algorithm tends to construct a topology that allocates a relatively average capacity to each target satellite. However, this topology is not the best solution to maximize the data flow of a specific target satellite, which will result in a reduction in the number of data streams received by a single satellite.

[0094] from Figure 8It can be seen that as the number of source satellites increases, the total coding capacity of the multi-source network also increases, but the rate of increase is significantly lower. The reason is that as the number of source satellites increases, the number of ISLs used by multiple source satellites also increases. When the capacity of these ISLs remains unchanged, it can only reduce the communication volume of each source satellite, so the rate of increase of the total coding capacity decreases.

[0095] Through the above multiple comparisons, it can be seen that in almost any scenario, our method is significantly better than the FCP method used as a comparison. The use of network coding technology in satellite multicast networks to improve network capacity is studied. Through the coding process, we can compress multiple data streams on the satellite, thereby minimizing the use of communication and storage resources. In addition, we also introduced a topology construction and traffic distribution method based on genetic algorithms to improve the performance of network coding. Simulation results show that adding optimization to the topology structure can significantly improve network capacity.

Claims

1. A method for constructing a wireless multicast satellite network topology and allocating traffic using a genetic algorithm. The wireless multicast satellite network is composed of multiple communication satellites. Each satellite establishes multiple inter-satellite wireless links (ISLs) with other satellites to become a network node. Each node can be used as a source node, a relay node or a destination node as required. In multicast transmission using network coding, multiple source satellites first encode information and transmit it to relay satellites connected through established links. The relay satellites then encode and forward these data packets. The destination satellite that needs to obtain data from one of the source satellites collects and decodes the data packets from the relay satellite to obtain the original information. The network in a short period of time is regarded as static. The total transmission time T is divided into N time slots. The length of each time slot is τ=T / N. When τ is small enough, the change of ISL characteristics is negligible. is the vertex set, represents the i-th satellite node in the t-th time slot, E, where S represents the source satellite set, R is the relay satellite set, D is the destination satellite set, and E is the set of edges, which includes a communication edge between two satellites in a satellite slot and a storage edge between two satellites in a satellite slot. The capacity of the communication edge is defined as the throughput of the ISL between two satellites. The capacity is expressed as, Where B is the ISL bandwidth, is the transmission power of the ith satellite in the tth time slot, is the average noise power, is the ISL gain between the i-th satellite and the j-th satellite, which can be calculated as: Where λ is the wavelength, G tr and G re are the transmit gain and receive gain respectively, is the distance between the i-th satellite and the j-th satellite, k is the Boltzmann constant, Γ is the noise temperature, and is expressed as The storage edge, whose capacity is the storage capacity STO of the i-th satellite i； It is characterized in that the established time extension graph G is used to describe the topological structure of the satellite network, and the vertex s of the time extension G of the multicast transmission problem from the source satellite to the target satellite flow distribution problem is (1) The vertex set D between the source and the destination, the network topology is G, and a set of binary To describe whether should be established, considering that each satellite has a limited number of antennas, with the constraints: To indicate the source satellite via relay satellite and transmission, and eventually to the target satellite Collected information flow, D s ∈D is the receiver from The set of target satellites for data, defined The traffic balance constraint of each relay satellite is expressed as: In addition, all information is sent by the source satellite and collected by the target satellite within a given time, so there are flow constraints on the source satellite and the target satellite: According to the linear network coding theory, During the transmission process, the amount of data transmitted is equal to the maximum flow of the ISL to different target satellites and cannot exceed the capacity of the ISL. Therefore, Similarly, for the data that has been stored by satellite i but not transmitted Their number cannot exceed the storage capacity, so there is After finding the constraints, network coding is used to construct the optimization problem of the multicast satellite network. The optimization problem is to optimize the scheme of establishing ISLs. Traffic distribution Require Encoding capacity In line with linear network coding theory, specifically, the optimization problem is: st(4)-(10) The optimization problem is solved by an ISL construction method based on genetic algorithm GA.

2. The wireless multicast satellite network topology construction and genetic algorithm traffic distribution method according to claim 1 is characterized in that: The ISL construction method based on the genetic algorithm GA to solve the optimization problem includes: individuals with different chromosomes are randomly generated to form an initial population, and the transmission performance of the network topology represented by each individual is evaluated as its fitness function to determine which individuals can be replicated, and the chromosomes of these individuals are crossed to simulate sexual reproduction in nature, so as to find other better solutions while retaining some advantages of the current network topology structure. At the same time, the population is periodically randomly updated to simulate gene mutations, and searches in unknown areas of the solution space are encouraged to prevent the algorithm from falling into a local optimal solution. The above process is repeated iteratively. After the population characteristics tend to be stable or the maximum number of iterations M is reached, the individual with the highest fitness is selected, and the ISL establishment set A represented by it is the solution obtained by this algorithm, which specifically includes the following steps: Step 1: Create an initial population, which is a set of randomly selected valid candidate solutions (individuals), each of which has a different randomly generated chromosome; build a complete topology to get the optimization variable A, and as long as two satellites are visible, build ISL and add ISL as an element to the set A, and encode A to chromosome g i If there are P individuals, there is a population G = {g i |i∈P}; Step 2: Calculate the fitness The fitness function value of each individual is calculated. This operation is performed once for the initial population and then repeated for each generation after applying the genetic operators of selection, crossover and mutation. Since the fitness of each individual is independent of other individuals, it can be calculated in parallel. In order to solve this problem, the optimization problem is used as the fitness function. The network topology represented by each individual chromosome is used to obtain the maximum multicast capacity that can be achieved by network coding and determine the value of each individual. Since the variable A is given by the chromosome, the fitness function is calculated as follows: st(6)-(10) Step 3: Selection, crossover and mutation Apply the genetic operators of selection, crossover and mutation to a population to produce a new generation based on the better individuals in the current generation. The purpose of selection is to select individuals that have advantages in the current population. The probability of an individual chromosome being selected for reproduction is related to its fitness f(g i ), the set p i is the probability that the chromosome of the i-th individual is selected, Crossover is responsible for producing offspring from selected individuals. This process is usually accomplished by two selected individuals exchanging parts of their chromatids to produce two new chromosomes representing the offspring, with the intersection being k and the parental chromosomes being g. i and g j , then the offspring g new It can be expressed as: g new =(g i1 ,g i2 ,…g ik ,g j(k+1) ,…,g jn ), the mutation operation randomly changes one or more chromosome values ​​(genes) of each newly created individual. The probability of mutation is usually very low. Uniform mutation is used. For chromosome g nk , the mutation operation is expressed as: g' nk =g nk +δ,with probabilityμ。 3. The wireless multicast satellite network topology construction and genetic algorithm traffic distribution method according to claim 2 is characterized in that: In step 2, perform the following operations to solve the maximum flow problem: Step 2-1: First, solve the flow distribution problem, i.e., the maximum flow problem, for each source node, and accumulate the solved flows to find the overflow edge; Step 2-2: Create a subgraph based on the maximum flow of different source nodes, and reduce the capacity of the corresponding edge in the subgraph according to the exceeded capacity; Step 2-3: Solve the maximum flow problem on each of the new subgraphs.

Images