Inter-satellite routing redirection method based on integration of integer linear programming and genetic algorithm

Abstract

The invention relates to an inter-satellite routing redirection method based on integral linear programming and genetic algorithm fusion, and belongs to the technical field of satellite network communication. The method mainly solves the technical problems that when a satellite network link fails, a traditional route redirection method is low in convergence speed, and real-time service continuity is difficult to guarantee. According to the technical scheme, a satellite network topology model is constructed and a random link fault is introduced through a hybrid optimization method combining integer linear programming (ILP) and a genetic algorithm (GA), a multi-constraint optimization target and a hard and soft constraint condition are defined by utilizing the ILP, the GA is guided to perform directional search operations such as population initialization, fitness calculation, selection, intersection and variation, and the robustness of the satellite network is improved. And an optimization path for avoiding the fault is autonomously planned. According to the method, the convergence speed, the path quality and the calculation efficiency of route redirection are effectively improved, the solution diversity is maintained through an evolution mechanism, the robustness and the reliability of a satellite network are remarkably enhanced, and technical support is provided for intelligent operation and maintenance of a large-scale constellation network.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of satellite network communication and relates to an intersatellite routing redirection method based on integration of integer linear programming and a genetic algorithm. BACKGROUND

[0002] With the acceleration of global digitalization, space information networks, as a key infrastructure for building space-earth integration networks, are playing an increasingly important role. Among them, intersatellite laser communication technology, with its high bandwidth, low delay, strong anti-interference and other advantages, has become the core transmission means of the new generation of satellite networks. However, satellite networks have highly dynamic characteristics, and link failures occur frequently. Traditional routing protection schemes based on static configuration or predetermined rules have been difficult to adapt to such extreme dynamics. When an intersatellite link is interrupted, traditional methods need to recalculate the global path, resulting in slow convergence speed and failing to meet the continuity requirements of real-time services. Therefore, how to improve the stability, reliability and adaptability of intersatellite communication systems has become a research frontier and core challenge in the field of intersatellite routing redirection.

[0003] In the prior art, researchers have proposed a variety of heuristic algorithms to optimize intersatellite routing. For example, Yang Jingwen et al. designed a heuristic intersatellite routing algorithm that integrates genetic algorithm and ant colony algorithm to optimize path cost and end-to-end delay under bandwidth and maximum tolerance delay constraints, but it does not solve the problem of topology division complexity caused by polar region link switching; Chen Quan et al. studied the low-orbit giant constellation satellite-ground integrated network from four aspects: topology dynamicity management, intersatellite forwarding hop evaluation, distributed congestion avoidance routing algorithm design, and gateway station layout optimization, but did not consider network robustness in extreme scenarios; Jing Y et al. proposed an energy-saving routing scheme based on genetic algorithm, which selects appropriate discharge depth of satellites to improve satellite life, but fixed number of sleep satellites in heavy load will lead to increased blocking rate, lacking dynamic adjustment mechanism. These methods generally rely on simple heuristic rules, are prone to local optimal solution, and do not fully consider the mathematical rigor of multi-constraint optimization problems.

[0004] Integer Linear Programming (ILP) as a rigorous mathematical modeling tool can convert the multi-constraint optimization requirements of inter-satellite routing redirection into objective functions and hard-soft constraint conditions, providing a lower bound of the global optimal solution. Genetic Algorithm (GA) is good at efficiently searching for feasible solutions in large-scale solution space. However, existing research fails to effectively combine the complementary advantages of ILP and GA, resulting in deficiencies in convergence speed, path quality and computational efficiency. The present application fuses ILP and GA to construct a hybrid optimization framework to solve the satellite network routing reconstruction problem under link failure scenarios, significantly enhancing the robustness and reliability of space communication systems. SUMMARY

[0005] Therefore, the present application aims to provide an inter-satellite routing redirection method based on the fusion of integer linear programming and genetic algorithm.

[0006] To achieve the above-mentioned purpose, the present application provides the following technical solutions:

[0007] An inter-satellite routing redirection method based on the fusion of integer linear programming and genetic algorithm, comprising the following steps:

[0008] Obtaining the satellite network topology, traffic demand set and fault link set;

[0009] Initializing the genetic algorithm (GA) population, wherein each individual represents a complete routing scheme;

[0010] For multiple generations, calculate the fitness of each individual, which is based on the integer linear programming (ILP) objective function, including the delay reward, load balancing reward, bandwidth compliance reward and constraint violation penalty term;

[0011] Performing selection operation based on the fitness to retain high fitness individuals;

[0012] Performing crossover operation on the selected individuals to generate offspring individuals;

[0013] Performing mutation operation on the offspring individuals; updating the population and repeating the above steps until the termination condition is met;

[0014] Outputting the individual with the highest fitness as the optimal routing scheme.

[0015] Furthermore, the objective function of the ILP is a maximization function: max{ω1time+ω2load+ω3band-λViolat}, where ω1, ω2, and ω3 are weight coefficients and ω1+ω2+ω3=1, λ is a penalty coefficient, time is the latency score, load is the load balancing score, band is the bandwidth compliance score, and Violat is the constraint violation score.

[0016] Furthermore, the formula for calculating the delay score (time) is as follows:

[0017]

[0018] Where D is the set of all traffic demands, E is the set of effective inter-satellite links, and f d,e The bandwidth allocated on link e for traffic demand d, t e For the fixed transmission delay of link e, dem d Let d be the bandwidth requirement for traffic flow.

[0019] Furthermore, the formula for calculating the load balancing score (load) is as follows:

[0020]

[0021] Among them B e Let be the maximum bandwidth capacity of link e, and ∈ be a very small constant to avoid division by zero.

[0022] Furthermore, the formula for calculating the bandwidth compliance score (band) is as follows:

[0023]

[0024] Among them B e This represents the maximum bandwidth capacity of link e.

[0025] Furthermore, the formula for calculating the constraint violation score Violat is as follows:

[0026]

[0027] Where E fall Let x be the set of faulty links. d,e For traffic demand d, determine whether to select link e as a binary variable, s d The source node is the one with traffic demand d.

[0028] Furthermore, in the initialization step of the genetic algorithm (GA) population, each individual is generated in the following manner: a path is generated with an 80% probability using the shortest path algorithm, and a random feasible path is generated with a 20% probability, and all paths avoid the set of faulty links.

[0029] Furthermore, the selection operation employs a tournament selection method, which involves randomly selecting multiple individuals and retaining the individual with the highest fitness as the parent.

[0030] Furthermore, the crossover operation is performed independently for each traffic demand, and the offspring individuals randomly inherit the path of parent generation 1 or parent generation 2, and the crossover satisfies the hard constraint condition of integer linear programming (ILP).

[0031] Furthermore, the mutation operation regenerates a new path for each traffic demand path with a 10% probability. The new path is created by a random path generation function and satisfies integer linear programming (ILP) constraints.

[0032] The beneficial effects of this invention are as follows:

[0033] (1) This invention effectively overcomes the inherent defect of traditional heuristic algorithms, which are prone to getting trapped in local optima, by deeply integrating integer linear programming and genetic algorithms. Integer linear programming provides a rigorous mathematical modeling framework for routing optimization problems and clarifies the search direction for the global optimum; while genetic algorithms fully utilize their global search capabilities to efficiently explore a broad solution space. The two algorithms complement each other, significantly improving the overall quality of routing solutions.

[0034] (2) This method significantly enhances the autonomous recovery capability of satellite networks in the event of link failure. When inter-satellite links are interrupted, the algorithm can quickly plan an optimized path to avoid the failure, ensuring the continuity and stability of communication services. This feature greatly improves the robustness and reliability of space communication systems in dynamic topology environments.

[0035] (3) This invention provides clear guidance for the search process of the genetic algorithm through the constraints set by integer linear programming. Hard constraints effectively filter out invalid solutions and avoid resource waste; soft constraints guide the search direction through the fitness function. This constraint guidance mechanism greatly improves the convergence speed of the algorithm and shortens the computation time of route redirection.

[0036] (4) This method exhibits excellent load balancing characteristics. By optimizing the load balancing reward term in the objective function, the algorithm can automatically balance the traffic load of each link, avoid network congestion, and thus improve the overall utilization efficiency of satellite network resources.

[0037] (5) The inherent evolutionary mechanism of genetic algorithms ensures the diversity of the population. Through operations such as selection, crossover, and mutation, the population can maintain sufficient diversity, avoid premature convergence, and ensure that the algorithm can adapt to various complex network scenarios. This diversity maintenance mechanism provides a strong guarantee for the long-term stable operation of satellite networks.

[0038] Other advantages, objectives, and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination, or may be learned from practice of the invention. The objectives and other advantages of the invention can be realized and obtained through the following description. Attached Figure Description

[0039] To make the objectives, technical solutions, and advantages of the present invention clearer, the preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings, wherein:

[0040] Figure 1 Design a flowchart for the intelligent optimization algorithm;

[0041] Figure 2 This represents the inter-satellite network topology.

[0042] Figure 3 A performance comparison of ACO, PSO, and GA algorithms;

[0043] Figure 4 This study compares the performance of various heuristic algorithms guided by the ILP framework. Detailed Implementation

[0044] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0045] The accompanying drawings are for illustrative purposes only and are schematic diagrams, not actual pictures. They should not be construed as limiting the invention. To better illustrate the embodiments of the invention, some parts in the drawings may be omitted, enlarged, or reduced, and do not represent the actual product dimensions. It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0046] In the accompanying drawings of the embodiments of the present invention, the same or similar reference numerals correspond to the same or similar components. In the description of the present invention, it should be understood that if terms such as "upper," "lower," "left," "right," "front," and "rear" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, they are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, the terms used to describe positional relationships in the drawings are only for illustrative purposes and should not be construed as limiting the present invention. For those skilled in the art, the specific meaning of the above terms can be understood according to the specific circumstances.

[0047] 1 Algorithm Design

[0048] 1.1 ILP Constraints

[0049] The core function of ILP is to transform the multi-constraint optimization requirements of inter-satellite routing redirection into a rigorous mathematical model, providing a lower bound for the optimal solution. Heuristic algorithms (ACO / PSO / GA) are responsible for efficiently searching for feasible solutions in a large-scale solution space. The core logic of their combination is: ILP defines the optimization objective plus hard / soft constraints; the heuristic algorithm performs a directional search within the ILP framework (avoiding invalid solutions); and then ILP performs local optimization on the heuristic results. That is, hard constraints are directly used to filter invalid routes, while soft constraints are integrated into the fitness function to guide optimization, ensuring that the heuristic algorithm searches only within the feasible solution space. This invention maximizes constraints from three aspects: latency, bandwidth, and load balancing. The higher the fitness value of the objective function result, the better the routing scheme. If hard constraints are violated, the penalty term will significantly reduce the fitness, and the heuristic algorithm will automatically eliminate such invalid solutions. Table 1 shows the constants and defined variables:

[0050] Table 1. Definitions of ILP Algorithm Constants and Variables

[0051] D Set of all traffic demands E Set of active inter-star links E fall ]]> Set of failed links dem d ]] Bandwidth demand of traffic demand d [cat e ]]> Fixed transmission delay of link e B e ]]> Maximum bandwidth capacity of link e f d,e ]]> Bandwidth allocated to traffic demand d on link e x d,e ]]> Whether traffic demand d chooses link e [["ω1", "ω2", "ω3"]] weighting factors (ω1+ω2+ω3=1) λ Constraint violation penalty coefficient

[0052] The objective function and constraints are as follows:

[0053] Objective function:

[0054] max{ω1time+ω2load+ω3band-λViolat} (1)

[0055]

[0056] Constraints:

[0057] x d,e =0, f d,e =0 (6)

[0058]

[0059]

[0060] Formula (1) is the objective function, which aims to guide the heuristic algorithm to achieve the three core objectives of low latency, load balancing, and bandwidth compliance. Formulas (2) to (5) are the latency score reward, load balancing reward, bandwidth compliance reward, and violation penalty score, respectively. Formulas (6) to (11) are hard constraints, which must be met, otherwise the solution is invalid. Formula (6) is a fault link constraint, which prohibits the use of fault links. Formula (7) is a link traffic constraint, which stipulates that the total bandwidth of traffic on link e does not exceed the link capacity. Formula (8) is a traffic conservation constraint, which stipulates that the inflow traffic from non-source satellites equals the outflow traffic. Formula (9) is a traffic conservation constraint, which stipulates that the inflow traffic from source satellites equals the outflow traffic. Formula (9) allows traffic to be allocated only when traffic d selects link e. Formula (10) is a binary constraint. Formulas (12) to (14) are soft constraints, which incorporate adaptive guidance to improve the robustness of the solution.

[0061] 1.2 Intelligent Optimization Algorithm Design

[0062] Genetic algorithms solve inter-satellite routing optimization problems by simulating biological evolution. They encode routing schemes as chromosomes and gradually evolve optimal solutions through operations such as selection, crossover, and mutation, as shown in the flowchart. Figure 1 As shown.

[0063] First, the population is initialized. Chromosomes represent a complete routing scheme for each individual, containing a set of paths for all traffic demands. Each gene corresponds to a path for a traffic demand. Paths are generated with an 80% probability using the shortest path algorithm and a 20% probability using a random feasible path, satisfying ILP constraints and avoiding faulty links. After the loop begins, the population fitness is calculated, the traffic load of each link is calculated, the maximum link utilization is optimized, ILP rewards and constraint violation penalties are set, and a fitness function is used for looping. A tournament selection process retains the best individual of the current generation, selecting the individual with the highest fitness as the parent. Crossover is performed independently for each traffic demand, and offspring randomly inherit paths from either parent 1 or parent 2. After crossover, offspring individuals must satisfy the ILP hard constraints. Each traffic path has a 10% probability of being regenerated using a random path generation function to create new paths and prevent premature convergence. After reaching the maximum number of generations, the historical best individual is used as the final routing scheme. The algorithm flow is as follows:

[0064] Algorithm: Inter-satellite route redirection optimization based on genetic algorithm

[0065] Inputs: satellite network, traffic demands, failed links, population size, maximum number of generations.

[0066] Output: Best-solution routing solution

[0067] 1) Create inter-satellite topology;

[0068] 2) Initialize the population: population = cell(population-size, 1), and loop i = 1: population-size; obtain the source and destination nodes, and generate the path using an algorithm based on the probability;

[0069] 3) Obtain the loop result: solution.('demand-d') = path, population{i} = solution;

[0070] 4) Initialize fitness history: fitness-history = zeros(generations, 1); best-fitness = 0;

[0071] 5) Loop gen = 1: generations; i = 1: population-size; Initialize maximum utilization and bandwidth violation count in the loop, obtain path and traffic requirements, and calculate fitness-scores(i) = 1 / (max-utilization + 0.001 + total-bandwidth-violation × 0.1);

[0072] 6) The result is [current-best-fitness, best-idx] = max(fitness-scores);

[0073] 7) Update fitness-history(gen)=current-best-fitness;

[0074] 8) Create a new population; retain elites: new-population{1} = population{best-idx};

[0075] 9) Tournament selection, loop i = 2: population-size; set tournament size tournament-size = 3, then randomly select tournament participants, find the best individual to become parent1, loop 3 times;

[0076] 10) Create an empty child child = struct(); loop d = 1: length(traffic-demands), if the probability is greater than 0.5, inherit from parent 1, otherwise inherit from parent 2; end the loop (crossing);

[0077] 11) Set the mutation rate to 0.1; loop d = 1:length(traffic-demands); if the mutation rate is less than 0.1, regenerate the path new-path = GenerateRandomPath(graph,source,destination), child.('de mand_d') = new-path; add the offspring to the new population new-population{i} = child; end the loop;

[0078] 12) Population update: population = new - population; returns best - solution;

[0079] 2 Experimental Platform and Results

[0080] 2.1 Experimental Platform

[0081] This experiment was implemented using MATLAB 2024a and the ILP cplex terminal, with a hardware environment of 13th generation Intel CPU. R Core TM The system uses an i5-13400 processor at 2.50 GHz and 16.0 GB of RAM. Experiments were conducted using Ant Colony Optimization (ACO), Particle Swarm Optimization (PSO), and a Genetic Algorithm (GA) under the guidance of the ILP framework for route redirection of satellite faulty paths. One hundred simulations were performed, and the experiments were compared and analyzed based on latency, maximum link utilization, and runtime cost. The results show that the GA guided by the ILP framework is superior.

[0082] 2.2 Simulation Comparison

[0083] (1) Inter-satellite network topology planning

[0084] This experiment uses 18 satellites in 3 orbital planes, with 6 satellites per plane, for a total of 40 links. It simulates the bandwidth of a real satellite network. Red indicates faulty links, and blue indicates normal links. Figure 2 As shown.

[0085] (2) Performance comparison of various heuristic algorithms

[0086] Performance comparison of different algorithms, for example Figure 3 As shown in the figure, after comprehensively evaluating the three indicators of total latency, maximum link utilization, and computation time, the results show that the genetic algorithm performs best, verifying its applicability in solving the inter-satellite routing redirection problem.

[0087] (3) Performance comparison of various heuristic algorithms under the guidance of the ILP framework

[0088] This invention designs and compares the performance of various algorithms within the ILP framework. For example... Figure 4 As shown, with Figure 3 Compared with the baseline results, under the guidance and constraints of ILP, the performance indicators of all algorithms have been effectively improved.

[0089] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. An inter-satellite route redirection method based on the fusion of integer linear programming and genetic algorithm, characterized in that: Includes the following steps: Obtain the satellite network topology, traffic demand set, and faulty link set; Initialize the genetic algorithm (GA) population, where each individual represents a complete routing scheme; For multiple generations, the fitness of each individual is calculated based on an integer linear programming (ILP) objective function, which includes latency rewards, load balancing rewards, bandwidth compliance rewards, and constraint violation penalties. A selection operation is performed based on the fitness to retain individuals with high fitness. Perform a crossover operation on the selected individuals to generate offspring individuals; Perform mutation operations on offspring individuals; Update the population and repeat the above steps until the termination condition is met; The individual with the highest fitness is output as the optimal routing scheme.

2. The inter-satellite route redirection method based on the fusion of integer linear programming and genetic algorithm as described in claim 1, characterized in that: The objective function of the ILP is a maximization function: max{ω1time+ω2load+ω3band-λViolat}, where ω1, ω2, and ω3 are weight coefficients and ω1+ω2+ω3=1, λ is the penalty coefficient, time is the latency score, load is the load balancing score, band is the bandwidth compliance score, and Violat is the constraint violation score.

3. The inter-satellite route redirection method based on the fusion of integer linear programming and genetic algorithm according to claim 2, characterized in that: The formula for calculating the delay score (time) is as follows: Where D is the set of all traffic demands, E is the set of effective inter-satellite links, and f d,e The bandwidth allocated on link e for traffic demand d, t e For the fixed transmission delay of link e, dem d Let d be the bandwidth requirement for traffic flow.

4. The inter-satellite route redirection method based on the fusion of integer linear programming and genetic algorithm according to claim 2, characterized in that: The formula for calculating the load balancing score (load) is as follows: Among them B e Let be the maximum bandwidth capacity of link e, and ∈ be a very small constant to avoid division by zero.

5. The inter-satellite route redirection method based on the fusion of integer linear programming and genetic algorithm according to claim 2, characterized in that: The formula for calculating the bandwidth compliance score (band) is as follows: Among them B e This represents the maximum bandwidth capacity of link e.

6. The inter-satellite route redirection method based on the fusion of integer linear programming and genetic algorithm according to claim 2, characterized in that: The formula for calculating the constraint violation score Violat is as follows: Where E fall Let x be the set of faulty links. d,e For traffic demand d, determine whether to select link e as a binary variable, s d The source node is the one with traffic demand d.

7. The inter-satellite route redirection method based on the fusion of integer linear programming and genetic algorithm according to claim 1, characterized in that: In the initialization step of the genetic algorithm (GA) population, each individual is generated in the following way: a path is generated with an 80% probability using the shortest path algorithm, a random feasible path is generated with a 20% probability, and all paths avoid the set of faulty links.

8. The inter-satellite route redirection method based on the fusion of integer linear programming and genetic algorithm according to claim 1, characterized in that: The selection operation employs a tournament selection method, which involves randomly selecting multiple individuals and retaining the individual with the highest fitness as the parent.

9. The inter-satellite route redirection method based on the fusion of integer linear programming and genetic algorithm according to claim 1, characterized in that: The crossover operation is performed independently for each traffic demand. The offspring individuals randomly inherit the path of parent generation 1 or parent generation 2, and the crossover satisfies the hard constraint condition of integer linear programming (ILP).

10. The inter-satellite route redirection method based on the fusion of integer linear programming and genetic algorithm according to claim 1, characterized in that: The mutation operation regenerates a new path for each traffic demand path with a 10% probability. The new path is created by a random path generation function and satisfies integer linear programming (ILP) constraints.

Images