A network routing planning method and system based on an improved ant colony algorithm

Abstract

The application discloses a network routing planning method and system based on an improved ant colony algorithm. First, the LPA* algorithm is used to search for effective routing on a grid map, which is used as initial pheromone setting of the ant colony algorithm, so that the guiding ability of a preselected area is improved. Then, the number of fiber sites, the length of routing and the node load are introduced, and a multi-factor loss function is defined to realize multi-objective optimization of fiber routing planning. Finally, the heuristic function definition of the multi-factor improved ACO algorithm is combined, and the pheromone evaporation factor is optimized, so that the algorithm performance is further improved. The application effectively solves the problems that the traditional fiber routing planning method is prone to falling into a local optimal solution and slow convergence speed, and has important practical application value for park fiber routing resource planning and management.

Description

Technical Field

[0001] This invention belongs to the field of routing planning technology, specifically involving route optimization based on a multi-factor intelligent biological search algorithm. Background Technology

[0002] Network routing optimization aims to find the optimal route based on network topology, node locations, and service requirements, while satisfying network performance, reliability, and scalability requirements. However, fiber optic network planning is a complex combinatorial optimization problem. Conventional routing optimization methods suffer from high computational complexity and are prone to getting trapped in local optima and slow convergence, resulting in inefficiency. Summary of the Invention

[0003] To address the aforementioned issues, this invention proposes a network routing planning method and system based on an improved ant colony algorithm. It utilizes a heuristic algorithm to search for an effective route on a grid map as the initial pheromone setting for the ant colony algorithm, enabling the ant colony to find feasible routes more quickly, further improving algorithm performance and demonstrating superior performance in handling complex scenarios.

[0004] A network routing planning method based on an improved ant colony algorithm, which achieves one of the objectives of this invention, includes the following steps:

[0005] S1. Calculate the transition probability of each ant moving from the current network node to every possible next network node based on the distance between two network nodes, the distance between the network node and the target node, and the load of the network node; the load of the network node is used to represent the network congestion degree of the node.

[0006] In the ant colony algorithm, all ants start from the initial node of the network where the route to be planned is to be completed; they select the next network node based on the transition probability until the first search is completed.

[0007] S2. Calculate the pheromone increment of the route where the network node is located based on the total loss value; the total loss value is used to represent the number of network devices on the network route traversed by the ant, the total load of the nodes traversed to complete one route search, and the total route length; the pheromone increment is used to calculate the transition probability;

[0008] S3. When the number of searches reaches the set limit or the search routes of all ants are consistent after multiple consecutive searches, the globally optimal fiber route from the starting node to the target node is obtained; otherwise, the pheromone between nodes is updated according to the pheromone increment and the pheromone between nodes in the previous iteration, the transition probability is calculated according to the updated pheromone between nodes, a new round of search is started according to the transition probability, and the process returns to step S2 after the new round of search is completed.

[0009] A network routing planning system based on an improved ant colony algorithm, which achieves the second objective of this invention, includes a transition probability calculation module, an ant colony initialization module, a pheromone increment calculation module, a transition probability update module, and an optimal network route judgment module.

[0010] The transfer probability calculation module is used to calculate the transfer probability of each ant moving from the current network node to each possible next network node based on the distance between two network nodes, the distance between the network node and the target node, and the load of the fiber network node.

[0011] The ant colony initialization module is used to initialize all ants in the ant colony algorithm to start from the starting node of the network to be planned; select the next network node according to the transition probability and the set pheromone concentration between nodes, until the first search is completed; the pheromone concentration between network nodes is used to guide the ants to search for network routes;

[0012] The pheromone increment calculation module is used to calculate the pheromone increment of the route where the fiber network node is located based on the total loss value.

[0013] The transition probability update module is used to update the pheromone between nodes based on the pheromone increment and the pheromone between nodes in the previous iteration, and to calculate the transition probability based on the updated pheromone between nodes.

[0014] The optimal network route determination module is used to determine whether the currently obtained route is the optimal network route. When the number of searches reaches the set limit or after multiple consecutive searches, all ants have the same search route, then the globally optimal network route from the starting node to the target node is obtained.

[0015] When the current route is not the optimal network route, the transition probability update module is called to update the transition probabilities between nodes. All ants start from the starting node of the network with the route to be planned and proceed to the next round of iteration based on the transition probabilities.

[0016] Beneficial effects:

[0017] 1. Introducing load values ​​between nodes to calculate the transition probability can make the searched network routes more optimal;

[0018] 2. By setting a multi-factor loss function, the actual route length cost, node access load cost, and number of network services accessed in network routing planning are described respectively. This transforms the routing planning problem into a multi-objective optimization problem. Through these multi-factor constraints, a multi-faceted optimized network route with the shortest route, relatively few traversed nodes, and relatively low route load is obtained.

[0019] 3. Considering that the adaptive volatility factor decreases continuously with the iterative changes of multiple factors in the experiment, setting an upper and lower limit for the volatility factor can prevent the algorithm from stalling due to the volatility factor being too small. Attached Figure Description

[0020] Figure 1 This is an embodiment of the method described in this invention;

[0021] Figure 2 It is the locally optimal route obtained by the heuristic method selected in this invention;

[0022] Figure 3 The final global network optimal route is obtained through the method described in this invention. Detailed Implementation

[0023] The following detailed embodiments are provided to explain the technical solutions of the claims of this invention, so that those skilled in the art can understand the claims. The scope of protection of this invention is not limited to the following specific embodiments. Any modifications made by those skilled in the art that incorporate the technical solutions of the claims but differ from the following detailed embodiments are also within the scope of protection of this invention.

[0024] S1. Calculate the transition probability of each ant moving from the current network node to every possible next network node based on the distance between two network nodes, the distance between the network node and the target node, and the load of the network node; the load of the network node is used to represent the network congestion degree of the node.

[0025] The probability of an ant moving to the next search node or link is determined by the route transition probability, and the next node is selected by roulette wheel selection.

[0026] The formula for the transfer probability of ants is as follows:

[0027]

[0028] In the formula:

[0029] The probability of the k-th ant moving from node i to every possible next node j;

[0030] τ ij (t): The pheromone concentration from node i to node j at the t-th iteration;

[0031] η ij (t): The heuristic function value from node i to node j at the t-th iteration;

[0032] α: Pheromone importance coefficient; in this embodiment, the values ​​are shown in Table 2;

[0033] β: Distance importance coefficient; in this embodiment, the values ​​are shown in Table 2;

[0034] v k : The set of all possible nodes that ant k can choose next when it is at node i.

[0035] The pheromone is used as a positive feedback mechanism to guide ants to choose the best route during the search process. The more pheromones on a route, the greater the probability that an ant will choose that route. Eventually, most ants will tend to choose the best route, thus obtaining the optimal route.

[0036] To better guide the ants towards the target point and find the route with the shortest number of hops, we need to adjust the heuristic function value η. ij (t) should be improved, and the improvement should not only consider the distance to the heuristic function value η ij The influence of (t) must also be considered in relation to the node load included in the search route. Theoretically, not only does the distance between nodes affect the heuristic function value, but the node load of the access route is also a key factor influencing the heuristic function value for selecting the next node. Therefore, in ant colony search, the heuristic function for probabilistic selection, i.e., the heuristic function value η, is crucial. ij The calculation of (t) needs to include the load values ​​of each node or link. j The load value of the node or link is used to represent the network congestion level of the node or link, and is set according to the actual network conditions.

[0037] Improved heuristic function value η ij The methods for calculating (t) include:

[0038]

[0039] D i,j =d i,j +d i,doal

[0040] Equation (3)

[0041] In the formula:

[0042] load j : Represents the load value of node j;

[0043] d i,j : The distance between optical crossover station i and optical crossover station j; in this embodiment, it is the Euclidean distance;

[0044] d i,goal : The distance between the current optical crossover station i and the target station goal; in this embodiment, it is the Euclidean distance;

[0045] In the formula D i,jIt integrates information between the current node and its neighboring nodes, as well as between the current node and the target node (goal). Compared to traditional methods, it adds an extra layer of heuristic information, which can improve search efficiency.

[0046] In the initialization of the ant colony algorithm, all ants start from the starting node of the network to be planned; the next network node is selected according to the transition probability until the first search is completed;

[0047] Traditional ACO algorithms require a uniform initial pheromone distribution, which can lead to blindness in the initial stages and slow convergence. To address these shortcomings, this embodiment combines a heuristic search algorithm with an initial locally optimal route based on the current network topology to guide a non-uniform initial pheromone distribution, improving the pre-selection capability of candidate regions during route planning. This improved method facilitates adjustment of the search speed from the current point to the target point and enhances the strength of heuristic information, thereby effectively reducing the blindness and arbitrariness of the ACO algorithm in the early stages of the search and improving global search capability and convergence speed.

[0048] The method for setting the initial pheromone for the ant colony algorithm is shown in the following formula:

[0049]

[0050] In the formula:

[0051] τ ij (0): The pheromone concentration from node i to node j at the 0th iteration;

[0052] C0: Initial pheromone concentration on the pre-search route;

[0053] ζ is the initial deviation pheromone decay coefficient;

[0054] The set of nodes on a locally optimal network route;

[0055] Q(j): represents the sum of Manhattan distances between node j and every node on the locally optimal network route; in this embodiment, the distance between nodes is the Manhattan distance, calculated as follows:

[0056]

[0057] In the formula:

[0058] (x j ,y j (x) represents the coordinates of point j; j' y j' ) represents the coordinates of point j'.

[0059] The locally optimal network route can be calculated from the network topology using existing algorithms. This embodiment uses the LPA* heuristic search algorithm, but this invention does not limit it.

[0060] In this embodiment, taking a closed fiber optic planning area as an example, a grid method is used to transform the network topology map into a grid map. The improved ant colony algorithm performs route search based on this grid map. The grid map construction method includes:

[0061] Given a network with N nodes and M links, and a raster map of R×R rows where each raster cell is a square with a side length of 1, then:

[0062] The total number of raster cells on the raster map is R×R. Each raster cell on the raster map can be represented by a two-dimensional coordinate (i,j), where i = 1, 2, ..., R, j = 1, 2, ..., R. Each raster cell on the raster map can be represented by a unique number B, where B = 1, 2, ..., R×R. The relationship between the number B and the coordinate (i,j) is as follows:

[0063] B = (i-1) × R + j

[0064] Equation (6)

[0065] Each grid cell on the raster map has a label t(b), where t(b) ∈ {0, 1, ..., R × R}. t(b) represents the node or link number in the fiber optic network corresponding to that grid cell, or it may be an obstacle point. Specifically:

[0066] If t(B) = 0, it means that the grid cell is an obstacle point, that is, it does not correspond to any node or link.

[0067] If t(B)∈{1,…,R×R}, then it means that the grid cell corresponds to the node or link numbered t(k) in the optical fiber network.

[0068] In this context, a node is a network device (such as an optical distribution box), a link is a "connection" between adjacent network devices, and an obstacle is a point on the network that cannot be traversed.

[0069] It should be noted that the map in this embodiment is an R×R square grid map. In fact, a non-square map can be made into a square grid map by adding grid cells. The newly added grid cells are set as obstacle points.

[0070] Specifically, based on the initial pheromone concentration, the ant colony completes the first round of searching; as follows Figure 2 In the grid map shown, optical crossover 1 (7, 0) is the starting point and optical crossover 20 (20, 16) is the ending point. The locally optimal network route is as follows: Figure 2As shown by the underlined lines, the node coordinates on this better initial route are shown in Table 1 below;

[0071]

[0072] Table 1

[0073] The relevant parameters of the ant colony algorithm are set according to the actual situation of the current network planning example, as shown in Table 2 below:

[0074]

[0075] Table 2

[0076] S2. Calculate the pheromone increment of the route where the network node is located based on the total loss value; the total loss value is used to represent the number of network devices on the network route traversed by the ant, the total load of the nodes traversed in one route search, and the total route length; the pheromone increment is used to calculate the pheromone concentration between nodes; the calculated pheromone concentration is used to update the transition probability between nodes according to equation (1); the search is performed according to the new transition probability in the next iteration;

[0077] In this embodiment, the pheromone matrix is ​​updated after the first round of 50 ants' search iterations. The pheromone increment is calculated using the ant-week system model, and the pheromone left over from the previous iteration is volatilized. After each iteration, the pheromone is updated according to the following formula:

[0078] τ ij (t+1)=(1-ρ(t))τ ij (t)+Δτ ij (t) Equation (7)

[0079]

[0080] In the formula:

[0081] ρ(t): The pheromone evaporation coefficient at the t-th iteration; in this embodiment, when t=0, the value of ρ is preset, and its value is related to ρ max equal;

[0082] Δτ ij (t): The pheromone increment from route i to j at the t-th iteration, where i and j are network nodes;

[0083] N: represents the total number of ants;

[0084] In the t-th iteration, the pheromone left by the k-th ant when routing from i to j, where i and j are network grid nodes;

[0085] The increment of pheromones Δτ ijThe calculation of (t) adopts the ant-cycle system model. At the same time, in order to address the practical problems encountered in fiber optic planning, a multi-factor loss function is introduced as a constraint on the change of pheromone, so that the search algorithm can take into account more factors such as cost and efficiency that are important in fiber optic route planning during iteration.

[0086] The expression for the pheromone increment per ant is as follows:

[0087]

[0088] Q: The total amount of pheromones carried by a single ant;

[0089] cost k (t): The total loss value considering various cost factors after the route taken by the kth ant in the t-th iteration;

[0090] To more accurately address practical problems encountered in fiber optic routing planning, a multi-factor loss function is established to describe the actual route length cost, node access load cost, and the number of optical distribution boxes used for route access. This transforms the routing planning problem into a multi-objective optimization problem, using these multi-factor constraints to find the optimal route that meets the requirements of network routing planning. The multi-factor loss function is cost. k The definition of (t) is shown in the following equation:

[0091] cost k (t)=a1×M k (t)+a2×Q k (t)+a3×L k (t) Equation (10)

[0092] In the formula:

[0093] cost k (t): The multi-factor loss value of the route taken by the kth ant in the t-th iteration;

[0094] M k (t): The number of optical distribution box nodes that the k-th ant passes through when it successfully completes a route search in the t-th iteration;

[0095] a1, a2, a3: Set weighting coefficients;

[0096] Q k (t): The total load of the optical distribution box nodes traversed by the k-th ant in one successful route search at the t-th iteration;

[0097] The calculation method is as follows:

[0098]

[0099] m: The total number of optical distribution box nodes in the network planning;

[0100] The current set of routing nodes successfully searched by Ant.

[0101] b q Additional load factor for routing; that is, the fixed load cost of routing connectivity, which is set according to actual needs;

[0102] L k (t): The total length of the route traversed by the k-th ant in a single route search during the t-th iteration, expressed mathematically as follows:

[0103]

[0104] Here, a multi-factor loss function, cost, is constructed. k (t) serves as a constraint on the change in pheromone, enabling the search algorithm to balance factors such as cost and efficiency during fiber optic route planning during iteration. This allows for the determination of a multi-faceted optimized fiber optic resource planning route that is shortest, requires fewer traversed nodes, and has a relatively low routing load.

[0105] Preferably, by combining the definition of the multi-factor loss function, the expression for the pheromone increment in the ant colony algorithm is improved, as shown in the following equation:

[0106]

[0107] In the formula:

[0108] Q represents the total amount of pheromones carried by a single ant;

[0109] cost min The lowest historical loss value since the iteration.

[0110] In another embodiment, considering the pheromone concentration τ in the experiment ij Both excessively high and low pheromone concentrations can lead to significant numerical errors in the calculation, potentially causing the iterative process to fail to converge and resulting in unreliable results. Therefore, in iterative algorithms, it is crucial to avoid large differences in pheromone values ​​and strive to keep them within the same order of magnitude to improve computational accuracy and efficiency. Preferably, an upper and lower limit for pheromone concentration is set for each routing node, as shown in the following formula:

[0111]

[0112] Preferably, τ max =C0×1.5, which can prevent the situation from becoming more likely to get trapped in local optima due to excessively high pheromone levels on some nodes or links after iteration; where C0 is the preset initial value of pheromone concentration;

[0113] Preferably, τ min =0.1 / r, which can prevent the model from losing its generalization search ability due to the low pheromone of some rarely visited nodes or links; where r represents the maximum number of iterations.

[0114] Based on general experience with the volatilization of chemical substances on the ground, the volatilization rate of pheromones should meet the following conditions: First, the higher the pheromone concentration, the greater the volatilization rate. This is reflected in the algorithm as the ratio of the maximum pheromone concentration to the upper limit of pheromone concentration, max(τ). ij (t)) / τ max The larger the value, the greater the volatile factor.

[0115] Second, the larger the area of ​​pheromone spread, the faster it evaporates; that is, the ratio of the successful route length searched by the ant to its maximum mobility, max(L... k The larger (t)) / 3*R is, the greater the volatility factor;

[0116] Third, the longer the pheromone stays, the faster it evaporates. That is, the larger the ratio (rt) / r of the current iteration number to the maximum iteration number, the larger the evaporation factor.

[0117] In another embodiment, considering that the adaptive volatile factor decreases continuously with the iterative changes of multiple factors in the experiment, it is also necessary to set an upper and lower limit for the volatile factor to prevent the algorithm from stalling due to the volatile factor being too small. The mathematical expression of the improved adaptive volatile factor is as follows:

[0118]

[0119] In the formula:

[0120] r: Maximum number of iterations;

[0121] R: The maximum length or width of the current network topology graph after it has been converted into a raster map;

[0122] max(L k (t)): The maximum total length of the route traversed by each ant to successfully complete one route search at the t-th iteration;

[0123] max(τ ij (t)): The maximum value in the pheromone matrix updated after each ant successfully completes a route search in the t-th iteration;

[0124] ρ min ρ max : Minimum and maximum values ​​of the volatility coefficient; ρ was determined experimentally. max , ρ min The search results are optimal when the values ​​are set to 0.9 and 0.5 respectively.

[0125] τ max : The maximum value of the set pheromone concentration;

[0126] b: The first set value, obtained by adjusting parameters according to specific experiments. The optimal value of b is 0.3.

[0127] S3. When the number of searches reaches a set limit or after multiple consecutive searches all ants have the same search route, then the globally optimal network route from the starting node to the target node is obtained, such as... Figure 3 As shown, each gray square represents a light distribution box station; otherwise, the pheromone between nodes is updated according to the pheromone increment and the pheromone between nodes in the previous iteration using equations (7) and (14) respectively. The transition probability is calculated according to the updated pheromone between nodes using equation (1). All ants are set at the starting node. The next network node is selected according to the updated transition probability. When all ants have moved from the starting node to the target node, a new round of search is completed, and the process returns to step S2.

[0128] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0129] This application also provides a network routing planning system based on an improved ant colony algorithm, including a transition probability calculation module, an ant colony initialization module, a pheromone increment calculation module, a transition probability update module, and an optimal network route judgment module.

[0130] The transition probability calculation module is used to calculate the transition probability of each ant moving from the current network node to every possible next network node based on the distance between two network nodes, the distance between the network node and the target node, and the load of the network node.

[0131] The ant colony initialization module is used to initialize all ants in the ant colony algorithm to start from the starting node of the network to be planned; select the next network node according to the transition probability and the set pheromone concentration between nodes, until the first search is completed; the pheromone concentration between network nodes is used to guide the ants to search for network routes;

[0132] The pheromone increment calculation module is used to calculate the pheromone increment of the route where the network node is located based on the total loss value; its calculation method is shown in equation (9) or (13), where equation (13) is a better calculation method.

[0133] The transition probability update module is used to update the pheromone between nodes based on the pheromone increment and the pheromone between nodes in the previous iteration, and to calculate the transition probability based on the updated pheromone between nodes; the calculation method is shown in Equation (1).

[0134] The optimal network route determination module is used to determine whether the currently obtained route is the optimal network route. When the number of searches reaches the set limit or after multiple consecutive searches, all ants have the same search route, then the globally optimal network route from the starting node to the target node is obtained.

[0135] When the current route is not the optimal network route, the transition probability update module is called to update the transition probabilities between nodes. All ants start from the starting node of the network with the route to be planned and proceed to the next round of iteration based on the transition probabilities.

[0136] In another embodiment, an initial pheromone concentration calculation module is also included, which is used to set the initial pheromone concentration between nodes based on the nodes and links on the locally optimal network route. The initial pheromone concentration is the pheromone concentration between each node during the initial iteration, and its calculation method is shown in Equation (4).

[0137] In another embodiment, the pheromone increment calculation module further includes a total loss value calculation module, which is used to calculate the network congestion level of all nodes of the route traversed in each iteration, the route length, and the number of nodes where the network device is located. The calculation method is shown in Equation (10).

[0138] The contents not described in detail in this specification are existing technologies known to those skilled in the art.

Claims

1. A network routing planning method based on an improved ant colony algorithm, characterized in that, The steps include the following: S1. Calculate the transition probability of each ant moving from the current network node to every possible next network node based on the distance between two network nodes, the distance between the network node and the target node, and the load of the network node; the load of the network node is used to represent the network congestion degree of the node. In the initialization of the ant colony algorithm, all ants start from the starting node of the network to be planned; the next network node is selected according to the transition probability until the first search is completed; S2. Calculate the pheromone increment of the route where the network node is located based on the total loss value; the total loss value is used to represent the number of network service nodes on the network route traversed by the ant, the total load of the nodes traversed in one route search, and the total route length; the pheromone increment is used to calculate the transition probability; S3. When the number of searches reaches the set limit or after multiple consecutive searches all ants have the same search route, the globally optimal network route from the starting node to the target node is obtained. Otherwise, update the pheromone between nodes based on the pheromone increment and the pheromone between nodes in the previous iteration, calculate the transition probability based on the updated pheromone between nodes, start a new round of search based on the transition probability, and return to step S2 when the new round of search is completed. The total loss value cost k The methods for calculating (t) include: ; In the formula: M k (t): The number of network service nodes that the k-th ant passes through when it successfully completes a route search in the t-th iteration; a1, a2, a3: Set weighting coefficients; Q k (t): The total load of the network service nodes traversed by the k-th ant in one successful route search at the t-th iteration; L k (t): The total length of the route traversed by the k-th ant when it successfully completes a route search in the t-th iteration.

2. The network routing planning method based on the improved ant colony algorithm as described in claim 1, characterized in that, Before step S1, the method further includes: setting the initial pheromone concentration between nodes based on the nodes and links on the locally optimal network route; the calculation method for the initial pheromone concentration includes: ; In the formula: The initial pheromone concentration from node i to node j during the 0th iteration; C 0: The initial pheromone concentration on the locally optimal network route; : The initial deviation pheromone decay coefficient set; The set of nodes on the locally optimal network route; Q ( j ): Represents the sum of Manhattan distances between node j and every node on the locally optimal network route.

3. The network routing planning method based on the improved ant colony algorithm as described in claim 1 or 2, characterized in that, In step S1, the method for calculating the transition probability includes: ; ; D i,j = d i,j + d i,goal ; In the formula: (t): represents the probability that the k-th ant moves from node i to node j in the t-th iteration; The pheromone concentration of route ij at the t-th iteration, where i and j are network nodes; : Indicates the pheromone concentration of route is at the t-th iteration; The heuristic function value from node i to node j at the t-th iteration; The heuristic function value from node i to node s at the t-th iteration; α: Pheromone importance coefficient; β: Distance importance coefficient; v k : The set of all possible nodes that ant k can choose next when it is at node i; load j : Represents the load value of node j; the load value is used to indicate the degree of network congestion of the node; d i,j The distance between node i and node j; d i,goal : The distance between node i and the target site goal.

4. The network routing planning method based on the improved ant colony algorithm as described in claim 1, characterized in that, In step S3, the method for calculating the pheromone increment between nodes includes: ; ; In the formula: The pheromone increment from route i to j during the t-th iteration, where i and j are network nodes; : The pheromone left by the k-th ant when routing from i to j during the t-th iteration, where i and j are network nodes; Q: The total amount of pheromones carried by a single ant; cost min The lowest historical loss value to date; cost k (t): The total loss value after considering various cost factors after the route taken by the k-th ant in the t-th iteration.

5. The network routing planning method based on the improved ant colony algorithm as described in claim 1, characterized in that, In step S3, the method for updating pheromones between nodes includes: ; ; In the formula: : The pheromone concentration of route ij at the t-th iteration, where i and j are network nodes; ρ(t): The pheromone evaporation coefficient at the t-th iteration; The pheromone increment from route i to j during the t-th iteration, where i and j are network nodes; : The pheromone left by the k-th ant when routing from i to j during the t-th iteration, where i and j are network nodes.

6. The network routing planning method based on the improved ant colony algorithm as described in claim 5, characterized in that, The methods for calculating the volatile coefficient ρ(t) of pheromones include: ; In the formula: r Maximum number of iterations; R : The maximum length or width of the current network topology map after it has been converted into a raster map; max ( L k ( t ): The maximum total length of the route traversed by each ant to successfully complete one route search in the t-th iteration; The maximum value in the pheromone matrix updated after each ant successfully completes a route search in the t-th iteration; ρ min ρ max : Minimum and maximum values ​​of the volatility coefficient; : The maximum value of the set pheromone concentration; b: First set value.

7. A network routing planning system based on an improved ant colony algorithm, which executes the network routing planning method based on the improved ant colony algorithm as described in claim 1, characterized in that, It includes a transition probability calculation module, an ant colony initialization module, a pheromone increment calculation module, a transition probability update module, and an optimal network route determination module; The transfer probability calculation module is used to calculate the transfer probability of each ant moving from the current network node to each possible next network node based on the distance between two network nodes, the distance between the network node and the target node, and the load of the network node. The ant colony initialization module is used to initialize all ants in the ant colony algorithm to start from the starting node of the network to be planned; select the next network node according to the transition probability and the set pheromone concentration between nodes, until the first search is completed; the pheromone concentration between network nodes is used to guide the ants to search for network routes; The pheromone increment calculation module is used to calculate the pheromone increment of the route where the network node is located based on the total loss value. The transition probability update module is used to update the pheromone between nodes based on the pheromone increment and the pheromone between nodes in the previous iteration, and to calculate the transition probability based on the updated pheromone between nodes. The optimal network route determination module is used to determine whether the currently obtained route is the optimal network route. When the number of searches reaches the set limit or after multiple consecutive searches, all ants have the same search route, then the globally optimal network route from the starting node to the target node is obtained.

8. The network routing planning system based on the improved ant colony algorithm as described in claim 7, characterized in that, It includes an initial pheromone concentration calculation module, which is used to set the initial pheromone concentration between nodes based on the nodes and links on the locally optimal network route. The initial pheromone concentration is the pheromone concentration between each node during the initial iteration.

9. The network routing planning system based on the improved ant colony algorithm as described in claim 7, characterized in that, The pheromone increment calculation module also includes a total loss value calculation module, which is used to calculate the network congestion level, route length, and number of nodes where network services are located for all nodes traversed in each iteration.

Images