Fish school-imitating random routing method for quantum key distribution network

By abstracting the quantum key routing process into fish school behavior patterns, combining key resources and node credibility, optimizing routing selection, the problems of low transmission efficiency and insufficient security in the QKD network are solved, and more efficient and secure are achieved Key distribution.

CN119071215APending Publication Date: 2024-12-03XIAN UNIV OF POSTS & TELECOMM
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202411053924.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-02
Publication Date
2024-12-03

Smart Images

  • Figure CN119071215A_ABST
    Figure CN119071215A_ABST
Patent Text Reader

Abstract

The invention relates to a fish school-imitating random routing method for a quantum key distribution network. The method comprises the following steps: abstracting a routing process of a global quantum key into an optimizing behavior of a fish school according to a key transmission principle and a routing transmission architecture model of a partially credible relay quantum key distribution network under the inspiration of a self-organizing behavior of the fish school, and constructing a fish school behavior imitating mode under a key transmission mechanism; wherein the fish school behavior simulation mode comprises an identification behavior, a foraging behavior, a gathering behavior and a mark diffusion behavior; establishing a fish shoal perception random routing model through a fish shoal behavior simulation mode; and solving the fish swarm perception random routing model to deploy network resources, thereby realizing balanced deployment of quantum key distribution network resources. According to the invention, key resources can be balanced, the service blocking rate is reduced, and the overall security of the quantum key distribution network is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of quantum communication technology, and in particular to a random routing method for simulating fish swarms in a quantum key distribution network. Background Art

[0002] Quantum key distribution (QKD) is a technology that uses the basic principles of quantum mechanics to provide theoretically unconditionally secure keys to both communicating parties. Due to the attenuation characteristics of single photons, the transmission distance of quantum keys is limited, and the emergence of quantum key distribution networks makes it possible to transmit quantum keys over long distances. At present, the main QKD networking methods in the world include: trusted relay QKD network, QKD network independent of measurement equipment, and quantum relay QKD network. Considering the security risks of malicious tampering and unauthorized access to relay nodes, some trusted relay networking methods that are more in line with actual scenarios have gradually become the preferred solution for the practicalization and popularization of QKD networks. However, in some trusted relay QKD networking methods, the transmission of quantum keys involves routing selection issues.

[0003] In related technologies, introducing random strategies into single-path or alternative multi-path routing methods can effectively improve security. The distance between nodes is used as a measure to prioritize neighboring nodes close to the target node for random routing selection. Although this method can reduce the consumption of key resources to a certain extent, it still has problems such as low transmission efficiency, low security, and network blocking probability.

[0004] Therefore, it is necessary to provide a new technical solution to improve one or more problems existing in the above solutions.

[0005] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present application, and therefore may include information that does not constitute the prior art known to ordinary technicians in the field. Summary of the invention

[0006] The purpose of this application is to provide a quantum key distribution network fish swarm simulation random routing method, thereby at least to some extent overcoming one or more problems caused by the limitations and defects of related technologies.

[0007] According to a first aspect of an embodiment of the present application, a quantum key distribution network fish swarm-like random routing method is provided, the method comprising:

[0008] According to the key transmission principle and routing transmission architecture model of the partially trusted relay quantum key distribution network, the routing process of the global quantum key is abstracted as the optimization behavior of a school of fish, and a school of fish-like behavior model under the key transmission mechanism is constructed; wherein the school of fish-like behavior model includes recognition behavior, foraging behavior, aggregation behavior and mark diffusion behavior;

[0009] According to the simulated fish school behavior pattern, a fish school perception random routing model is established;

[0010] The fish school sensing random routing model is solved to allocate network resources.

[0011] In an embodiment of the present application, the method of establishing a fish school perception random routing model according to the simulated fish school behavior pattern includes:

[0012] Before route selection, the controller obtains the remaining key of the link passing through the next-hop neighbor node link by querying the status information of each neighbor node in the routing table, and identifies the neighbor node through the path threshold, where:

[0013]

[0014] μ=v×T r (2);

[0015] In the formula, e k represents the neighbor node link, represents the remaining key amount of the link, μ represents the path threshold, k0 represents the initial key amount of the key pool, represents the quantum key generation rate, T r Indicates the service time slot requirement, T k Indicates the key update period, v r represents the key consumption rate, v represents the key consumption rate per service unit time, R represents the key service request set, and r represents the service;

[0016] Introduce the distance evaluation factor and sort the k neighbor nodes according to the distance d i Arrange in descending order, the expression of the distance evaluation factor is as follows:

[0017]

[0018] Where W j represents the distance evaluation factor,

[0019] In an embodiment of the present application, the method of establishing a fish school perception random routing model according to the simulated fish school behavior pattern includes:

[0020] The routing cost is evaluated from the perspectives of key resource availability and node credibility;

[0021] The evaluation of the routing cost from two aspects of key resource availability and node credibility includes:

[0022] The quantum key generation rate and key freshness constitute the link resource cost:

[0023] C r =aC KGR +bC KF (4);

[0024]

[0025] In the formula, C r represents the link resource cost, C KGR represents the quantum key generation rate, C KF represents the key freshness; and is the capacity of the key pool at both ends of the link, where a and b represent the key resource influence weights, satisfying a+b=1, a=2 / 3, b=1 / 3;

[0026] The expression of the routing trust cost is:

[0027]

[0028] In the formula, C T represents the trusted cost of routing, N p represents the set of candidate food nodes, v i represents the routing relay node, l(v i ) represents a trusted Boolean value;

[0029] By querying N p The trusted Boolean value of the relay node in the middle route is used to calculate the security cost of selecting the current node route, and the trusted node is added to the trusted node set, otherwise it is added to the untrusted node set.

[0030] In an embodiment of the present application, the method of establishing a fish school perception random routing model according to the simulated fish school behavior pattern includes:

[0031] The node food concentration is calculated for the candidate neighbor nodes in the trusted node set and the untrusted node set respectively. The expression of the node food concentration is as follows:

[0032]

[0033] In the formula, η(v i ) represents the food concentration of the node, α represents the distance weight coefficient, β represents the route evaluation cost weight coefficient, |α|+|β|=1;

[0034] The node food concentration is normalized to obtain the normalized node food concentration, and then the routing foraging behavior is performed, and the neighboring nodes are randomly selected by probability; wherein the expressions of the probabilities are as follows:

[0035] η * (v i )=x p1 +x p2 (9);

[0036]

[0037] Where η * (v i ) represents the normalized node food concentration, x p1 represents the normalized eigenvalue of the distance evaluation factor matrix, x p2 represents the normalized eigenvalue of the routing cost matrix, P(v i ) represents probability.

[0038] In an embodiment of the present application, solving the fish school sensing random routing model to allocate network resources includes:

[0039] Initialize the quantum key distribution network, collect network topology information, key pool resource information, node information, and add it to the routing table in the form of data packets;

[0040] When both communicating parties receive the task request of forwarding the key, the fish swarm abstracts the routing process of the global quantum key into the message of the fish swarm's optimization behavior and performs unified routing calculation on the key relay task through the server;

[0041] Traverse the neighboring nodes of the next hop and calculate the remaining key amount of the link between the node and each neighboring node and link threshold μ;

[0042] like Then add the neighbor node to the candidate food node set N p ;

[0043] Judge N p The number of neighbor nodes described in;

[0044] If N p If there are multiple neighbor nodes, the congestion factor δ is set according to the network scale, and N p Number of markings of neighbor nodes in i ;

[0045] If i >δ, exclude it N p ;

[0046] Update Np , calculate the node food concentration η(v i ) and normalize it;

[0047] Then with probability P(v i ) Perform routing selection and update the mark number n i =n i +1, and use the link key in the quantum key pool to the global key K c Perform encryption operations and update network resources;

[0048] If K c When relayed to the destination node, the quantum key terminal connection between both parties is released.

[0049] In the embodiment of the present application, the determination of N p The number of neighbor nodes in the above-mentioned embodiment also includes:

[0050] If there is a neighbor node, the link key in the quantum key pool is used to generate the global key K c Perform encryption operations and update network resources.

[0051] In the embodiment of the present application, the determination of N p The number of neighbor nodes in the above-mentioned embodiment also includes:

[0052] If the neighbor node does not exist, the call loss is recorded.

[0053] In the embodiment of the present application, if N p If there are multiple neighbor nodes, the congestion factor δ is set according to the network scale, and N p Number of markings of neighbor nodes in i , also includes:

[0054] If i ≤δ, then calculate the node food concentration η(v i ) and normalize it.

[0055] In the embodiment of the present application, the neighboring nodes of the next hop are traversed to calculate the remaining key amount of the link between the node and each neighboring node. With the link threshold μ, it also includes:

[0056] like Then the neighbor node does not participate in routing selection.

[0057] The technical solution provided by the embodiments of the present application may have the following beneficial effects:

[0058] In one embodiment of the present application, through the above method, combined with the key transmission principle and routing transmission architecture model of the partially trusted relay quantum key distribution network, the routing process of the global quantum key is abstracted as the optimization behavior of the fish school, and the fish school behavior model under the key transmission mechanism is constructed. The fish school-perceived random routing model is established through the fish school behavior model, and the fish school-perceived random routing model is solved to allocate network resources. This method balances key resources to a certain extent, reduces the service blocking rate, and improves the overall security of the quantum key distribution network.

[0059] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] The drawings herein are incorporated into the specification and constitute a part of the specification, illustrate embodiments consistent with the present application, and together with the specification are used to explain the principles of the present application. Obviously, the drawings described below are only some embodiments of the present application, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0061] Figure 1 A flowchart schematically showing a method for random routing of a quantum key distribution network simulating a fish swarm in an exemplary embodiment of the present application;

[0062] Figure 2 The schematic diagram of the transmission principle of a partially trusted relay key in an exemplary embodiment of the present application is shown;

[0063] Figure 3 Schematically showing a partially trusted relay QKD transmission architecture diagram in an exemplary embodiment of the present application;

[0064] Figure 4 The self-organizing behavior of a school of fish in nature in an exemplary embodiment of the present application is schematically shown;

[0065] Figure 5 A flowchart schematically illustrates the steps of solving a fish school sensing random routing model in an exemplary embodiment of the present application;

[0066] Figure 6 A schematic diagram of a mesh topology structure of a partially trusted relay QKD network in an exemplary embodiment of the present application is shown;

[0067] Figure 7 Schematically showing a comparison diagram of routing time of various methods in an exemplary embodiment of the present application;

[0068] Figure 8 A schematic diagram showing a comparison of routing costs of various methods in an exemplary embodiment of the present application is shown;

[0069] Fig. 9 A schematic diagram showing a comparison of the average number of routing hops and key consumption of each method in an exemplary embodiment of the present application;

[0070] Fig.10 Schematically showing the average key consumption of different QKD routing methods in an exemplary embodiment of the present application;

[0071] Fig.11 A diagram schematically showing a comparison of the key distribution success rates of various methods under different numbers of services in an exemplary embodiment of the present application;

[0072] Fig.12 Schematically illustrates the overall security probability of each algorithm under the medium attack mode in the exemplary embodiment of the present application;

[0073] Fig.13 Schematically illustrating the key leakage probability of each algorithm in the fixed node attack mode in the exemplary embodiment of the present application;

[0074] Fig.14 Schematically illustrates the introduction of marker diffusion behavior in an exemplary embodiment of the present application;

[0075] Fig.15 It is schematically shown that no mark diffusion behavior is introduced in the exemplary embodiment of the present application;

[0076] Fig.16 A diagram schematically shows a comparison of link key consumption trends during Random algorithm routing in an exemplary embodiment of the present application;

[0077] Fig.17 A comparison diagram of link key consumption trends during routing using the RIP-SRA method in an exemplary embodiment of the present application is schematically shown;

[0078] Fig.18 A comparison diagram of link key consumption trends during routing in the method in an exemplary embodiment of the present application is schematically shown;

[0079] Fig.19 A comparison diagram of link key consumption trends during routing using the MP-GA method in an exemplary embodiment of the present application is schematically shown;

[0080] Fig. 20 A comparison diagram of link key consumption trends during routing using the LCN-MPA method in an exemplary embodiment of the present application is schematically shown. DETAILED DESCRIPTION

[0081] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this application will be more comprehensive and complete and fully convey the concepts of the example embodiments to those skilled in the art. The described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.

[0082] In addition, the accompanying drawings are only schematic illustrations of the present application and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and thus their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor devices and / or microcontroller devices.

[0083] This example implementation provides a quantum key distribution network fish swarm-like random routing method. Figure 1 As shown in , the method includes: step S101 to step S103.

[0084] Among them, step S101: according to the key transmission principle and routing transmission architecture model of the partially trusted relay quantum key distribution network, the routing process of the global quantum key is abstracted as the optimization behavior of a school of fish, and a school of fish-like behavior pattern under the key transmission mechanism is constructed; wherein the school of fish-like behavior pattern includes recognition behavior, foraging behavior, aggregation behavior and mark diffusion behavior.

[0085] Step S102: establishing a fish school perception random routing model according to the simulated fish school behavior pattern.

[0086] Step S103: Solve the fish school sensing random routing model to allocate network resources.

[0087] In one embodiment of the present application, through the above method, combined with the key transmission principle and routing transmission architecture model of the partially trusted relay quantum key distribution network, the routing process of the global quantum key is abstracted as the optimization behavior of the fish school, and the fish school behavior model under the key transmission mechanism is constructed. The fish school-perceived random routing model is established through the fish school behavior model, and the fish school-perceived random routing model is solved to allocate network resources. This method balances key resources to a certain extent, reduces the service blocking rate, and improves the overall security of the quantum key distribution network.

[0088] Next, we will refer to Figures 1 to 4 Each step of the above method in this example implementation is described in more detail.

[0089] In step S101, according to the key transmission principle and routing transmission architecture model of the partially trusted relay quantum key distribution network, the routing process of the global quantum key is abstracted as the optimization behavior of a school of fish, and a simulated school of fish behavior model under the key transmission mechanism is constructed.

[0090] Among them, the key transmission principle of the partially trusted relay quantum key distribution network includes the following:

[0091] The main function of the partially trusted relay-based quantum key distribution network is to achieve end-to-end key transmission between non-adjacent nodes. Figure 2 As shown, suppose the sender Alice needs to send the original session key K c Transmitted to the receiver Bob to achieve the purpose of sharing the quantum key, the key transmission path is as follows:

[0092] Alice→QRN1→QRN2→QRN3→Bob→QRN2→QRN3→Bob, in the key transmission path, it needs to pass through three quantum relay nodes (QRN), each relay node is composed of quantum devices, optical devices and quantum key pools, and several quantum channels and classical channels are required between nodes to ensure the distribution of keys. Among them, QRN1, QRN2 and QRN3 are all relay nodes.

[0093] The specific key relay process is as follows:

[0094] (1) Alice and QRN1 share the global key K c QRN1 and QRN2 generate link key K1 through quantum key protocol. c Then, the link key K1 stored in the quantum key pool is used to perform a one-time pad (OTP) to secure K1. c Encryption is performed;

[0095] (2) Using XOR encryption algorithm K c Encryption, the encrypted K c1 Transmitted to QRN2, then QRN2 uses Decrypt in the same way and recover the global key K c .

[0096] (3) In the same way, QRN2 uses link key K2 to c Perform the encryption operation and decrypt at QRN3 to recover the global key K c . Through the quantum channel, K c Transmitted to Bob, both parties use K cConduct confidential communications.

[0097] The routing transmission architecture model of the partially trusted relay quantum key distribution network includes the following:

[0098] like Figure 3 As shown in the figure, the routing transmission architecture model of the partially trusted relay QKD network adopts a 4-layer architecture. The quantum layer is the core layer of the QKD network, responsible for creating and transmitting quantum keys, and sending the generated keys to the key management layer; the key management layer is responsible for managing and distributing the stored keys, and collecting and coordinating the status information of each node and link through the key management server (KMS); the key provision layer is responsible for delivering the generated quantum keys to users, and the SDN controller performs centralized routing calculations and sends the calculation results to each quantum node; the application layer is responsible for realizing confidential communication services between users.

[0099] Different from the trusted relay QKD network model, the routing transmission architecture model of this part of the trusted relay QKD network is composed of trusted and untrusted nodes in a certain proportion in the quantum layer. Among them, the trusted nodes have higher security and credibility, while the untrusted nodes refer to the nodes whose relay devices are vulnerable to physical attacks by eavesdroppers due to improper placement or poor protection. When the key is relayed under such nodes, there is a risk of key leakage.

[0100] In nature, some fish often live in groups. Individuals in the school will not only actively adjust their positions in the school, but also coordinate with other members to make intelligent clustering behaviors to change the shape of the school. This self-organizing characteristic helps to improve the overall foraging efficiency and survival of the school. Inspired by the self-organizing behavior of fish schools, this application uses the artificial fish school algorithm to imitate the intelligent behavior of fish to quickly obtain the optimal solution as a theoretical basis. Combined with the characteristics of the scarcity of quantum key resources, the routing process of the global quantum key is abstracted as the optimization behavior of the fish school, and this process is recorded as FS (fish selection). FS is integrated in the SDN server to be responsible for centralized routing calculations and to uniformly allocate network resources.

[0101] The fish school behavior patterns under the key transmission mechanism constructed in the present application include recognition behavior, foraging behavior, aggregation behavior and mark diffusion behavior.

[0102] Among them, the identification behavior includes: the ability of fish to perceive objects and make food judgments. In the QKD network, FS traverses the remaining key quantity in QKP, and screens out nodes that do not meet business requirements by setting a link threshold μ, otherwise it adds the candidate food node set N p ,This priority filtering mechanism is helpful to reduce the service blocking rate.,Wherein, the candidate food node set is also the candidate food node set.

[0103] Foraging behavior includes: fish instinctively choose areas with high food concentration to feed. Each node in the QKD network has its corresponding node food concentration η(V i ), which is determined by the distance factor and the cost function. The node food concentration is used as the routing evaluation index. FS gives priority to nodes with large node food concentration for routing. This optimization mechanism is conducive to improving the efficiency of key distribution.

[0104] Aggregation behaviors include: fish schools gather in areas with high node food concentration, which will intensify the competition for resources. The above foraging behavior will cause the business to be concentrated on nodes and links with high node food concentration. In order to avoid falling into local optimality and excessive resource consumption, random thinking is introduced. By normalizing the node food concentration, FS selects the routing of neighbor nodes in the form of probability. This randomness ensures routing optimization while reducing the probability of network congestion.

[0105] The tag diffusion behavior includes: adaptive behavior to avoid excessive competition within the fish school. When selecting a node, FS will first mark the node. When the subsequent business selects the node again, it will query the node's tag count n through the routing table. i ,If the number of marking reaches the congestion threshold δ, FS will select neighbor nodes with fewer markings for relaying. ,This diffusion mechanism balances the network load to a certain extent.

[0106] In step S102, a fish school perception random routing model is established according to the simulated fish school behavior pattern.

[0107] Before establishing a random routing model based on the simulated fish school behavior pattern, this application conducted corresponding research on the linear programming routing transmission architecture model based on the maximum food concentration. For the service request set R, its objective function is:

[0108]

[0109] The formula shows that within the constraints, the global key will select the node with the largest node food concentration for relaying. The node food concentration in the objective function is satisfy And get the following formula:

[0110]

[0111] In the formula, α represents the distance weight coefficient, represents the quantum key pool capacity, l i Represents the routing relay node v i The trusted Boolean value, β represents the routing cost weight coefficient, d i Represents the routing relay node v i The distance to the destination node;

[0112] The constraints of the objective function of the linear programming routing transmission architecture model based on maximum food concentration are as follows:

[0113]

[0114] 0≤|V t |≤|V| (15);

[0115] |E|≥|h| (16);

[0116]

[0117] Where, T k Represents the key update period, T r Business time slot requirement, T d represents the duration of the key request set, V t Represents the set of relay nodes participating in routing, v r represents the key consumption rate per business unit time, Indicates link e within service duration t k The amount of remaining keys on the quantum key distribution network, E represents the set of quantum key distribution network links, V represents the set of quantum key distribution network nodes, and h represents the set of optical paths for business routing selection.

[0118] Formula (13) is used to ensure that a single link e k The number of keys in the key pool is not less than the key consumption required for the key request; Formula (14) ensures that in time period T d The total number of keys required by the key request set R should not exceed the total number of keys of the QKD links in the topology; Formula (15) ensures that the number of nodes passed by the route should not exceed the total number of nodes in the topology; Formula (16) ensures that the number of links traversed by the key request set R should not exceed the total number of links; Formula (17) ensures that the number of links traversed by the set R should not exceed the product of the total number of links in the topology and the number of link sets traversed by a single service r.

[0119] Although the ILP routing transmission architecture model based on the maximum food concentration provides a minimum cost routing solution, its computational complexity is high and it does not have good scalability in large-scale OKD networks. Therefore, this application establishes a fish school perception random routing model.

[0120] In step S1021, a fish school perception random routing model is established according to the simulated fish school behavior pattern, including:

[0121] Before route selection, the controller obtains the remaining key of the link passing through the next-hop neighbor node link by querying the status information of each neighbor node in the routing table, and identifies the neighbor node through the path threshold, where:

[0122]

[0123] μ=v×T r (2);

[0124] In the formula, e k Indicates a link, represents the remaining key amount of the link, μ represents the path threshold, k0 represents the initial key amount of the key pool, represents the quantum key generation rate, T r Indicates the service time slot requirement, T k Indicates the key update period, v r represents the key consumption rate, v represents the key consumption rate per service unit time, R represents the key service request set, and r represents the service;

[0125] Introduce the distance evaluation factor and sort the k neighbor nodes according to the distance d i Arrange in descending order, the expression of the distance evaluation factor is as follows:

[0126]

[0127] Where W j represents the distance evaluation factor,

[0128] In step S1022, a fish school perception random routing model is established according to the simulated fish school behavior pattern, including:

[0129] The routing cost is evaluated from the perspectives of key resource availability and node credibility;

[0130] The routing cost is evaluated from two aspects: key resource availability and node credibility, including:

[0131] The quantum key generation rate and key freshness constitute the link resource cost:

[0132] C r =aC KGR +bC KF (4);

[0133]

[0134] In the formula, C r represents the link resource cost, C KGR represents the quantum key generation rate, C KFrepresents the key freshness; and is the capacity of the key pool at both ends of the link, where a and b represent the key resource influence weights, satisfying a+b=1, a=2 / 3, b=1 / 3;

[0135] The expression of the routing trust cost is:

[0136]

[0137] In the formula, C T represents the trusted cost of routing, N p represents the set of candidate food nodes, v i represents the routing relay node, l(v i ) represents a trusted Boolean value;

[0138] By querying N p The trusted Boolean value of the relay node in the middle route is used to calculate the security cost of selecting the current node route, and the trusted node is added to the trusted node set, otherwise it is added to the untrusted node set.

[0139] In step S1023, a fish school perception random routing model is established according to the simulated fish school behavior pattern, including:

[0140] The node food concentration is calculated for the candidate neighbor nodes in the trusted node set and the untrusted node set respectively. The expression of the node food concentration is as follows:

[0141]

[0142] In the formula, η(v i ) represents the food concentration of the node, α represents the distance weight coefficient, β represents the route evaluation cost weight coefficient, |α|+|β|=1.

[0143] Substituting formulas (5), (6) and (7) into formula (8), the calculation formula for node food concentration is as follows:

[0144]

[0145] In order to speed up the routing convergence, the node food concentration needs to be normalized and the distance evaluation factor W calculated by k neighbor nodes is j and routing cost function value C i =1 / (C r +2C T ) is recorded in the node food concentration parameter table, as shown in Table 1:

[0146] Table 1 Candidate food node set N p Food concentration parameter table

[0147]

[0148] According to the routing evaluation index W in Table 1 j and C i , establish the corresponding feature matrix F k×2 :

[0149]

[0150] In the formula, r k1 Represents the characteristic value of the distance evaluation factor, r k2 Indicates the routing cost characteristic value.

[0151] For the feature matrix F k×2 The routing feature values ​​in are normalized to obtain a normalized feature matrix. The matrix element in the first column after normalization is recorded as r * p1 :

[0152]

[0153] Similarly, we get the normalized matrix element r of the second column feature. * p2 , and the weight coefficients of the routing evaluation index in formula (8) (including the distance weight coefficient and the routing evaluation cost weight coefficient) are used to obtain the weighted food concentration normalized matrix. The element x in the normalized matrix pq It is expressed as:

[0154]

[0155] The node food concentration is normalized to obtain the normalized node food concentration, and then the routing foraging behavior is performed, and the neighboring nodes are randomly selected by probability; the expressions of the probabilities are as follows:

[0156] η * (v i )=x p1 +x p2 (9);

[0157]

[0158] Where η * (V i ) represents the normalized node food concentration, x p1 represents the normalized eigenvalue of the distance evaluation factor matrix, x p2 represents the normalized eigenvalue of the routing cost matrix, P(v i ) represents probability.

[0159] In step S103, the fish school sensing random routing model is solved to allocate network resources.

[0160] In step S1031, the fish school sensing random routing model is solved to allocate network resources, such as Figure 4 As shown, including:

[0161] Initialize the quantum key distribution network, collect network topology information, key pool resource information, node information, and add it to the routing table in the form of data packets;

[0162] When both communicating parties receive the task request of forwarding the key, the fish swarm abstracts the routing process of the global quantum key into the message of the fish swarm's optimization behavior and performs unified routing calculation on the key relay task through the server;

[0163] Traverse the next-hop neighbor nodes and calculate the remaining key amount of the link between the node and each neighbor node and link threshold μ;

[0164] like Then add the neighbor node to the candidate food node set N p ;

[0165] Judge N p The number of neighbor nodes in ;

[0166] If N p If there are multiple neighbor nodes in the network, the crowding factor δ is set according to the network scale, and N p Number of markings of neighbor nodes in i ;

[0167] If i >δ, exclude it N p ;

[0168] Update N p , calculate the node food concentration η(v i ) and normalize it;

[0169] Then with probability P(v i ) Perform routing selection and update the mark number n i =n i +1, and use the link key in the quantum key pool to the global key K c Perform encryption operations and update network resources;

[0170] If K c When relayed to the destination node, the quantum key terminal connection between both parties is released.

[0171] In step S1032, the link key in the quantum key pool is used to generate the global key Kc Perform encryption operations, update network resources, and then include:

[0172] If K c If the destination node is not reached, the data packet is sent and the routing table is updated.

[0173] In step S1033, the N p The number of neighbor nodes in the network also includes:

[0174] If there is a neighbor node, the link key in the quantum key pool is used to align the global key K c Perform encryption operations and update network resources.

[0175] In step S1034, the N p The number of neighbor nodes in the network also includes:

[0176] If there is no neighbor node, the call loss is recorded.

[0177] In step S1035, if N p If there are multiple neighbor nodes in the network, the crowding factor δ is set according to the network scale, and N p Number of markings of neighbor nodes in i , also includes:

[0178] If i ≤δ, then calculate the node food concentration η(v i ) and normalize it.

[0179] In step S1036, the next-hop neighbor nodes are traversed to calculate the remaining key amount of the link between the node and each neighbor node. With the link threshold μ, it also includes:

[0180] like Then the neighboring nodes do not participate in routing selection.

[0181] The present application is further described below through simulation and analysis.

[0182] In order to evaluate the performance of the QKD network routing method, according to the topological characteristics of the QKD backbone network and the metropolitan area network, this application uses a 14-node 21-link trusted / untrusted hybrid relay mesh topology for simulation experiments, such as Figure 6As shown in the figure, the physical length of each optical fiber link is marked in the network, in km. In order to accommodate each optical path connection request in the QKD network, it is assumed that there are enough wavelength channels on each optical fiber, and the low-loss c-band (1530-1565nm) is used to transmit quantum signals. A 200GHz protection bandwidth is set between the quantum channel and the classical channel to reduce the four-wave mixing effect, and the wavelength allocation adopts the First-Fit method.

[0183] In order to simulate the business scenarios in real situations, the arrival interval of business requests, key update rate and key quantity requirements are set to follow Poisson distribution within a certain range. The source and sink nodes in the network are trusted nodes by default, and other nodes select trusted / untrusted relays according to uniform distribution. The other parameter settings are shown in Table 2:

[0184] Table 2 Simulation parameter settings

[0185]

[0186]

[0187] Algorithm feasibility analysis

[0188] In order to verify the feasibility of the proposed heuristic method, the method was verified from three aspects: routing resource consumption level, routing cost and routing time. The performance difference between the fish school-aware random routing method and the ideal ILP routing transmission architecture model and the random routing benchmark method (Random algorithm) was analyzed.

[0189] To evaluate the resource consumption level of each routing solution, Figure 6 The relationship between the number of different service requests and the average number of hops and key consumption is simulated. From equations (2) and (10), it can be seen that the introduction of the distance evaluation factor effectively reduces the number of hops in ILP and the routing selection of this method, thereby reducing the key consumption during routing, while the Random algorithm randomly selects routes from all possible paths between the source node and the target node. Figure 7 Further verification shows that the average routing hops of the ILP model and this method are much smaller than that of the Random algorithm. When the business volume is 250, the method proposed in this application saves about 21.9% of the key consumption compared with the Random algorithm.

[0190] Figure 8 The relationship between the proportion of trusted nodes T and routing cost under each method is simulated. Figure 8It can be seen that when T is the same, the routing cost is ranked as: Random>this method>ILP. When the business volume is 200 and T=50%, the routing cost of this method is about 42.6% lower than that of the Random algorithm; when the business volume is 150 and T=25%, the routing cost of this method is 16.3% higher than that of the ILP model. This is because the ILP model is essentially a minimum cost routing strategy, and this method introduces a random strategy within the optimization range to avoid the algorithm from falling into the local optimum. Although the randomness of the routing balances the consumption of key resources, it increases the routing cost to a certain extent.

[0191] Fig. 9 The running time of each routing scheme under different numbers of business requests is simulated. From 9, it can be seen that the routing time of the ILP model is much longer than the Random algorithm and the fish school-aware random routing method proposed in this application. When the business volume is 200, the ILP model Figure 8 The running time on the topology is 448 seconds, while the running time of the Random algorithm and this method is 0.27 seconds and 0.63 seconds respectively. As the number of services increases and the network scale expands, the routing time will continue to increase. Therefore, the ILP model is not suitable for solving large-scale QKD network routing problems. This method effectively improves routing efficiency while ensuring routing performance.

[0192] Algorithm Comparative Analysis

[0193] In order to further verify the performance of the fish school-aware random routing method, this application conducts simulation tests from the aspects of key consumption, key distribution success rate, routing security and resource balance, and adopts a random routing benchmark method (Random), a multi-path routing benchmark method (MP-GA), an improved RIP random routing method (RIP-SRA) and a multi-path routing method based on the least common node (LCN-MPA) for comparative analysis.

[0194] Table 3 simulates the routing results of each routing method. It can be seen from Table 3 that the ranking of the routing hops of each method is: Random>this method>MP-GA=LCN-MPA=RIP-SRA. Among them, loops will appear when the Random algorithm is routed, and the MP-GA, LCN-MPA and RIP-SRA methods all find multiple shortest paths of 4 hops through the Dijkstra method, but the routing results are relatively fixed and are not suitable for the problem of only one shortest path between the source and the destination node. Although the average hop count of this algorithm is 4.6, which is slightly higher than the method with the shortest path of 4 hops, the routing of this method is more flexible, which reduces the occurrence of loops to a certain extent, and will not cause the problem of being unable to route due to network restrictions.

[0195] Table 3 Routing selection results of each QKD routing method

[0196]

[0197]

[0198] The key consumption reflects the degree of key resource consumption in the routing process, which is used to evaluate the energy consumption level of the routing method. The global key size is set to 100 bits, and the number of rerouting times n is set to 1. Fig.10 The key consumption of each algorithm is simulated when the key quantity is sufficient and the relay key quantity is insufficient in a single service. Fig.10 It can be seen that when the key quantity is sufficient, the average key consumption of this method is slightly higher than that of the RIP-SRA method, but lower than that of the other three methods; when the key quantity of 1-4 hops is insufficient, the average key consumption of this method is 480 bits, which is significantly less than that of other routing methods, and the key consumption is more stable than that of the RIP-SRA method. This is because this method excludes links with insufficient key quantity from the set of routing candidate food nodes, avoiding key waste caused by rerouting.

[0199] The success rate of quantum key distribution is an important indicator to measure the quality of QKD network routing methods. It represents the ratio of the number of services that have completed key distribution to the total number of services. Fig.11 The success rate of quantum key distribution of each method under different numbers of business requests was simulated. Fig.11 It shows that when the traffic volume is 20, the key distribution success rate of each method remains above 80%. When the traffic volume increases, each method will fail to distribute keys due to insufficient keys, among which the MP-GA method has the largest decrease. When the traffic volume is large, the key distribution success rate of this method is more obvious than other multi-path and random routing algorithms.

[0200] In a partially trusted relay QKD network, attackers will adopt different attack strategies to steal key information from untrusted nodes. To maximize the attack effect, attackers often use equal probability attack strategies when the topology is unknown, and then attack key nodes in a targeted manner after knowing the topology. Fig.12The routing security of each method under different attack modes was simulated when the proportion of topological trusted nodes T was 25%, and the overall routing security probability of different numbers of attackers attacking untrusted relay nodes with equal probability was simulated. When the number of attackers increased, the overall security probability of each algorithm showed a downward trend. When the number of attackers was the same, the security probability of this method was increased by about 8% and 2% compared with the Random and RIP-SRA methods, respectively, but the security of the MP-GA and LCN-MPA methods was significantly better than the other three random routing methods. This is because the multi-path method distributes the global key to multiple paths for transmission, and the key is completely leaked when all paths are intercepted, but the high security comes at the cost of increased key consumption. Among the random routing methods, this method has better security performance, which effectively improves routing security while reducing key consumption.

[0201] Fig.13 The key leakage probability of each method when the attacker adopts a fixed node attack strategy is simulated. As can be seen from the figure, the key leakage probability of the MP-GA and LCN-MPA methods is always 100%, and the key leakage probability of the RIP-SRA method is 88% when attacking nodes 5 and 12, and the other attack methods are also 100%. The routing strategies of the above methods are easy to expose the key to the attacker, and the security is almost completely invalid. The key leakage probability of this method is relatively low, and the key leakage probability is 78% when attacking nodes 2 and 5. The leakage probability of the other attack methods is maintained at around 65%. This is because the attacker must control all adjacent nodes near the source node to completely intercept the key information, and the routing security performance is guaranteed to a certain extent.

[0202] Link load uniformity (LLU) is a key indicator to measure the resource usage balance of routing methods. It refers to the standard deviation of the average consumption of key resources of each link per unit time. Fig.14 and Fig.15 The influence of the diffusion mechanism of this method on the balance of link keys was simulated. When the label diffusion behavior was not introduced, the corresponding LLU value in the link key pool was 408.6243, and the range of the remaining key amount was 1175 bits. After the introduction of this behavior, the LLU value became 270.2673, the range of the remaining key amount was reduced to 1048 bits, and the distribution of the remaining key of the link was more balanced. The diffusion mechanism of this method effectively balanced the use of key resources by reducing the probability of repeatedly selecting a link for relaying and reducing the routing failure caused by the rapid consumption of keys.

[0203] Figures 16 to 20 The key resource consumption of each method on the key link is simulated. Fig.16 and Fig.17It can be seen that the link key consumption of the Random and RIP-SRA methods is not uniform. At the end of key relay, the remaining keys on links 9-14 are all less than 280 bits, while the remaining keys on links 7-12 are all greater than 1050 bits. Too much remaining key leads to a waste of key resources, while too little remaining key leads to the failure of key distribution. Fig.19 and Fig. 20 The MP-GA and LCN-MPA methods will consume link keys evenly and quickly during relaying, but the consumption rate is too fast, which leads to a sharp drop in the amount of remaining keys in the key pool. The amount of remaining link keys after the relay is less than 170 bits, and key distribution fails due to insufficient keys in links 9-14. Fig.18 It can be seen that the key consumption trend of this method is relatively uniform and gentle. After the relay task is completed, the key amount on the link is relatively sufficient, and the range of the remaining key amount is only 214 bits. Therefore, this method can effectively reduce the occurrence of extreme situations such as key waste and key shortage, and has good key balance.

[0204] This application establishes a fish swarm behavior model and routing transmission architecture model under the key transmission mechanism, takes into account the impact of multi-attribute evaluation indicators on routing performance, gives priority to nodes that are close to the target node and have low key usage costs for key relay, and introduces a random strategy for routing selection. Finally, the feasibility of this method is verified through simulation. The results show that the fish swarm random routing method for quantum key distribution network proposed in this application effectively reduces the service blocking rate and key consumption during quantum key distribution while ensuring routing security, and balances the use of key resources in the network to a certain extent, improving the efficiency of quantum key distribution.

[0205] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. The present application is intended to cover any modification, use or adaptation of the present application, which follows the general principles of the present application and includes common knowledge or customary technical means in the art that are not disclosed in the present application.

Claims

1. A random routing method for quantum key distribution network simulating fish swarms, It is characterized in that The method includes: Inspired by the self-organizing behavior of fish schools, according to the key transmission principle and routing transmission architecture model of partially trusted relay quantum key distribution network, the routing process of global quantum key is abstracted as the optimization behavior of fish schools, and a fish school-like behavior model under the key transmission mechanism is constructed; wherein, the fish school-like behavior model includes recognition behavior, foraging behavior, aggregation behavior and mark diffusion behavior; According to the simulated fish school behavior pattern, a fish school perception random routing model is established; The fish school sensing random routing model is solved to allocate network resources.

2. According to claim 1, the quantum key distribution network fish swarm random routing method, It is characterized in that The method of establishing a fish school perception random routing model according to the simulated fish school behavior pattern includes: Before route selection, the controller obtains the remaining key of the link passing through the next-hop neighbor node link by querying the status information of each neighbor node in the routing table, and identifies the neighbor node through the path threshold, where: μ=v r ×T r (2); In the formula, e k Indicates a link, represents the remaining key of the link, μ represents the path threshold, k 0 Indicates the initial key quantity of the key pool. represents the quantum key generation rate, T r Indicates the service time slot requirement, T k Indicates the key update period, v r represents the key consumption rate, v represents the key consumption rate per service unit time, R represents the key service request set, and r represents the service; The distance evaluation factor is introduced to sort the k neighbor nodes in descending order according to the distance from the routing relay node to the destination node. The expression of the distance evaluation factor is as follows: Where W j represents the distance evaluation factor, d j Indicates the physical distance between routing relay nodes.

3. According to claim 2, the quantum key distribution network fish swarm random routing method, It is characterized in that The method of establishing a fish school perception random routing model according to the simulated fish school behavior pattern includes: The routing cost is evaluated from the perspectives of key resource availability and node credibility; The evaluation of the routing cost from two aspects of key resource availability and node credibility includes: The quantum key generation rate and key freshness constitute the link resource cost: C r =aC KGR +bC KF (4); In the formula, C r represents the link resource cost, C KGR represents the quantum key generation rate, C KF represents the key freshness; and is the capacity of the key pool at both ends of the link, where a and b represent the key resource influence weights, satisfying a+b=1, a=2 / 3, b=1 / 3; The expression of the routing trust cost is: In the formula, C T represents the trusted cost of routing, N p represents the set of candidate food nodes, v i represents the routing relay node, l(v i ) represents a trusted Boolean value; By querying N p The trusted Boolean value of the relay node in the middle route is used to calculate the security cost of selecting the current node route, and the trusted node is added to the trusted node set, otherwise it is added to the untrusted node set.

4. According to claim 3, the quantum key distribution network fish swarm random routing method, It is characterized in that The method of establishing a fish school perception random routing model according to the simulated fish school behavior pattern includes: The node food concentration is calculated for the candidate neighbor nodes in the trusted node set and the untrusted node set respectively. The expression of the node food concentration is as follows: In the formula, η(v i ) represents the node food concentration, α represents the distance weight coefficient, β represents the routing evaluation cost weight coefficient, |α|+|β|=1; The node food concentration is normalized to obtain the normalized node food concentration, and then the routing foraging behavior is performed, and the neighboring nodes are randomly selected by probability; wherein the expressions of the probabilities are as follows: or * (v i )=x p1 +x p2 (9); Where η * (v i ) represents the normalized node food concentration, x p1 represents the normalized eigenvalue of the distance evaluation factor matrix, x p2 represents the normalized eigenvalue of the routing cost matrix, P(v i ) represents probability.

5. According to claim 4, the method for random routing of a quantum key distribution network simulating a fish swarm, It is characterized in that The solving the fish school sensing random routing model to allocate network resources includes: Initialize the quantum key distribution network, collect network topology information, key pool resource information, node information, and add it to the routing table in the form of data packets; When both communicating parties receive the task request of forwarding the key, the fish swarm abstracts the routing process of the global quantum key into the message of the fish swarm's optimization behavior and performs unified routing calculation on the key relay task through the server; Traverse the neighboring nodes of the next hop and calculate the remaining key amount of the link between the node and each neighboring node and link threshold μ; like Then add the neighbor node to the candidate food node set N p ; Judge N p The number of neighbor nodes described in; If N p If there are multiple neighbor nodes, the congestion factor δ is set according to the network scale, and N p Number of markings of neighbor nodes in i ; If i >δ, exclude it N p ; Update N p , calculate the node food concentration η(v i ) and normalize it; Then with probability P(v i ) Perform routing selection and update the mark number n i =n i +1, and use the link key in the quantum key pool to the global key K c Perform encryption operations and update network resources; If K c When relayed to the destination node, the quantum key terminal connection between both parties is released.

6. According to claim 5, the method for random routing of a quantum key distribution network simulating a swarm of fish, It is characterized in that The global key K is constructed by using the link key in the quantum key pool. c Perform encryption operations, update network resources, and then include: If K c If the destination node is not reached, the data packet is sent and the routing table is updated.

7. According to claim 5, the method for random routing of a quantum key distribution network simulating a fish swarm, It is characterized in that The pair of judgments N p The number of neighbor nodes in the above-mentioned embodiment also includes: If there is a neighbor node, the link key in the quantum key pool is used to generate the global key K c Perform encryption operations and update network resources.

8. According to claim 5, the method for random routing of a quantum key distribution network simulating a fish swarm, It is characterized in that The pair of judgments N p The number of neighbor nodes in the above-mentioned embodiment also includes: If the neighbor node does not exist, the call loss is recorded.

9. According to claim 5, the method for random routing of a quantum key distribution network simulating a fish swarm, It is characterized in that If N p If there are multiple neighbor nodes, the congestion factor δ is set according to the network scale, and N p Number of markings of neighbor nodes in i , also includes: If i ≤δ, then calculate the node food concentration η(v i ) and normalize it.

10. According to claim 5, the method for random routing of a quantum key distribution network simulating a fish swarm, It is characterized in that The traversal of the neighboring nodes of the next hop calculates the remaining key amount R of the link between the node and each neighboring node ek With the link threshold μ, it also includes: like Then the neighbor node does not participate in routing selection.

Citation Information

Cited By

  • Centralized quantum key relay network and key distribution method

    CN119766443A