Authentication evaluation processing method based on DSN network

Abstract

The invention discloses an authentication evaluation processing method based on a DSN network, and the method comprises the steps: initializing particle swarm parameters, defining a node arrangement sequence of each potential transmission path, and filtering illegal identifiers based on a node trust evaluation library; in combination with judgment of the position hopping rate of the node sequence of each potential transmission path and the position hopping rate threshold, pseudo path nodes are eliminated, and an effective node set is generated; constructing a fitness function; after the node arrangement sequence and the particle speed of the potential transmission path are initialized, initial fitness is calculated, and individual optimum and global optimum are recorded; according to particle swarm parameters, a fitness function, an individual optimal Pbest and a global optimal Gbest, in combination with a trust guiding strategy, particle swarm optimization processing is carried out, iterative output is carried out to obtain a target path, and a high-reliability and high-performance transmission path is searched in a DSN network by utilizing a particle swarm optimization (PSO) algorithm and in combination with node trust evaluation and path dynamic stability analysis.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of distributed DSN network detection, and particularly relates to a DSN network-based authentication evaluation processing method. BACKGROUND

[0002] With the rapid development of information technology, network security problems have gradually become the focus of attention in various industries. In the field of modern identity authentication, traditional authentication methods mainly rely on static information such as passwords and smart cards for user identity verification. However, these traditional authentication methods face many challenges, especially in protecting user privacy and preventing information leakage. In distributed computing systems, communication and authentication processes between nodes are uncertain and are vulnerable to Byzantine faults. Especially in a distributed environment, nodes may fail or behave maliciously, and traditional authentication mechanisms cannot effectively deal with Byzantine fault tolerance problems.

[0003] For the above technical defects, the person skilled in the art encrypts the reconstructed key by using lattice cryptography algorithm; generates a quantum random number label containing a timestamp and a node topology; binds the token with double factors by using a hyper-elliptic curve signature algorithm; embeds a self-destruct counter in the token and initializes the countdown. And combine the deployment of deep packet inspection engine in the service node, execute: parse the asymmetric nested protocol header in the token; verify the Bell inequality correlation of the quantum random number label; use a predetermined tolerance threshold to detect the space-time coordinate offset; activate the defense chain when the offset is detected to be out of limit.

[0004] Specifically, the threat index TI of the current authentication session is calculated, which is the weighted integral of the multi-dimensional risk assessment parameters; the response level of the defense chain is dynamically adjusted according to the TI value; when TI exceeds the first threshold, multi-node collaborative tracing is started; when TI exceeds the second threshold, the operation of physical layer signal interference is triggered for optimization processing. However, in the above optimization processing process, ensuring efficient and secure transmission of information is an important cornerstone of building a reliable system. In the above distributed computing system (such as a distributed DSN network), dynamic optimization of data transmission paths is the key to improving the overall performance, stability and security of the system, which is the authentication evaluation of the DSN network. The ideal data transmission path needs to meet the requirements of low delay and high throughput while effectively avoiding the risks brought by malicious nodes, faulty nodes and network state fluctuations.

[0005] However, the prior art has significant technical defects in dynamically adjusting and evaluating the DSN network transmission path, and it is difficult to meet the efficient and secure transmission requirements in complex dynamic environments.

[0006] Firstly, the existing method often relies too much on static network topology information or a single dimension of historical indicators (such as simple historical interaction success rate) when evaluating the credibility of nodes or the reliability of paths. This leads to an inability to accurately and dynamically reflect the comprehensive trust conduction ability of nodes in the real network environment. For example, a node may exhibit unstable behavior patterns (such as fluctuation in interaction success rate or occasional violation of behavior signs) due to device aging, environmental interference, or intermittent attacks, but traditional static evaluation or single indicator model is difficult to capture such dynamic changes, and it is easy to misjudge it as a reliable node or fail to identify its risks in time. This deficiency in evaluating the trust conduction rate between nodes makes the path selection process unable to accurately filter potential unsafe or unreliable nodes, which poses a security risk.

[0007] In addition, the existing path optimization algorithm usually focuses on limited and isolated performance indicators (such as only considering average delay or average packet loss rate) when evaluating path performance, lacking comprehensive consideration of real-time health status and its fluctuation characteristics of the path. They are difficult to effectively integrate multi-dimensional information reflecting path instantaneous load (real-time load rate), communication timeliness (communication delay rate), transmission reliability (packet loss rate), and path overall stability (path fluctuation anomaly index). The singleness and lag of such evaluation lead to slow reaction and inaccurate adjustment of the algorithm when facing real-time fluctuations of network state (such as sudden traffic, link instantaneous outage, node performance jitter). It cannot identify and avoid those "sub-health" or "sick" path segments with extremely high current load, dramatic delay increase, severe packet loss, or extremely poor stability.

[0008] Due to the above deficiencies in trust evaluation and health status evaluation, the path optimization process is difficult to effectively utilize the trust relationship between nodes (trust conduction matrix) to guide the path exploration direction (lack of trust guidance strategy), and when a node performance degradation (low node priority score) is found in the discovered path, there is a lack of an efficient global priority list mechanism based on comprehensive indicators (trust conduction rate, H, F) to guide how to accurately select a better adjacent node for exchange adjustment. SUMMARY

[0009] The purpose of the present application is to provide a DSN network-based authentication evaluation processing method, which solves the above technical problems pointed out in the prior art.

[0010] The present application provides a DSN network-based authentication evaluation processing method, comprising the following operation steps:

[0011] Initialize particle swarm parameters; the particle swarm parameters include N potential transmission paths, the maximum number of iterations , learning factors c1 and c2, and inertia weight w;

[0012] Definition of the node arrangement sequence Xᵢ=[ID1,ID2,…,ID m ] of each potential transmission path;

[0013] The node identifiers {ID k} in the node arrangement sequence of each potential transmission path are filtered based on the node trust evaluation library to exclude illegal identifiers and exclude pseudo-path nodes based on the position jump rate of the node sequence of each potential transmission path and the judgment of the position jump rate threshold, to generate a set of valid nodes;

[0014] The path health degree response value constructed using the real-time load rate, communication delay rate, and packet loss rate of each valid node of each potential transmission path in the set of valid nodes is combined with the path fluctuation anomaly index to construct a fitness function;

[0015] The node arrangement sequence and particle velocity of the potential transmission path are randomly initialized, the initial fitness is calculated, and the individual optimal Pbest and global optimal Gbest are recorded;

[0016] According to the particle swarm parameters, the fitness function, the individual optimal Pbest and the global optimal Gbest, and the trust guidance strategy, the particle swarm optimization process is iteratively output to obtain the target path.

[0017] Preferably, according to the particle swarm parameters, the fitness function, the individual optimal Pbest and the global optimal Gbest, and the trust guidance strategy, the particle swarm optimization process is iteratively output to obtain the target path, including the following operation steps:

[0018] The iteration number is increased by 1, the current iteration number is obtained, and the particle velocity and the node arrangement sequence of the potential transmission path are updated using the trust guidance strategy based on the particle position under the current iteration number and the individual optimal Pbest and the global optimal Gbest, to obtain a new node arrangement sequence of the potential transmission path;

[0019] The fitness is recalculated based on the node arrangement sequence of the potential transmission path and the individual optimal Pbest and the global optimal Gbest under the current iteration number are updated;

[0020] According to the maximum iteration number and the fitness threshold, if the termination condition is met, the potential transmission path corresponding to Gbest is output as the target path, otherwise the above operation is returned, the iteration number is increased by 1, and the iteration process is performed again until the target path is output.

[0021] Preferably, the trust guidance strategy is used to update the particle velocity and the node arrangement sequence of the potential transmission path according to the node arrangement sequence of the potential transmission path under the current iteration number and the individual optimal Pbest and the global optimal Gbest, to obtain a new node arrangement sequence of the potential transmission path, including the following operation steps:

[0022] The trust conduction rate between each two nodes is obtained by analyzing the interaction between each two nodes in the historical time period.

[0023] The trust conduction matrix of the node sequence of the current path is constructed based on the trust conduction rate.

[0024] The particle velocity at the next iteration is calculated based on the trust conduction matrix, the node arrangement sequence of the potential transmission path under the current iteration number and the individual optimal Pbest, and the new node arrangement sequence of the potential transmission path is calculated based on the particle velocity under the current iteration number.

[0025] Preferably, the trust conduction rate between each two nodes is obtained by analyzing the interaction between each two nodes in the historical time period, including the following operation steps:

[0026] The number of interactions between the a th node and the b th node in the historical time period is collected, including the number of successful interactions and the total number of interactions; the interaction success rate is calculated.

[0027] The behavior flag of the a th node and the b th node in the historical time period is collected; the behavior flag includes the compliance behavior flag and the violation behavior flag; the behavior consistency index of the a th node and the b th node is calculated based on the behavior flag.

[0028] The trust conduction rate is calculated based on the interaction success rate and the behavior consistency index in the last corresponding interaction success time in the historical time period.

[0029] Preferably, after the individual optimal Pbest and the global optimal Gbest under the current iteration number are obtained by recalculating the fitness based on the node arrangement sequence of the potential transmission path and updating, the following operation steps are further included:

[0030] The global priority list is constructed by adjusting the adjacent trust conduction rate, the path health degree response value H and the path fluctuation anomaly index F for each node of each potential transmission path.

[0031] Preferably, the global priority list is constructed by adjusting the adjacent trust conduction rate, the path health degree response value H and the path fluctuation anomaly index F for each node of each potential transmission path, including the following operation steps:

[0032] obtaining the k nodes of the ith potential transmission path, and calculating the node priority score of the kth node through the trust conduction rate, the path health response value H and the path fluctuation anomaly index F;

[0033] establishing a local priority list corresponding to each node in the current ith potential transmission path based on the node arrangement sequence and the node priority score of the current ith potential transmission path;

[0034] determining the kth node of the current ith potential path as an adjustment node when the node priority score of the kth node of the current ith potential path is less than the node priority score of the kth node in the global optimal Gbest;

[0035] According to the local priority list, the adjustment node is exchanged and adjusted through the comparison of the trust conduction rates of adjacent nodes to obtain an updated local priority list;

[0036] From each updated local priority list, the corresponding potential transmission path with the highest fitness is selected as the global priority list.

[0037] Preferably, according to the local priority list, the adjustment node is exchanged and adjusted through the comparison of the trust conduction rates of adjacent nodes to obtain an updated local priority list, including the following operation steps:

[0038] According to the path topology matrix, the adjacent nodes of each adjustment node are obtained, and an adjacent node set of the adjustment node is constructed;

[0039] Obtaining the historical interaction information between the adjustment node and each adjacent node in the adjacent node set to calculate the adjustment trust conduction rate;

[0040] Obtaining a first screening condition, and determining the adjacent node as a node to be exchanged when the adjustment trust conduction rate is greater than or equal to a preset first adjustment threshold value;

[0041] Obtaining a second screening condition, and searching the trust degree in the node trust evaluation library for the node to be exchanged; and determining the node to be exchanged as a target exchange node when the trust degree is greater than or equal to a second adjustment threshold value;

[0042] Exchanging the adjustment node and the target exchange node to generate an adjusted local priority list; checking whether the node priority score in the adjusted local priority list meets the exchange verification rule, and if so, returning to the above operation for re-exchange processing until an updated local priority list is output.

[0043] Preferably, the exchange verification rule refers to the node priority score of the kth node in the current ith potential path being greater than or equal to the node priority score of the kth node in the global optimal Gbest.

[0044] Compared with the prior art, the embodiments of the present application have at least the following technical advantages:

[0045] It can be known from the above-mentioned authentication evaluation processing method based on the DSN network provided by the present application that, in specific application, firstly, the authentication evaluation prepositioning and path screening processing are performed, that is, after the node sequence of the potential path is defined, the node-level authentication processing is immediately performed, and each identifier {ID k} in the sequence is authenticated in real time by relying on the node trust evaluation library, and any illegal node identifier marked as malicious, faulty or unauthorized is filtered out, and the authentication and authorization check of the node identity and state is performed at the source of path construction;

[0046] Further, by analyzing the "position jump rate" (that is, the possibility of direct connection of adjacent nodes in the physical or logical topology) of the node sequence and comparing it with the set threshold, the pseudo-path nodes with unreasonable structure or fraudulent connection are excluded, and the authenticity and feasibility of the path topology structure are preliminarily evaluated; then, after the above-mentioned two-step strict authentication (node identity / state) and evaluation (path structure rationality), the "effective node set" of each potential path is finally generated, so that all subsequent optimization operations are based on the authenticated and evaluated, legal and structurally reasonable nodes and path segments; further, the embodiments of the present application also construct a path topology matrix based on the effective node set and the connection relationship thereof, and take the link trust value as the matrix element, the trust value fuses the historical successful transmission rate and the trust time decay factor, dynamically quantifies the reliability of the link, and reflects the "trustworthiness" historical performance of the link between the authenticated nodes, which is a key indicator for evaluating the continuous reliability of the link after authentication; further, the embodiments of the present application calculate the path fluctuation anomaly index F by analyzing the change (fluctuation of the number of node connections) of the path topology matrix in the continuous time window, and quantify the historical stability of the path topology structure; the maximum fitness function value of the direct target optimized by the particle swarm algorithm, in all potential paths that have passed the preliminary authentication and evaluation, finds the path with the optimal comprehensive authentication evaluation result (reliability, stability, performance), the "trust guiding strategy" introduced in the particle velocity update rule strengthens the preference of the optimization process to the high-trust and high-stability area, and the particle not only learns from the historical optimal (Pbest) and the group historical optimal (Gbest), but also is influenced by the trust and stability information implied in the path, so that the search is guided to converge to the path area with higher authentication evaluation comprehensive score more quickly;

[0047] Finally, when the particle swarm algorithm meets the termination condition, the output global optimal solution Gbest corresponding to the node sequence (transmission path) is the path with the highest comprehensive authentication evaluation score found in the entire optimization process. BRIEF DESCRIPTION OF DRAWINGS

[0048] Figure 1 A main flowchart of a kind of authentication evaluation processing method based on DSN network;

[0049] Figure 2 A filtering illegal identifier simulation schematic diagram in a kind of authentication evaluation processing method based on DSN network;

[0050] Figure 3 A filtering illegal identifier simulation schematic diagram in a kind of authentication evaluation processing method based on DSN network;

[0051] Figure 4 A global priority list establishment flow simulation schematic diagram in a kind of authentication evaluation processing method based on DSN network. DETAILED DESCRIPTION

[0052] The technical solutions of the present application will be described clearly and completely below in conjunction with the drawings. Obviously, the described embodiments are part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0053] First of all, the present application provides a kind of authentication evaluation processing method based on DSN network, which is for the evaluation of the distributed DSN network;The main function of the distributed DSN network is to collect identity authentication data and other feature data from access terminal in real time to carry out security authentication network, the network needs to be based on the authentication evaluation processing of DSN network to evaluate the stability and security, for this, an embodiment of the present application relates to a kind of authentication evaluation processing method based on DSN network, see detailed in embodiment one.

[0054] The present application will be described in further detail below through specific embodiments and in conjunction with the drawings.

[0055] Embodiment one

[0056] As shown in Figure 1 , the present application embodiment one provides a kind of authentication evaluation processing method based on DSN network, including the following operation steps:

[0057] Step S10: initialize particle swarm parameters;The particle swarm parameters include N potential transmission paths, maximum iteration times , learning factor c1 and c2 and inertia weight w;

[0058] Step S20: defining the node arrangement sequence Xᵢ=[ID1,ID2,…,ID m ] of each potential transmission path;

[0059] Step S30: filtering illegal identifiers based on the node identifiers {ID k} in the node arrangement sequence of each potential transmission path through the node trust evaluation library and excluding pseudo-path nodes in combination with the judgment of the position jump rate of the node sequence of each potential transmission path and the position jump rate threshold, to generate a valid node set;

[0060] It should be noted that the above potential path refers to the forwarding path that may exist between the sending node and the receiving node during the DSN network data forwarding process. The above illegal identifier refers to a node that is marked as malicious, faulty or unauthorized, for example;

[0061] Step S40: constructing a fitness function Fitness=α·H+β·(1-H)+γ·F using the path health response value constructed by the real-time load rate, communication delay rate and packet loss rate of each valid node of each potential transmission path in the valid node set in combination with the path fluctuation anomaly index, where H is the path health response value (the path health response value H is calculated by weighted summation of the real-time load rate L k , the communication delay D k , and the packet loss rate P k ), F is the path fluctuation anomaly index, and α, β, γ are weight coefficients;

[0062] It should be noted that the embodiments of the present application first establish a path topology matrix based on each path and the corresponding valid node (convert the physical connection relationship of the current path into a mathematical representation, and at the same time, incorporate the link trust value to quantify the transmission reliability) to encode the path topology structure into a sparse matrix (the sparse matrix is the above path topology matrix), only adjacent nodes have non-zero values, and at the same time, the historical success rate and the time decay of trust are considered by combining the trust value, so as to dynamically reflect the link reliability; then, T consecutive time windows are selected, and the number of topology links (the number of directly adjacent nodes) in the consecutive time windows is counted for each node in the path topology matrix; the absolute change amount (i.e., the absolute value of the change amount) of the number of topology links in the adjacent time windows in the consecutive time windows is calculated; after normalization of the absolute change amount, a historical fluctuation matrix is constructed; the path fluctuation anomaly index F is calculated based on the historical fluctuation matrix and the path topology matrix;

[0063] The time decay of the trust is to consider that the trust will decay over time (the more distant historical interaction, the less impact on the current trust degree) when calculating the link trust value in the path, to ensure that the link trust value in the path topology matrix can dynamically reflect the latest network state and node reliability, avoid over-reliance on old historical data, so that the calculated path fluctuation anomaly index F is more accurate (the higher the F value, the greater the path historical fluctuation and instability); the reason for introducing the time decay of the trust is that, for example, a once-trusted node may be invaded, malfunctioned or malicious, and old successful records cannot guarantee current reliability, or the interaction data a long time ago has low value for predicting the current behavior of the node.

[0064] Step S50: randomly initializing the node arrangement sequence of the potential transmission path and the particle velocity, calculating the initial fitness and recording the individual optimal Pbest and the global optimal Gbest;

[0065] The particle velocity represents the trend of the potential transmission path in the path search process and the variation intensity in the trend.

[0066] It should be noted that the embodiments of the present application gradually search for an optimal transmission path by the initialized node arrangement sequence of the potential transmission path, according to the initialized path search trend and the variation intensity of the path search trend direction.

[0067] Step S60: obtaining the current iteration number, updating the particle velocity and the node arrangement sequence of the potential transmission path according to the particle position under the current iteration number and the individual optimal Pbest and the global optimal Gbest by using the trust guidance strategy, to obtain a new node arrangement sequence of the potential transmission path;

[0068] It should be noted that the embodiments of the present application update the particle velocity and the node arrangement sequence of the potential transmission path according to the particle position under the current iteration number and the individual optimal Pbest and the global optimal Gbest by using the trust guidance strategy, to obtain a new node arrangement sequence of the potential transmission path, that is, the core iteration processing operation in the particle swarm algorithm process, to drive the node arrangement sequence of the current potential transmission path to move to a better solution (i.e., the new node arrangement sequence of the transmission path) according to the experience (i.e., the individual optimal Pbest), the group experience (i.e., the global optimal Gbest) and the current state (the node arrangement sequence of the current potential transmission path) of the current potential transmission path, and to guide the search to a more reliable direction position by using the trust guidance strategy, to improve the accuracy of subsequent optimization.

[0069] Step S70: Recalculating the fitness based on the node arrangement sequence of the potential transmission path and updating the individual optimal Pbest and global optimal Gbest under the current iteration number;

[0070] Step S80: According to the maximum iteration number judging with the fitness threshold, if the termination condition is met, the potential transmission path corresponding to Gbest is output as the target path, otherwise, returning to S60.

[0071] It should be noted that the above embodiments of the present application use particle swarm optimization algorithm (PSO) to find a high-reliability and high-performance transmission path in DSN network by combining node trust evaluation and path dynamic stability analysis;

[0072] Specifically, first, the particle swarm optimization algorithm is set by initializing the particle swarm parameters to set the basic rules and control parameters of the operation, which determines the range of the search space, the dynamic characteristics of the search process (such as the balance between exploration and development), and the termination condition of the algorithm, laying the foundation for the entire optimization process; further, the abstract "transmission path" is concretized into a mathematically processable representation, and each path is modeled as an ordered node ID sequence (Xᵢ), which provides a clear operation object for subsequent node filtering, path topology matrix construction, fitness calculation, and position updating of the particle swarm;

[0073] Further, as shown in Figure 2 , the node identifiers {ID k} in the node arrangement sequence of each potential transmission path are filtered based on the node trust evaluation library, and the node arrangement sequence jump rate and position jump rate threshold (such as 1.2) of each potential transmission path are used to exclude pseudo-path nodes, generating a valid node set. As shown in Figure 3 , the node sequence of each path defined in step S20 is verified for security and reasonableness, and illegal identifiers (such as nodes marked as malicious, faulty, or unauthorized) are filtered based on the node trust evaluation library to ensure that the path only contains trusted nodes. By checking the node arrangement sequence jump rate of the potential transmission path of the node sequence (i.e., whether the adjacent nodes are allowed to be directly connected in the physical topology or logical topology) and comparing it with the position jump rate threshold, path segments with node sequences that are physically / logically impossible or unreasonable (pseudo-path nodes) are excluded, and finally an effective node set is generated for each potential path. The set only contains legal and reasonable nodes that have passed trust verification and physical / logical location verification, reducing the search space for subsequent calculations and improving efficiency and result reliability; further, the fitness function defines a comprehensive standard for evaluating the pros and cons of a transmission path, quantifying the overall performance of the path and guiding the PSO algorithm to search for a better path;

[0074] Wherein, based on the effective node set and its physical connection relationship, a sparse matrix (the sparse matrix is the path topology matrix described above) is established, the matrix elements fuse the link trust value (calculated dynamically by the historical success rate and the time decay of trust), and the transmission reliability of the link is quantized. In T consecutive time windows, the number of topology connections (i.e. the number of directly adjacent nodes) of each node in each window in the matrix is counted, the absolute change amount of the number of topology connections of each node between adjacent time windows is calculated, and the absolute change amount is normalized to form a matrix reflecting the connection change history of the node.

[0075] Based on the historical fluctuation matrix and the current path topology matrix, F is calculated and analyzed, which quantifies the abnormal degree of fluctuation of the path topology structure (node connection relationship) in history. Wherein, the higher the F value is, the more unstable the connection relationship of the path is in history, and the more frequent and severe the abnormal fluctuation is, which guides the algorithm to select a path with high long-term stability;

[0076] Then, the particle swarm optimization process is started, a random starting position (Xᵢ sequence) and a starting speed (Vᵢ) are set for each particle (representing a path), points (particles) are randomly scattered in the defined search space (node sequence space), the speed Vᵢ determines the trend and strength of the movement (change of node sequence) of the particle in the initial iteration, the initial fitness value of each particle is calculated, Pbestᵢ is used to record the best position (node arrangement sequence of the transmission path) found by each particle so far, and Gbest records the best position (node arrangement sequence of the transmission path) found by the entire particle swarm so far; then, the particle (path) is driven to move to a better solution according to its own experience (Pbest), the experience of the group (Gbest) and the current state, so that the particle tends to move to the Pbest or Gbest direction containing a high-trust node / link, or the speed update amplitude is affected by the trust value, and through the speed update formula, a balance is achieved between exploring new areas (global search) and developing known good areas (local search). The trust guiding strategy aims to guide the search to a more reliable direction, and by evaluating the effect of the particle movement (path change) and updating the historical optimal record, it ensures that the algorithm always remembers and tracks the best solution (individual optimal and global optimal) found so far; finally, by checking whether the preset termination condition (the termination condition is whether the maximum number of iterations set in S10 is reached , or the fitness of Gbest converges to a threshold value, etc.), the iteration optimization process of the scheme is realized, and it is ensured that the algorithm ends in a limited number of steps and gives the final result;

[0077] The above embodiments of the present application first perform authentication evaluation prepositioning and path screening processing, i.e. after defining the node sequence of the potential path (S20), the key step S30 of node-level authentication processing is immediately performed, which relies on the node trust evaluation library to evaluate each identifier {ID kReal-time authentication is performed to filter out any illegal node identifier marked as malicious, faulty or unauthorized, and authentication and authorization checks of node identity and status are performed at the source of path construction; further, in step S30, by analyzing the "location jump rate" of the node sequence (i.e. the possibility of direct connection of adjacent nodes in physical or logical topology) and comparing it with the set threshold, the pseudo-path nodes with unreasonable structure or fraudulent connection are excluded, and the authenticity and feasibility of the path topology structure are preliminarily evaluated; then, after the above two strict authentication (node identity / status) and evaluation (path structure rationality), the "effective node set" of each potential path is finally generated, ensuring that all subsequent optimization operations are based only on authenticated and evaluated, legal and structurally reasonable nodes and path segments;

[0078] Further, the above processing steps construct a path topology matrix based on the effective node set and its connection relationship, take the link trust value as the matrix element, and dynamically quantify the reliability of the link by fusing the historical successful transmission rate and the trust time decay factor, reflecting the "trustworthiness" historical performance of the link between the authenticated nodes, which is a key indicator for evaluating the continuous reliability of the authenticated link; further, the above processing steps S20-S40 calculate the path fluctuation anomaly index F by analyzing the changes (fluctuation of the number of node connections) of the path topology matrix in the continuous time window, quantifying the historical stability of the path topology structure;

[0079] In the following steps S50-S80, the direct target maximum fitness function value optimized by the particle swarm algorithm is used to find the path with the optimal comprehensive authentication and evaluation result (reliability, stability, performance) among all the potential paths that have passed the preliminary authentication and evaluation, and the "trust guiding strategy" introduced in the particle velocity update rule further strengthens the preference of the optimization process for high-trust and high-stability areas, and the particles not only learn from their own historical optimum (Pbest) and group historical optimum (Gbest), but also are influenced by the trust and stability information implied in the path, guiding the search to converge faster to the path area with higher comprehensive authentication and evaluation score; finally, when the algorithm meets the termination condition (step S80), the node sequence (transmission path) corresponding to the global optimal solution Gbest output is the path with the highest comprehensive authentication and evaluation score found in the entire optimization process.

[0080] Specifically, in step S60, the trust guiding strategy is used to update the particle velocity and the node arrangement sequence of the potential transmission path according to the node arrangement sequence of the potential transmission path in the current iteration number and the individual optimum Pbest and the global optimum Gbest, to obtain a new node arrangement sequence of the potential transmission path, including the following operation steps:

[0081] Step S61: obtaining the trust conduction rate between each two nodes based on the interaction analysis between each two nodes in the historical time period;

[0082] S62: constructing the trust conduction matrix of the node sequence of the current path based on the trust conduction rate;

[0083] S63: obtaining the particle velocity in the next iteration based on the trust conduction matrix, the node arrangement sequence of the potential transmission path in the current iteration number and the individual optimal Pbest, and calculating the node arrangement sequence of the new potential transmission path based on the particle velocity in the current iteration number.

[0084] The calculation method of the particle velocity in the next iteration is as follows:

[0085] ;

[0086] In the formula, M= ;

[0087] In the formula, w is the inertia weight (or the velocity retention coefficient, which is used to control the retention proportion of the particle velocity in the last iteration), is the velocity of the ith potential transmission path at time t, is the learning factor (the learning factor of the individual cognitive term), is the random factor (the random factor of the individual cognitive term), is the path health degree response value, F is the path fluctuation anomaly index, is the trust conduction matrix;

[0088] is the node arrangement sequence individual optimal of the ith potential transmission path, is the node arrangement sequence of the ith potential transmission path at the tth iteration, is the learning factor (the learning factor of the global cognitive term), is the random factor (the random factor of the global cognitive term), is the global trust matrix, is the Hadamard product (matrix element-by-element multiplication), is the tensor product operator, is the node arrangement sequence global optimal of the ith potential transmission path, is the number of nodes of the potential transmission path, is the trust conduction rate (i.e. the trust influence of node k on node k+1), is the trust decay factor, is the path health degree response value fluctuation value, is the transmission delay coefficient, is the link availability rate;

[0089] It should be noted that the trust decay factor is used to quantify the speed of decay of trust over time, and the trust decay factor is a key parameter for landing the "time decay consideration of trust" in S40 into a computable mathematical model. By adjusting the value of the trust decay factor , the dynamic nature of different network environments can be flexibly adapted.

[0090] The above embodiments of the present application first analyze the interaction between each two nodes in the historical time period, calculate the trust conduction rate between each two nodes, understand the change of trust relationship between nodes, and thus provide a basis for subsequent path selection; further, the trust conduction rate is used to construct a trust conduction matrix to represent the trust transmission between different nodes, and provide more accurate trust information for subsequent particle swarm search; then, based on the trust conduction matrix and the node arrangement sequence of the current particle, the Pbest and Gbest are combined to calculate the new particle velocity, and the node arrangement sequence is updated, the direction of particle movement is influenced by the trust conduction matrix, and it is ensured that the search process is guided to a more reliable path.

[0091] Specifically, in step S61, the trust conduction rate between each two nodes is obtained by analyzing the interaction between each two nodes in the historical time period, including the following operation steps:

[0092] Step S611: Collect the interaction times between the a-th node and the b-th node in the historical time period, and the interaction times include the successful interaction times and the total interaction times; the interaction success rate is calculated;

[0093] Step S612: Collect the behavior flags of the a-th node and the b-th node in the historical time period; the behavior flags include compliance behavior flags and violation behavior flags; based on the behavior flags, the behavior consistency index of the a-th node and the b-th node is calculated;

[0094] It should be noted that the above behavior consistency index reflects the similarity of the behavior patterns of the a-th node and the b-th node in terms of "compliance" and "violation" in a given historical time period, which is a value calculated by statistical analysis or pattern matching of the behavior indicators (compliant behavior indicators and violation behavior indicators) of the two nodes in the same time period, which represents the coordination or similarity of the two nodes in the historical behavior pattern. Specifically, it measures whether the two nodes tend to make the same type of judgment (compliant or non-compliant) when facing similar situations, whether the compliant behavior or non-compliant behavior of the two nodes has relevance or consistency in terms of time, frequency or pattern, reflecting the degree of consistency between node a and node b in terms of "rules" or "rules". Among them, the highly consistent behavior (whether compliant or non-compliant) may reflect the similar interpretation of the two nodes to the network rules, the similar compliance intention (or destruction intention), the continuous behavior consistency may imply that there is some contact, cooperation relationship (such as belonging to the same trusted organization, running the same strategy) or mutual influence (such as one node imitating the behavior of another node) between the nodes, on the contrary, low consistency may indicate that the nodes are independently operated.

[0095] Step S613: Calculate the trust transmission rate by combining the last corresponding interaction success time in the historical time period with the interaction success rate and the behavior consistency index.

[0096] It should be noted that the above embodiment of the present application first collects the number of interactions between the a-th node and the b-th node in the historical time period, and calculates the interaction success rate, which reflects the communication quality and interaction reliability between the two nodes. The more successful interactions, the more stable the cooperation between the two nodes, and the higher the interaction success rate, which provides preliminary basic data for the calculation of the trust transmission rate, reflecting the basic trust degree between nodes; further, by collecting the behavior indicators between each two nodes, the behavior consistency index is calculated, which provides additional security data for the final calculation of the trust transmission rate.

[0097] Further, in step S613, the trust transmission rate is comprehensively obtained by combining the success time of the last interaction in the historical period, the interaction success rate and the behavior consistency index. The trust transmission rate is the final index for measuring the trust degree between nodes. By combining the interaction success rate and the behavior consistency, the trust transmission rate can accurately reflect the trust transmission degree between nodes. The addition of the success time makes the index consider the historical relationship and trend between nodes, so as to obtain a more time-effective trust transmission rate, which provides reliable trust data for subsequent particle swarm optimization path update, guiding the particle swarm to preferentially select the path with high trust degree in the search process.

[0098] Specifically, in step S70, after the fitness of the potential transmission path is recalculated based on the node arrangement sequence of the potential transmission path and the individual optimal Pbest and global optimal Gbest in the current iteration number are updated, the following operation steps are further included:

[0099] Step S71: constructing a global priority list based on the adjacent trust conduction rate, the path health degree response value H and the path fluctuation anomaly index F of each node of each potential transmission path;

[0100] It should be noted that the above embodiment of the present application reorders the nodes in each potential transmission path, evaluates the priority and adjusts the list, dynamically optimizes the order of the nodes in each path by using the trust conduction rate between each pair of nodes in the trust conduction matrix, thereby further improving the reliability and stability of the transmission path in the entire DSN network. Specifically, for each node (denoted as the jth node) of each potential transmission path (denoted as the ith path), a priority score function based on the adjacent trust conduction rate is constructed. The priority score considers the trust conduction between adjacent nodes in the path and combines the influence of the path health degree response value H and the fluctuation anomaly index F, so that the arrangement of each node in the path can meet the overall dynamic reliability and stability requirements. After completing the local and global node ordering and priority adjustment, the fitness of the entire transmission path is recalculated using the new node arrangement sequence to determine whether to accept the re-planned path or continue iteration adjustment. For details, see steps S711-S715.

[0101] Specifically, in step S71, a global priority list is constructed based on the adjacent trust conduction rate, the path health degree response value H and the path fluctuation anomaly index F of each node of each potential transmission path, including the following operation steps:

[0102] Step S711: obtaining k nodes of the ith potential transmission path, and calculating the node priority score of the kth node by using the trust conduction rate, the path health degree response value H and the path fluctuation anomaly index F;

[0103] The calculation method of the node priority score is:

[0104] ;

[0105] In the formula, is the trust conduction rate from the k-1th node to the kth node in the current ith potential transmission path, is the trust conduction rate from the kth node to the k+1th node in the current ith potential transmission path, , and are weight coefficients, and + + = 1 ;

[0106] Step S712: based on the node arrangement sequence of the current ith potential transmission path and the node priority score, a local priority list corresponding to each node in the current ith potential transmission path is established;

[0107] Step S713: when it is judged that the node priority score of the kth node in the current ith potential path is less than the node priority score of the kth node in the global optimal Gbest, it is determined that the kth node of the current ith potential path is an adjustment node;

[0108] It should be noted that the above embodiment of the present application compares the node priority score of the node of sequence k in the current ith potential path (sequence k is the node ranked kth in the node arrangement sequence, that is, the kth node in the ith potential path) with the node priority score of the node of sequence k in the global optimal Gbest, compares the node scores of the corresponding positions in each potential path, and when the node priority score of the kth node in the current ith potential path is significantly smaller (significantly smaller means that the node priority score of the kth node in the current ith potential path plus the priority score parameter threshold is still less than the node priority score of the kth node in the global optimal Gbest), it is indicated that there is optimization space for the node, and the node is marked as an adjustment node for subsequent adjustment processing.

[0109] Step S714: according to the local priority list, the adjustment node is adjusted by comparing the trust conduction rate of adjacent nodes to obtain an updated local priority list;

[0110] Step S715: from each updated local priority list, the corresponding potential transmission path with the highest fitness is selected as the updated local priority list, and the updated local priority list is selected as the global priority list.

[0111] It should be noted that the above steps of the embodiments of the present application first calculate the priority score of each node of each potential transmission path, and the trust conduction rate of adjacent nodes is considered in the calculation, i.e., the trust conduction rate between the k-1th and kth nodes and the trust conduction rate between the kth and k+1th nodes, and the priority score of the node is evaluated by combining the trust conduction rate, the path health degree, and the fluctuation anomaly index F. The above processing operation provides the basis for node sorting and adjustment for the subsequent steps. Further, based on the priority score of each node obtained in step S711, a local priority list of the current potential transmission path is constructed. The local priority list is for all nodes in a specific path, and the nodes are sorted according to the priority score. Each node in the path is assigned a ranking based on the current network state, which provides a basis for subsequent node adjustment (steps S713 and S714). Further, by comparing the priority score of the node in the current ith potential path with the node score of the global optimal solution Gbest, it is determined whether there is a node to be adjusted. Specifically, if the priority score of a node in the current path is significantly lower than the corresponding node score in Gbest (i.e., the priority score of the current node plus the priority score parameter threshold is still lower than the priority score of the global optimal solution), the node will be marked as a node to be adjusted, and potential performance bottlenecks or optimization spaces are found, and the nodes that need to be adjusted are marked. Further, for the node to be adjusted, the node to be adjusted is exchanged or adjusted according to the comparison of the local priority list and the trust conduction rate, and the transmission path is further optimized to ensure that the order of the path meets the overall reliability and stability requirements, thereby improving the fitness of the entire transmission path.

[0112] Specifically, as Figure 4As shown, the embodiment of the application obtains the path health degree response value H and the path fluctuation anomaly index F through the path evaluation process S40, quantifies the real-time performance and historical stability characteristics of the path, and transmits them to the global optimization process operation S70. The global optimization process operation S70 provides the complete node sequence of the global optimal path (Gbest) determined in the current iteration cycle to the node priority scoring operation (i.e., the node priority scoring of S711), and the node-level optimization provides an authoritative reference frame. After the node priority scoring process operation of S711, the Gbest node sequence is received, and the trust conduction rate, health degree H, and fluctuation index F are combined to perform node priority scoring calculation on each potential path. Then, an optimization proposal containing a node exchange adjustment scheme is generated and submitted to the global priority list acquisition process operation S715. Further, S715 evaluates the optimization schemes submitted by each path, selects the optimization path with the highest fitness as a new generation of global optimal path candidate, and feeds back the new candidate path to S70. After obtaining the new candidate path in S70, the fitness calculation operation S40 is triggered to reevaluate the path, and H (path health degree response value) and F (path fluctuation anomaly index) are recalculated based on the updated path topology, completing the optimization closed loop.

[0113] The traditional PSO only updates the entire path sequence (particle position) in path optimization, and cannot perform local repair on specific low-quality nodes in the path. The embodiment of the application introduces the node priority scoring mechanism through the processing operations S711-S715, refines the path optimization granularity from “entire path” to “single node” through the trust conduction rate (ensuring the continuity of the trust chain between nodes), the path health degree response value (reflecting the current state of node load, delay, etc.), and the path fluctuation anomaly index (avoiding historically unstable nodes), and directly improves the local reliability of the path.

[0114] Specifically, in step S714, the exchange adjustment processing of the to-be-adjusted node is performed through the comparison of the trust conduction rates of adjacent nodes according to the local priority list, and an updated local priority list is obtained, including the following operation steps:

[0115] Step S7141: Obtain the adjacent nodes of each to-be-adjusted node according to the path topology matrix, and construct the adjacent node set of the to-be-adjusted node;

[0116] Step S7142: Calculate the adjustment trust conduction rate from the historical interaction information between the to-be-adjusted node and each adjacent node in the adjacent node set;

[0117] It should be noted that the historical interaction information includes an interaction success rate between the to-be-adjusted node and the adjacent node, and a behavior consistency index.

[0118] Step S7143: Obtain a first screening condition, and determine the adjacent node as a to-be-exchanged node when the adjusted trust conduction rate is greater than or equal to a preset first adjustment threshold.

[0119] Step S7144: Obtain a second screening condition, and search for a trust degree of the to-be-exchanged node in the node trust evaluation library; and determine the to-be-exchanged node as a target exchanged node when the trust degree is greater than or equal to a second adjustment threshold. That is, the first screening condition refers to screening the adjacent node whose adjusted trust conduction rate meets the standard (i.e., the adjusted trust conduction rate is greater than or equal to the preset first adjustment threshold in the above step S7143), and the second screening condition refers to further screening the target exchanged node according to the trust degree corresponding to the trust degree of the to-be-exchanged node in the node trust evaluation library (i.e., the trust degree is greater than or equal to the second adjustment threshold in the above step S7144).

[0120] Step S7145: Exchange the to-be-adjusted node and the target exchanged node, and generate an adjusted local priority list; check whether the node priority score in the adjusted local priority list meets the exchange verification rule, and if so, return to the above operation (i.e., the processing of step S7141) for re-exchange processing until an updated local priority list is output.

[0121] The exchange verification rule refers to that the node priority score of the kth node in the current ith potential path is greater than or equal to the node priority score of the kth node in the global optimal Gbest.

[0122] It should be noted that the embodiment of the present application fuses the interaction success rate and the behavior consistency to ensure that the conduction rate reflects the real-time trust state, ensures the stability of the new node and the upstream and downstream trust chain through the trust conduction rate (core optimization goal), selects the “easy integration” node preferentially to reduce the path disturbance by quantifying the exchange cost according to the path topology matrix (i.e., the closer to the to-be-adjusted node, the more convenient to replace, and the lower the exchange cost), and verifies the trust improvement, stability guarantee and length constraint of the exchanged path from a global perspective through the exchange verification rule, thereby avoiding the overall deterioration caused by local optimization.

[0123] In summary, the authentication evaluation processing method based on the DSN network according to the embodiment of the present application finds a high-reliability and high-performance transmission path by using the particle swarm optimization algorithm (PSO) in the DSN network, combining node trust evaluation and path dynamic stability analysis.

[0124] In the specific implementation process, by combining the interaction analysis in the historical time period, the trust transmission rate between each two nodes is calculated, the change of the trust relationship between the nodes is understood, and the basis for subsequent path selection is provided; further, the trust transmission rate is used to construct a trust transmission matrix to represent the trust transmission between different nodes, and more accurate trust information is provided for the subsequent particle swarm search;

[0125] In the further implementation process, the trust transmission rate is calculated through the interaction success rate analysis and the behavior consistency index analysis between each two nodes, which provides reliable trust data for the path update in the subsequent particle swarm optimization, and guides the particle swarm to preferentially select the path with high trust degree in the search process;

[0126] Further, based on the global priority list of each node of each potential transmission path, a node priority scoring mechanism is introduced, the trust transmission rate (guaranteeing the continuity of the trust chain between nodes), the path health degree response value (reflecting the current state of the node such as load and delay), and the path fluctuation anomaly index (avoiding historical unstable nodes) are used to refine the path optimization granularity from "the whole path" to "single node", and the local reliability of the path is directly improved.

[0127] Further, by fusing the interaction success rate and the behavior consistency of each node to be adjusted, it is ensured that the transmission rate reflects the real-time trust state, the trust transmission rate ensures that the new node and the upstream and downstream trust chain are stable, the exchange cost is quantified by the path topology matrix, the "easy integration" node is preferentially selected, and the path disturbance is reduced; the trust improvement, stability guarantee and length constraint of the path after exchange are verified from the global perspective by the exchange verification rule, and local optimization leading to overall degradation is avoided.

[0128] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; the ordinary skilled in the art can modify the technical solutions described in the above embodiments, or replace some or all of the technical features with equivalent ones; these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. An authentication assessment processing method based on a DSN network, characterized in that, The following steps are included: Initialize particle swarm parameters; the particle swarm parameters include N potential transport paths and the maximum number of iterations. Learning factors c1 and c2 and inertia weight w; Define the node permutation sequence Xᵢ=[ID1,ID2,…,ID] for each potential transmission path. m ]; The node identifier {ID} in the node arrangement sequence of each potential transmission path k The system filters out illegal identifiers based on the node trust assessment library and excludes pseudo-path nodes by combining the judgment of the position jump rate and position jump rate threshold of the node order of each potential transmission path, thereby generating a valid node set. The fitness function is constructed by combining the path health response value obtained by constructing the real-time load rate, communication latency rate and packet loss rate of each effective node in the effective node set of each potential transmission path with the path fluctuation anomaly index. The node arrangement sequence and particle velocity of potential transmission paths are randomly initialized, the initial fitness is calculated, and the individual best Pbest and the global best Gbest are recorded. Based on the particle swarm parameters, fitness function, individual best Pbest and global best Gbest combined with the trust-guided strategy, particle swarm optimization is performed and iterative output is obtained to obtain the target path.

2. The authentication evaluation processing method based on a DSN network according to claim 1, characterized in that, The target path is obtained through iterative particle swarm optimization based on particle swarm parameters, fitness function, individual best (Pbest) and global best (Gbest) combined with a trust-guided strategy. The process includes the following steps: Increment the iteration count by 1 to obtain the current iteration count. Based on the particle position and the individual best Pbest and the global best Gbest under the current iteration count, update the particle velocity and the node arrangement sequence of the potential transmission path using a trust-guided strategy to obtain the new node arrangement sequence of the potential transmission path. The fitness is recalculated and updated based on the node permutation sequence of potential transmission paths to obtain the individual best Pbest and the global best Gbest at the current iteration number; Based on the maximum number of iterations The algorithm is compared with the fitness threshold. If the termination condition is met, the potential transmission path corresponding to Gbest is output as the target path. Otherwise, the above operation is returned, the iteration count is incremented by 1, and the iteration process is repeated until the target path is output.

3. The authentication assessment processing method based on a DSN network according to claim 2, characterized in that, Based on the node arrangement sequence of the potential transmission paths at the current iteration number and the individual optimal Pbest and global optimal Gbest, a trust-guided strategy is used to update the particle velocity and the node arrangement sequence of the potential transmission paths to obtain a new node arrangement sequence of the potential transmission paths. The process includes the following steps: The trust transmission rate between each pair of nodes is obtained by combining the interaction analysis between each pair of nodes within a historical time period. Construct the trust transmission matrix of the node sequence of the current path based on the trust transmission rate; The particle velocity for the next iteration is calculated based on the trust transmission matrix, the node arrangement sequence of the potential transmission path under the current iteration number, and the individual optimal Pbest. The node arrangement sequence of the new potential transmission path is then calculated based on the particle velocity under the current iteration number.

4. The authentication assessment processing method based on a DSN network according to claim 3, characterized in that, The trust transmission rate between each pair of nodes is obtained by combining the interaction analysis between each pair of nodes within a historical time period, including the following steps: Collect the number of interactions between node a and node b within a historical time period. The number of interactions includes the number of successful interactions and the total number of interactions. The success rate of the interaction was calculated. Collect behavioral markers for the a-th and b-th nodes within a historical time period; behavioral markers include compliant behavioral markers and non-compliant behavioral markers; The behavioral consistency index between node a and node b is calculated based on behavioral markers. The trust transmission rate is calculated by combining the last corresponding successful interaction time in the historical time period with the interaction success rate and behavior consistency index.

5. The authentication assessment processing method based on a DSN network according to claim 4, characterized in that, After recalculating the fitness based on the node permutation sequence of potential transmission paths and updating it to obtain the individual best Pbest and the global best Gbest for the current iteration number, the following steps are also included: A global priority list is constructed by adjusting each node of each potential transmission path based on the adjacency trust transmission rate, path health response value H, and path fluctuation anomaly index F.

6. The authentication assessment processing method based on a DSN network according to claim 2, characterized in that, A global priority list is constructed by adjusting each node of each potential transmission path based on the adjacency trust transmission rate, path health response value H, and path fluctuation anomaly index F, including the following steps: Obtain the k nodes of the i-th potential transmission path, and calculate the node priority score for the k-th node using the trust transmission rate, path health response value H, and path fluctuation anomaly index F. Based on the node arrangement sequence and node priority score of the current i-th potential transmission path, a local priority list corresponding to each node in the current i-th potential transmission path is established. When the node priority score of the kth node in the current i-th potential path is less than the node priority score of the kth node in the global best Gbest, the kth node in the current i-th potential path is determined to be a node to be adjusted. The local priority list is adjusted by comparing the trust transmission rate of the node to be adjusted through adjacent nodes according to the local priority list, resulting in an updated local priority list. The updated local priority list, which is the potential transmission path with the highest fitness selected from each updated local priority list, is the global priority list.

7. The authentication assessment processing method based on a DSN network according to claim 6, characterized in that, The local priority list is adjusted by comparing the trust transmission rate of the node to be adjusted through its neighboring nodes, resulting in an updated local priority list. The steps include: Obtain the adjacent nodes of each node to be adjusted based on the path topology matrix, and construct the set of adjacent nodes of the node to be adjusted. The adjustment trust transmission rate is calculated by obtaining the historical interaction information between the node to be adjusted and each neighboring node in the set of neighboring nodes; Obtain the first screening condition, and when the adjusted trust transmission rate is greater than or equal to the preset adjustment first threshold, determine the adjacent node as the node to be exchanged; Obtain the second filtering condition and search the node trust evaluation database for the node to be exchanged; when the trust score is greater than or equal to the adjusted second threshold, determine the node to be exchanged as the target exchange node; Swap the node to be adjusted with the target node to generate an adjusted local priority list; check whether the priority scores of the nodes in the adjusted local priority list meet the swap verification rules. If so, return to the above operation and perform the swap process again until the updated local priority list is output.

8. The authentication evaluation processing method based on a DSN network according to claim 7, characterized in that, The exchange verification rule refers to the node priority score of the kth node in the current i-th potential path being greater than or equal to the node priority score of the kth node in the globally optimal Gbest.

9. A storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the steps of the authentication assessment processing method based on a DSN network as described in any one of claims 1-8.

Images