Automated implementation of tor network penetration and anonymous tracing methods based on computer models

Abstract

The application relates to the technical field of computer network security, in particular to a TOR network penetration and anonymous traceability method based on computer model automatic implementation, which comprises the following steps: monitoring TOR node and link state, analyzing delay characteristics to generate optimization parameters, extracting penetration demand to optimize routing, dynamically compensating errors to generate control instructions, adjusting paths to compare consistency, and outputting anonymous state labels. In the application, by acquiring node distribution in real time and dynamically analyzing link delay difference, the network penetration path can be timely identified and optimized, the real-time adaptability of path selection is improved, by continuously compensating routing errors, the influence of path deviation on anonymous consistency is reduced, by combining with periodic collection of path consistency change trend, delay fluctuation in the penetration process is reduced, the anonymous path convergence rate and penetration path matching precision are improved, the anonymous state is continuously maintained under a complex network environment, and the anonymous security and path stability in the data transmission process are enhanced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer network security, and in particular to a method for automatically implementing TOR network penetration and anonymous tracing based on a computer model. BACKGROUND

[0002] The technical field of computer network security involves protecting computer systems from various network attacks and threats, ensuring the confidentiality, integrity, availability of data, and stability and reliability of network services. This technical field includes various key security technologies and methods, including encryption technology, identity authentication, access control, firewall technology, intrusion detection and defense, virtual private network (VPN), etc. The purpose is to ensure that data transmission is not threatened and invaded by external threats, and to prevent information leakage and tampering in network communication. With the rapid development of the Internet and the Internet of Things, computer network security technology is facing increasingly complex attack forms, and protecting user privacy and anonymity has become one of the research focuses in this field.

[0003] Among them, the traditional TOR network penetration and anonymous tracing method refers to using TOR network to realize the anonymization of user network activities, ensuring that the online identity and location of the user are not tracked and exposed. The traditional method hides the user's real IP address by establishing multiple relay nodes and ensures the security of communication content through encryption technology. However, this method faces problems such as high network penetration difficulty, high delay, and incomplete anonymity in actual application. In view of the technical problems, the subject proposes a method for automatically implementing TOR network penetration and anonymous tracing, which mainly optimizes the penetration process of TOR network through automation technology, reduces network delay, and improves penetration efficiency, while anonymizing the tracing path. The traditional solution relies on manual configuration and fixed relay nodes, and cannot dynamically adapt to changes in complex network environments.

[0004] The existing technology relies on pre-set relay nodes for path construction, and the node selection process cannot real-time perceive node load changes, which can easily cause path delay to increase when the network topology dynamically changes. In addition, under the condition of fixed path, the penetration control parameters lack a dynamic adjustment mechanism, which makes it difficult to correct the path deviation problem in time under the condition of node load mutation or link delay fluctuation. For example, in the scenario of high node load or link quality deterioration, the existing path is difficult to maintain the consistency of the anonymous path, which can easily expose the abnormal connection of the node and cause the anonymous path to fail, and cannot effectively adapt to complex and variable network environments, affecting the continuity of the overall network penetration efficiency and anonymity. SUMMARY

[0005] The purpose of the present application is to solve the shortcomings in the prior art and to propose a method for automatically implementing TOR network penetration and anonymous tracing based on a computer model.

[0006] In order to achieve the above object, the present application adopts the following technical scheme: the TOR network penetration and anonymous tracing method based on computer model automatic implementation includes the following steps:

[0007] S1: obtain TOR network node distribution data, real-time monitor network link delay characteristics and transit node load state information, generate link optimization parameters by analyzing the differential characteristics of link delay fluctuation events and combining the multi-path transmission delay ratio index;

[0008] S2: based on the link optimization parameters, extract the penetration demand direction, optimize the path by screening the dynamic routing capability of the transit node, and obtain the penetration control command;

[0009] S3: according to the penetration control command, perform sequence parallel optimization of distributed transit node constraints, analyze the routing error distribution characteristics in the penetration process, combine the amplitude limiting protection algorithm, and dynamically compensate the routing error to obtain the amplitude limiting optimized routing penetration control instruction;

[0010] S4: call the amplitude limiting optimized routing penetration control instruction, compare the consistency of real-time penetration path and target anonymous path, if there is deviation, adjust the penetration control parameter, correct the transit node connection feedback, and obtain the routing penetration control command set.

[0011] As a further scheme of the present application, the link optimization parameters include link delay fluctuation range, multi-path delay ratio threshold, link fluctuation frequency range, the penetration control command includes penetration demand direction, dynamic routing optimization path, routing distribution consistency, the amplitude limiting optimized routing penetration control instruction includes routing error distribution characteristics, distributed transit node constraint condition, amplitude limiting protection algorithm optimization result, and the routing penetration control command set includes penetration control parameter adjustment amount, transit node feedback correction amount, and anonymous path consistency correction amount.

[0012] As a further scheme of the present application, the link optimization parameter acquisition step is specifically:

[0013] S111: obtain TOR network node distribution data, extract link delay characteristics and transit node load state information, generate link configuration optimization index by analyzing the differential characteristics of link delay fluctuation events and combining the multi-path transmission delay ratio index;

[0014] S112: based on the link configuration optimization index, detect the link delay event, analyze the influence law of differential link delay on penetration efficiency, identify the key delay characteristic change trend, and generate the link initial parameter group;

[0015] S113: According to the link initial parameter set, the matching relationship between the delay characteristics and the transit node load capacity is combined to evaluate the penetration demand direction and generate link optimization parameters.

[0016] As a further scheme of the present application, the obtaining step of the penetration control command is specifically:

[0017] S211: Based on the link optimization parameters, the instantaneous routing response curve of the transit node is extracted, the penetration demand direction is identified, the delay characteristics are identified, the penetration interval is determined, and the penetration mapping value set is obtained.

[0018] S212: The penetration mapping value set is called to monitor the response characteristics of the penetration demand node, analyze the load loss distribution and routing stability performance of the transit node in the penetration process, filter the optimal penetration node through the load loss amplitude, and obtain the penetration steady-state node set.

[0019] S213: According to the penetration steady-state node set, the load distribution characteristics and routing transmission efficiency consistency sample of the penetration node are extracted, the load loss difference and routing expansion degree of the corresponding node are combined, the routing response optimization index is calculated, and the penetration control command is obtained.

[0020] As a further scheme of the present application, the obtaining step of the amplitude-limited optimized routing penetration control instruction is specifically:

[0021] S311: According to the penetration control command, the distributed transit node data on the path is extracted and converted into a transit node sequence matrix, whether the node routing jump is within the limit range is judged according to the routing interval, and the routing constraint balance is obtained.

[0022] S312: The routing constraint balance is called to extract the dynamic error distribution characteristics of the transit node, the routing and load error are normalized respectively, the error fluctuation interval is identified, and the dynamic offset distribution coefficient set is obtained.

[0023] S313: According to the dynamic offset distribution coefficient set, the distributed transit node constraint sequence is combined to extract the routing coupling value, load adjustment amplitude and control update frequency of the node, identify the node routing adjustment interval, calculate the routing control strength index, configure the node with the optimal value as the control basis, and generate the amplitude-limited optimized routing penetration control instruction.

[0024] As a further scheme of the present application, the obtaining step of the routing penetration control command set is specifically:

[0025] S411: Call the limited optimization of the route penetration control instruction, compare the consistency of the current penetration path and the target anonymous path, judge by the main route angle deviation, the secondary route amplitude offset and the route graph trend deviation, and obtain the anonymous path consistency deviation value;

[0026] S412: According to the anonymous path consistency deviation value, adjust the current penetration control parameter group, analyze the difference of the route output amplitude and the load loss under the differentiated parameter adjustment result, calculate the route deviation adjustment metric value, select the control parameter combination, and obtain the optimal penetration control adjustment amount;

[0027] S413: Call the optimal penetration control adjustment amount, combine the current relay node feedback vector for correction, reconstruct the matching relationship between the differentiated route component amplitude and the controller response, optimize the penetration adjustment sequence, and obtain the route penetration control command set.

[0028] As a further scheme of the application, the method further comprises a step S5:

[0029] S5: Based on the route penetration control command set, collect the consistency change of the penetration path and the target anonymous path in the continuous period, analyze whether the change trend is in a convergence state, determine the stability period after the penetration execution, and output the anonymous maintenance state label;

[0030] The anonymous maintenance state label comprises an anonymous path consistency, a penetration adjustment stability identifier, and a load loss consistency.

[0031] As a further scheme of the application, the acquisition step of the anonymous maintenance state label is specifically:

[0032] S511: Based on the route penetration control command set, collect the link end penetration path change and the anonymous path change output by the relay node instantaneous state model, extract the anonymous path consistency and the overlap length, analyze the change amplitude of the overlap length in the period, and generate a period overlap amplitude sequence;

[0033] S512: Call the period overlap amplitude sequence, extract the overlap amplitude difference sequence in the continuous period, identify the stable trend according to the difference value change polarity, combine the penetration adjustment stability threshold, judge whether the overlap change converges to a single trend, and obtain the penetration fusion stable trend value;

[0034] S513: According to the penetration fusion stable trend value, identify the coupling relationship between the route distribution characteristics and the relay node dynamic error in the stable period, and combine the anonymous path consistency length ratio to output the anonymous maintenance state label.

[0035] Compared with the prior art, the application has the advantages and positive effects that:

[0036] In the application, by acquiring node distribution information in real time and dynamically analyzing link delay difference characteristics, the network penetration path can be effectively identified and optimized, the transmission path direction is screened in combination with the transit node load state, the path selection has stronger real-time adaptability, the influence of path deviation on anonymity consistency is reduced through continuous compensation and adjustment of route error distribution, the continuity of path stability is further improved, the network delay fluctuation influence in the penetration process is effectively reduced through periodic collection of path consistency change trend and dynamic adjustment of node connection parameters, the anonymity path convergence rate and penetration path matching accuracy are improved, the anonymous state can be continuously maintained in the complex network environment, and the anonymity security and path stability in the data transmission process are improved. BRIEF DESCRIPTION OF DRAWINGS

[0037] Figure 1 It is a main step schematic diagram of the application;

[0038] Figure 2 It is a link optimization parameter acquisition flowchart in the application;

[0039] Figure 3 It is a penetration control command acquisition flowchart in the application;

[0040] Figure 4 It is a route penetration control instruction acquisition flowchart after amplitude optimization in the application;

[0041] Figure 5 It is a route penetration control command set acquisition flowchart in the application;

[0042] Figure 6 It is an anonymous maintenance state label acquisition flowchart in the application. DETAILED DESCRIPTION

[0043] In order to make the purpose, technical scheme and advantages of the application more clear, the application is further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the application, and are not used to limit the application.

[0044] In the description of the application, it should be understood that the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only used to facilitate the description of the application and simplify the description, and therefore cannot be understood as indicating or implying that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as limiting the application. In addition, in the description of the application, the meaning of "a plurality of" is two or more, unless otherwise specifically limited.

[0045] Embodiment one

[0046] Referring to Figure 1 The present application provides a technical solution: an automatic TOR network penetration and anonymous tracing method based on a computer model, comprising the following steps:

[0047] S1: Obtain TOR network node distribution data, monitor network link delay characteristics and transit node load state information in real time, generate link optimization parameters by analyzing the differential characteristics of link delay fluctuation events and combining the multi-path transmission delay ratio index;

[0048] S2: Based on the link optimization parameters, extract the penetration demand direction, optimize the path by screening the dynamic routing capability of the transit node, and obtain the penetration control command;

[0049] S3: According to the penetration control command, perform sequence parallel optimization of distributed transit node constraints, analyze the routing error distribution characteristics in the penetration process, combine the limiting protection algorithm, and dynamically compensate the routing error to obtain the limiting optimized routing penetration control instruction;

[0050] S4: Call the limiting optimized routing penetration control instruction, compare the consistency of the real-time penetration path and the target anonymous path, if there is a deviation, adjust the penetration control parameters, correct the transit node connection feedback, and obtain the routing penetration control command set;

[0051] S5: Based on the routing penetration control command set, collect the consistency changes of the penetration path and the target anonymous path in the continuous period, analyze whether the change trend is in a convergent state, determine the stability period after the penetration is executed, and output the anonymous maintenance state label.

[0052] The link optimization parameters include link delay fluctuation range, multi-path delay ratio threshold, and link fluctuation frequency range. The penetration control command includes penetration demand direction, dynamic routing optimization path, and routing distribution consistency. The limiting optimized routing penetration control instruction includes routing error distribution characteristics, distributed transit node constraint conditions, and limiting protection algorithm optimization results. The routing penetration control command set includes penetration control parameter adjustment amount, transit node feedback correction amount, and anonymous path consistency correction amount. The anonymous maintenance state label includes anonymous path consistency, penetration adjustment stability identifier, and load loss consistency.

[0053] Referring to Figure 2 The acquisition step of the link optimization parameters is specifically:

[0054] S111: Obtain the TOR network node distribution data, extract the link delay characteristics and transit node load state information, generate the link configuration optimization index by analyzing the differential characteristics of link delay fluctuation events and combining the multi-path transmission delay ratio index;

[0055] By scanning the global Tor network nodes, the real-time topology and connection state of the entry node (EntryNode), the relay node (MiddleNode) and the exit node (ExitNode) are obtained, for example, five main relay nodes A, B, C, D and E are obtained, each node records its current active connection number, CPU utilization, memory occupancy and network I / O rate, the transmission path of each active connection is detected in time, the time stamp of the data packet from the source end to the target end is recorded, the two-way transmission delay of each data packet is calculated, and the time stamp of all nodes is calibrated through the network clock synchronization protocol (such as NTP) to ensure the accuracy of the delay calculation, the link delay characteristics and the transit node load state information of each node are extracted, such as the average delay of the link from node A to node B is 80ms, the peak delay reaches 150ms, and the average load rate of node B is 70%, the peak load reaches 95%, the differential characteristics of the link delay fluctuation event are analyzed, for example, set the delay fluctuation detection threshold value as 20ms, when the delay of a certain link changes more than the threshold value within 1 minute, it is identified as a delay fluctuation event, for example, the delay of link A-B suddenly jumps from 80ms to 110ms, the difference is 30ms, which exceeds the threshold value, so it is identified as a fluctuation event, combined with the multi-path transmission delay ratio index, its calculation method is that for the same source to destination multiple available paths, the average transmission delay of each path is calculated, then the ratio of the average transmission delay of each path to the minimum average transmission delay of all paths is taken as the delay ratio of the path, for example, the delay of path 1 is 100ms, the delay of path 2 is 120ms, and the delay of path 3 is 150ms, then the delay ratio of path 1 is 1, the delay ratio of path 2 is 1.2, and the delay ratio of path 3 is 1.5, and the link configuration optimization index is generated.

[0056] S112: Based on the link configuration optimization index, the link delay event is detected, the influence law of differential link delay on penetration efficiency is analyzed, the key delay characteristic change trend is identified, and the initial parameter group of the link is generated.

[0057] Based on the link configuration optimization index, the link delay event occurring in the network is accurately detected. In specific implementation, the delay event detection threshold is set. When the link delay exceeds 1.5 times of the historical average delay or the absolute value exceeds 100 ms, it is determined as a delay event. For example, the average delay of a certain link is 70 ms. If the instantaneous delay reaches 120 ms, because 120 ms exceeds 1.5 times (105 ms) of 70 ms, it is determined as a delay event. The influence of the differentiated link delay on the penetration efficiency of the Tor network is analyzed in depth. The penetration efficiency is quantified by the ratio of the number of successful anonymous connections established through the Tor network to the total number of anonymous connections attempted per unit time. If the delay of a certain path increases by 50 ms, resulting in a decrease in penetration success rate from 95% to 80%, the influence of the delay on the penetration efficiency is recorded as a negative significant influence. The key delay characteristic change trend is identified. For example, it is found through observation that when the CPU utilization of a certain relay node exceeds 90% for 5 minutes, the delay of all links through the node generally increases by more than 30%, which indicates that high CPU utilization is the key feature leading to delay increase. The link initial parameter group is generated.

[0058] S113: According to the link initial parameter group, the matching relationship between the delay characteristic and the load capacity of the relay node is combined to evaluate the penetration demand direction, and the link optimization parameter is generated.

[0059] According to the link initial parameter group, the matching relationship between the delay characteristic and the load capacity of the relay node is combined to evaluate the penetration demand direction of the Tor network. The specific evaluation process is to divide the delay characteristic into three intervals: “low delay (<50 ms)”, “medium delay (50-150 ms)”, and “high delay (>150 ms)”. At the same time, the load capacity of the relay node is divided into three intervals: “low load (<40%)”, “medium load (40%-70%)”, and “high load (>70%)”. For example, the link initial parameter group shows that the delay is 60 ms (medium delay), and the load rate of the associated relay node is 30% (low load). Secondly, through the pre-set matching rule, for example, when the target anonymous service request originates from the European region and the expected traffic is large, the penetration demand direction is determined to preferentially select low-delay and low-to-medium-load relay nodes. If it is actually monitored that most of the existing links are high-delay and high-load, it is evaluated that the current penetration capacity and demand are not matched, and the link optimization parameter is generated.

[0060] Please refer to Figure 3 , the acquisition step of the penetration control command is specifically:

[0061] S211: Based on the link optimization parameter, the instantaneous routing response curve of the relay node is extracted, the penetration demand direction is identified, the penetration interval is determined in combination with the delay characteristic identification reference, and the penetration mapping value group is obtained.

[0062] Based on the link optimization parameters, the instantaneous routing response curve of each relay node in the Tor network is extracted, which represents the relationship between the node routing table entry update speed and the routing convergence time under a given input traffic. For example, apply a test traffic of 100 Mbps to node X, record that its routing table entries are updated by 50 within 1 second, and the routing convergence time is 200 ms. Identify the current penetration demand direction, for example, by analyzing the geographical location of user requests, the type of target service (such as streaming or file download), and the expected anonymity level, determine that the current main penetration direction is large-flow anonymous communication in North America, and identify the delay characteristic reference, which is set as follows: for links with a delay of less than 50 ms, identify as "low-delay links"; for links with a delay between 50 ms and 150 ms, identify as "medium-delay links"; for links with a delay higher than 150 ms, identify as "high-delay links". Determine the penetration interval, for example, if the current penetration demand direction is low-delay anonymous communication, preferentially select the interval with low-delay links and a relay node load rate lower than 50% as the penetration interval, and obtain the penetration mapping value group, which is a multi-dimensional array.

[0063] S212: Call the penetration mapping value group, monitor the response characteristics of the penetration demand node, analyze the load loss distribution and routing stability performance of the relay node in the penetration process, select the optimal penetration node through the load loss amplitude, and obtain the penetration steady-state node set;

[0064] The call penetration mapping value group monitors the response characteristics of the penetration demand node, which is evaluated by tracking the complete time from the user request to receiving the anonymous service response, for example, after an anonymous file download request is issued, the time taken from the request to the start of the file download is recorded as 150ms, the load loss distribution and routing stability performance of the transit node in the penetration process are analyzed, the load loss distribution is quantified by calculating the percentage decrease in data throughput due to processing delay, queue overflow, etc. when the node receives and forwards data packets, for example, the input throughput of a certain node is 100Mbps, and the output throughput is 95Mbps, then the load loss is 5%, the routing stability performance is evaluated by the routing table item jitter frequency and routing convergence time, for example, a certain node updates the routing table item 10 times in 1 minute, and each convergence time is within 200ms, which indicates that its routing stability is good, the optimal penetration node is selected by the load loss amplitude, the load loss amplitude screening reference is set, when the load loss amplitude of the node is less than 5%, it is considered that the node is the optimal node, if the load loss amplitude is between 5% and 10%, it is considered that the node is suboptimal, if it is higher than 10%, it is considered that the node is not suitable, for example, the load loss amplitudes of existing nodes A, B and C are 3%, 6% and 12% respectively, according to the screening reference, node A is screened as the optimal penetration node, and the penetration steady-state node set is obtained.

[0065] S213: According to the penetration steady-state node set, the load distribution characteristics and routing transmission efficiency consistency sample of the penetration node are extracted, combined with the load loss difference and routing expansion degree of the corresponding node, and the formula is used:

[0066]

[0067] The routing response optimization index is calculated, and the penetration control command is obtained;

[0068] Wherein, N represents the routing response optimization index, L i represents the load loss difference of the i-th penetration node, R i represents the routing expansion degree value of the i-th penetration node, and n represents the number of nodes participating in this round of calculation in the penetration steady-state node set;

[0069] According to the penetrating steady-state node set, the load distribution characteristics and routing transmission efficiency consistency samples of the penetrating nodes are extracted. The load distribution characteristics refer to the fluctuation of the node processing traffic within a period of time. For example, the load of node A fluctuates between 30% and 60% within the past 1 hour, and the average load is 45%. The routing transmission efficiency consistency sample refers to the stability performance of the data packet transmission success rate and average delay of the same node at different time periods or different traffic modes. For example, the transmission success rate of node A is 99% during the morning peak period, and the average delay is 80 ms. The transmission success rate is 99.5% during the night low period, and the average delay is 75 ms, which indicates that the transmission efficiency consistency is high. Combined with the load loss difference and routing expansion degree of the corresponding node, the load loss difference refers to the difference between the load input and output of the node within a unit of time. For example, the input traffic of node A is 100 MB / s, and the output traffic is 98 MB / s, so the load loss difference is 2 MB / s. The routing expansion degree value refers to the product of the number of reachable paths and the average path cost from the node. For example, the number of reachable paths of node A is 10, and the average path cost is 5 (number of hops), so the routing expansion degree is 50. The formula is The routing response optimization index is calculated to obtain the penetrating control command.

[0070] wherein N represents the routing response optimization index, which is used to quantify the optimization degree of the current routing response. The smaller the value, the higher the optimization degree and the better the routing efficiency.

[0071] L i represents the load loss difference of the i-th penetrating node, and its unit is MB / s. The larger the value, the worse the node load capacity and the greater the traffic processing loss.

[0072] represents the square root of the routing expansion degree value of the i-th penetrating node. The larger the value, the better the node routing reachability and connection breadth.

[0073] represents a kind of composite relationship between the load loss of the node and its routing expansion degree, which aims to evaluate the ability of the node to process load while ensuring routing coverage.

[0074] represents the average value of the load loss and the routing expansion degree, which is used for comparison with the composite relationship term to measure the balance of the two.

[0075] represents the summation of all participating nodes for calculation, which summarizes the composite performance indicators of all nodes.

[0076] represents the sum of the load loss difference and the routing expansion degree of all participating computing nodes, serving as the denominator for normalization processing to ensure that the exponential result is within a reasonable range;

[0077] The advantage of the formula is that by comprehensively considering the load loss difference L i and the routing expansion degree R i of the node, the actual performance of the node can be more comprehensively evaluated. Traditional methods only focus on a single indicator, but this formula can identify nodes that can effectively handle loads and have good network reachability by comparing the product and average value of the two, avoiding the selection of nodes with good routing reachability but poor load handling capacity, or nodes with strong load handling capacity but low routing reachability, thereby achieving more accurate and efficient penetration node selection in the overall method and optimizing the performance of anonymous communication in the Tor network.

[0078] To perform the calculation, it is assumed that three nodes are selected from the penetration steady-state node set for calculation, namely node 1, node 2, and node 3. The load loss difference and routing expansion degree values are shown in Table 1.

[0079] Table 1: Penetration steady-state node set data table

[0080]

[0081] As shown in Table 1, the load loss difference of node 1 is 2.5 MB / s, and the routing expansion degree is 64; the load loss difference of node 2 is 3.0 MB / s, and the routing expansion degree is 49; the load loss difference of node 3 is 2.0 MB / s, and the routing expansion degree is 81.

[0082] Substitute the data in Table 1 into the formula for calculation:

[0083] For node 1:

[0084] For node 2:

[0085] For node 3:

[0086] L3+R3=2.0+81=83.0;

[0087] Calculate the sum of the numerator and denominator:

[0088] Calculate the routing response optimization index N:

[0089] The result shows that the calculated routing response optimization index is about 0.2072, and the smaller the routing response optimization index is, the higher the optimization degree of the routing response is, and the better the current routing efficiency is. The index will be an important part of the penetration control command and will be used to guide subsequent routing selection and adjustment.

[0090] Referring to Figure 4 The step of obtaining the routing penetration control command after the amplitude optimization is specifically:

[0091] S311: According to the penetration control command, extract the distributed relay node data on the path, and convert it into a relay node sequence matrix. According to the routing interval, determine whether the node routing jump is within the limit range, and obtain the routing constraint balance quantity.

[0092] According to the penetration control command, extract the distributed relay node data on the current routing path. The data includes the ID, geographic location, current connection number, bandwidth utilization rate and hop number of each relay node, etc. For example, a penetration path consists of node A, node B and node C, and their data are as follows: node A (ID: N101, location: New York, connection number: 500, bandwidth utilization rate: 70%, hop number: 1), node B (ID: N102, location: London, connection number: 650, bandwidth utilization rate: 85%, hop number: 2), node C (ID: N103, location: Tokyo, connection number: 400, bandwidth utilization rate: 60%, hop number: 3). Convert the data into a relay node sequence matrix. Each row of the matrix represents a relay node, and each column represents a specific attribute value, for example: According to the routing interval, determine whether the node routing jump is within the limit range. The routing interval refers to the hop number or geographic distance between two consecutive relay nodes. The limit range is determined according to the empirical value and network topology characteristics. For example, it is stipulated that the hop number interval between two consecutive relay nodes should not exceed 3 hops, or the geographic distance should not exceed 5000 kilometers. If the hop number from node A to node B is 1 and the hop number from node B to node C is 1, both are within the limit range of 3 hops, the routing constraint balance quantity is obtained.

[0093] S312: Call the routing constraint balance quantity, extract the dynamic error distribution characteristics of the relay node, normalize the routing and load error respectively, identify the error fluctuation interval, and obtain the dynamic offset distribution coefficient array.

[0094] Call routing constraint balance quantity, extract the dynamic error distribution characteristics of the transit node, which is quantified by monitoring the error rate and delay jitter of the node when processing routing requests, for example, a certain transit node processed 10000 routing requests in 1 hour, of which 10 requests occurred errors, and the average delay jitter was 5ms, the routing error and load error were normalized respectively, the routing error normalization processing is to divide the original routing error rate by the maximum error rate (for example 100%), so that the numerical range is between 0 and 1, for example, the routing error rate is 0.1%, and the normalized value is 0.001; The load error normalization processing is to divide the original load fluctuation amplitude by the maximum load fluctuation amplitude (for example 100%), so that the numerical range is between 0 and 1, for example, the load fluctuation amplitude is 10%, and the normalized value is 0.1, identify error fluctuation interval, which is defined as setting the dynamic error "low fluctuation" (0-0.2), "medium fluctuation" (0.2-0.5) and "high fluctuation" (0.5-1) three intervals on the basis of normalized error value, for example, when the normalized routing error of a certain node is 0.15 and the normalized load error is 0.25, the routing error falls in the low fluctuation interval, and the load error falls in the medium fluctuation interval, get the dynamic offset distribution system, which contains the normalized routing error and load error of each transit node, and the error fluctuation interval information of the node.

[0095] S313: According to the dynamic offset distribution system, combined with the distributed transit node constraint sequence, extract the routing coupling value, load adjustment amplitude and control update frequency of the node, identify the routing adjustment interval, use the formula:

[0096]

[0097] Calculate the routing control strength index, and configure the node with the optimal value as the control basis to generate the amplitude limited optimized routing penetration control instruction;

[0098] Wherein, P represents the routing control strength index, G i represents the load value of the i-th node, G avg represents the load average value of all nodes, H represents the control update frequency coefficient, Δt represents the routing adjustment interval time, and m represents the number of nodes.

[0099] According to the dynamic offset distribution system, combined with the distributed relay node constraint sequence, the sequence contains the specific constraint values of each relay node in terms of route jump, bandwidth limit and connection upper limit, for example, the jump constraint of node A is 3 hops at most, the bandwidth upper limit is 1 Gbps, and the connection upper limit is 1000 connections, the system extracts the route coupling value, load adjustment amplitude and control update frequency of the node, the route coupling value refers to the frequency and consistency degree of route information exchange between the node and the adjacent node, for example, node A exchanges route information with adjacent nodes 10 times per second, and the consistency of its route table with adjacent nodes reaches 98%; the load adjustment amplitude refers to the ability range of the node to dynamically adjust its processing traffic, for example, the load adjustment amplitude of node A is positive and negative 20%, that is, it can temporarily increase or decrease 20% of the traffic processing capacity, the control update frequency refers to the frequency of the system updating control instructions for the node, for example, the control update frequency of node A is once every 10 seconds, identify the node route adjustment interval, which is the shortest time interval for the system to adjust the route strategy of a single node, for example, the route adjustment interval of node A is 5 seconds, and the formula is used. Calculate the route regulation intensity index, and configure the node with the optimal value as the control basis to generate the route penetration control instruction after amplitude optimization.

[0100] Wherein, P represents the route regulation intensity index, which is used to quantify the necessity and intensity of route regulation of a single node, and the larger the value is, the more the node needs route regulation, and the greater the regulation intensity is;

[0101] G i represents the load value of the ith node, and its unit is %, which indicates that the current load of the node accounts for a percentage of its maximum capacity;

[0102] G avg represents the average load value of all nodes, and its unit is %, which is used to measure the deviation degree of the load of a single node relative to the overall network load;

[0103] |G i -G avg |represents the absolute difference between the load value of a single node and the average load value of all nodes, reflecting the abnormal degree of the load of the node;

[0104] represents the sum of squares of the difference between the load values of all nodes and the average value, which is used to measure the dispersion degree of the load distribution of the entire network;

[0105] represents the load standard deviation, which is used as the denominator for normalization processing, so that the P value is not affected by the absolute value of the load;

[0106] H represents the control update frequency coefficient, which is a dimensionless constant, used to adjust the influence weight of control update frequency on the regulation intensity index. For example, when the network environment changes frequently, a higher H value can be set to make the regulation intensity index more sensitive to control update frequency. H is set to an integer between 1 and 5, and H is 2 here;

[0107] Δt represents the routing adjustment interval, with a unit of seconds, representing the shortest time interval for adjusting the routing strategy of the node;

[0108] By introducing the load standard deviation as the denominator, the normalization of the node load deviation is realized, so that the routing regulation intensity index P can better reflect the relative degree of individual node load anomaly, rather than the absolute size. Combined with the control update frequency coefficient H and the routing adjustment interval Δt, the regulation intensity not only considers load balancing, but also integrates the regulation urgency in the time dimension, so that the routing adjustment can be more intelligent in the overall method, and the nodes with large load deviation and frequent update are prioritized for regulation, ensuring the stability and efficiency of the anonymous network;

[0109] To calculate, assume that there are 3 nodes participating in routing regulation, and their load values, control update frequency coefficients and routing adjustment intervals are shown in Table 2.

[0110] Table 2: Routing regulation related parameters

[0111]

[0112] As shown in Table 2, the load value of node 1 is 80%, and the routing adjustment interval is 5 seconds; the load value of node 2 is 60%, and the routing adjustment interval is 10 seconds; the load value of node 3 is 70%, and the routing adjustment interval is 8 seconds, and the control update frequency coefficient H is set to 2.

[0113] First, calculate the load average G of all nodes avg :

[0114] Next, calculate the denominator

[0115] Now calculate the routing regulation intensity index P of each node: for node 1:

[0116] For node 2:

[0117] For node 3:

[0118] The results show that the routing regulation intensity index of node 1 is the highest (about 0.2829), indicating that the load deviates from the average value greatly and the adjustment interval is short, and it needs to be regulated most; the regulation intensity index of node 2 is the second (about 0.1414); the regulation intensity index of node 3 is 0, indicating that its load is at the average level, and no special regulation is needed at present. The node configuration with the optimal value (for example, the node with a P value greater than 0.25) is configured as the control basis, and the regulation parameter configuration of node 1 is configured as an important part of the routing penetration control instruction after the amplitude limiting optimization, and the instruction will include specific parameters such as load balancing strategy, routing path adjustment suggestion and control update frequency of node 1, so as to ensure that the node performs optimal routing penetration control within the amplitude limiting range.

[0119] Please refer to Figure 5 The acquisition step of the routing penetration regulation command set is specifically:

[0120] S411: Call the routing penetration control instruction after amplitude limiting optimization, compare the consistency of the current penetration path and the target anonymous path, judge by the main route angle deviation, the secondary route amplitude offset and the route map trend deviation, use the formula:

[0121]

[0122] Get the anonymous path consistency deviation value;

[0123] Wherein, Q represents the anonymous path consistency deviation value, Δθ represents the main route angle deviation, ΔY represents the secondary route amplitude offset, ΔS k represents the kth route map trend deviation, and M represents the total number of deviations in the route map;

[0124] Call the routing penetration control instruction after amplitude limiting optimization, compare the consistency of the current penetration path and the target anonymous path. The current path is the actual transit node sequence of the user to the target service, and the target path is the optimal path preset according to the anonymous demand. The consistency is judged by three indexes: main route angle deviation (path main direction angle difference), secondary route amplitude offset (local branch number and length difference), and route map trend deviation (overall evolution trend difference), for example, by comparing the long-term change curves of node load, delay and other indexes on the two paths, it is judged whether the trend is consistent, and the formula Get the anonymous path consistency deviation value;

[0125] wherein Q represents an anonymity path consistency deviation value, the larger the value, the greater the deviation of the current penetration path from the target anonymity path, the worse the anonymity preservation ability; Δθ represents the main routing angle deviation, its unit is degree, which quantifies the angle deviation of the current path and the target path in the main routing direction; ΔY represents the secondary routing amplitude offset, which is a dimensionless value, representing the relative difference of the number and length of the current path branch routing and the target path; ΔS k represents the kth routing graph trend deviation, which is a dimensionless value, quantifying the deviation degree of the current path and the target path in a certain routing graph trend, for example, it can be the difference value obtained by comparing the average hop number trend or the average delay trend of the two paths in a certain time period; M represents the total number of deviations in the routing graph, for example, if the load trend, delay trend and hop number trend are compared, then M = 3; represents the simple sum of all routing graph trend deviations;

[0126] represents the square root of the square sum of all routing graph trend deviations, which is used as the denominator for normalization processing, so that the Q value is not affected by the absolute value of the trend deviation;

[0127] The advantage of the formula is that by comprehensively considering the main routing angle deviation Δθ, the secondary routing amplitude offset ΔY and the routing graph trend deviation ΔS k , the consistency of the current penetration path and the target anonymity path can be comprehensively evaluated, overcoming the limitations of single index evaluation, especially by introducing the square root of the square sum of the routing graph trend deviation in the denominator, the normalization of the trend deviation is realized, so that the Q value can better reflect the relative difference of the path consistency, so that the anonymity preservation state can be more accurately judged in the overall system or method, providing a more reliable basis for subsequent routing adjustment.

[0128] In order to calculate, it is assumed that a current penetration path and a target anonymity path are compared, and the following data is obtained: deviation type value main routing angle deviation Δθ (degree) is 15, secondary routing amplitude offset ΔY is 0.2 routing graph trend deviation ΔS1 (load trend deviation) is 0.1 routing graph trend deviation ΔS2 (delay trend deviation) is 0.05 routing graph trend deviation ΔS3 (hop number trend deviation) is 0.02;

[0129] wherein the routing graph trend deviation M = 3;

[0130] The above data is substituted into the formula for calculation: the numerator calculation is:

[0131] The denominator calculation is:

[0132] Calculate the anonymous path consistency deviation value Q:

[0133] The results indicate that the anonymity path consistency deviation is approximately 135.3. When the Q value is below the preset threshold of 20, it indicates that the current path is highly consistent with the target anonymity path, and anonymity is well maintained. When the Q value is between 20 and 100, it indicates that there is a certain degree of deviation, requiring fine-tuning. When the Q value is above 100, as shown in this example, it indicates that the current path deviates significantly from the target anonymity path, and anonymity is at risk, requiring immediate and substantial adjustments. This anonymity path consistency deviation will serve as the basis for subsequent adjustments to the current penetration control parameter set.

[0134] S412: Based on the anonymous path consistency deviation value, adjust the current penetration control parameter group, analyze the difference between the route output amplitude and load loss under the differentiated parameter adjustment results, calculate the route deviation adjustment metric value, select the control parameter combination, and obtain the optimal penetration control adjustment amount;

[0135] Based on the anonymous path consistency deviation value, the current penetration control parameter set is adjusted. This parameter set includes the weight of routing nodes, path selection priority, and traffic allocation ratio. For example, if the anonymous path consistency deviation value is too high (e.g., 135.3), it indicates a significant deviation in the anonymity of the current path. This will reduce the weight of node B in the current path and increase the priority of backup nodes with better anonymity (e.g., nodes Y and Z). The differences in routing output amplitude and load loss under the differentiated parameter tuning results are analyzed. The routing output amplitude refers to the effective anonymous traffic volume through the path after parameter adjustment. For example, by reducing the weight of node B by 10%, the routing output amplitude of the path is found to decrease from 50Mbps to 40Mbps. The load loss difference refers to the difference in the effective anonymous traffic volume through the path after parameter adjustment. Before and after the adjustment, the load loss of the transit node changes. For example, after the adjustment, the load loss of node B decreases from 5% to 3%. The routing deviation adjustment metric is calculated. This metric quantifies the degree of improvement in the anonymous path consistency deviation value after each parameter adjustment. The calculation method is: the anonymous path consistency deviation value before adjustment minus the anonymous path consistency deviation value after adjustment. For example, if the deviation value before adjustment is 135.3 and after adjustment is 90.5, then the adjustment metric value is 44.8. The combination of control parameters is selected. For example, through multiple iterations of adjustment and metric calculation, it is found that the combination of reducing the weight of node B by 10% and increasing the priority of node Y by 20% reduces the anonymous path consistency deviation value to 90.5 and minimizes the difference in load loss, thus obtaining the optimal penetration control adjustment amount.

[0136] S413: Call the optimal penetration control adjustment amount, combine it with the current relay node feedback vector for correction, reconstruct the matching relationship between the differential routing component amplitude and the controller response, optimize the penetration adjustment sequence, and obtain the routing penetration control command set;

[0137] The optimal penetration control adjustment is invoked. The relay node feedback vector contains real-time data such as load, latency, packet loss rate, and routing table update frequency for each relay node. For example, if the optimal penetration control adjustment suggests reducing traffic to node B, but node B's feedback vector shows a sharp drop in its load, the adjustment will be corrected, diverting traffic to nodes with more balanced loads. The matching relationship between the differential routing component amplitude and the controller response is reconstructed. The differential routing component amplitude refers to the proportion of traffic allocated to a specific routing path for different anonymity levels or service types. For example, for services requiring extremely high anonymity... The communication of routing components will have a higher amplitude. The controller response refers to the strength and frequency of the adjustment instructions output by the controller based on the network status and anonymity requirements. For example, when a 20% increase in network latency is detected, the controller response threshold is reduced by 5% to trigger routing adjustments earlier and optimize the penetration adjustment sequence. The penetration adjustment sequence is a series of ordered fine-tuning operation instructions for relay nodes and routing paths. For example, first, the traffic of node B is reduced by 10%, and then the reduced traffic is diverted to nodes Y and Z. Then, the overall anonymity deviation value is checked to obtain the routing penetration control command set.

[0138] Please see Figure 6 The specific steps for obtaining anonymous state tags are as follows:

[0139] S511: Based on the route penetration control command set, collect the changes in the penetration path at the link end and the changes in the anonymous path output by the instantaneous state model of the relay node, extract the consistency and overlap length of the anonymous path, analyze the change amplitude of the overlap length within the period, and generate a periodic overlap amplitude sequence.

[0140] Based on the route penetration control command set, the link-end penetration path change refers to the change in the actual route path and hop count from the user equipment to the Tor network entry node. For example, user A's request changed from passing through the China Telecom network to passing through the China Unicom network in the past hour, and from 5 hops to 6 hops. The anonymous path change output by the transit node instantaneous state model refers to the deviation of the actual path from the preset anonymous path when anonymous data packets are transmitted between transit nodes within the Tor network due to load balancing or node failure. For example, the path that should have passed through nodes XYZ actually passed through XAZ. Anonymous path consistency and overlap length are extracted. Anonymous path consistency refers to the similarity between the actual penetration path and the ideal anonymous path. For example, by calculating the Jaccard similarity coefficient, if the proportion of shared nodes between the two paths reaches 80%, the consistency is high. Overlap length refers to the number of consecutive transit nodes or the shared geographical distance between the actual penetration path and the ideal anonymous path. For example, within a 5-minute period, the overlap length of an anonymous path fluctuates from the initial 5 nodes to 3 nodes, and then back to 4 nodes, with a change range of 2 nodes, generating a periodic overlap amplitude sequence.

[0141] S512: Call the periodic overlap amplitude sequence, extract the overlap amplitude difference sequence within the continuous period, identify the stable trend based on the polarity of the difference change, combine with the penetration adjustment stability threshold, determine whether the overlap change converges to a single trend, and obtain the penetration fusion stable trend value.

[0142] The overlapping amplitude sequence of the cycles is called to extract the overlapping amplitude difference sequence within consecutive cycles. This sequence records the difference between the overlapping amplitudes of two adjacent cycles. For example, if the overlapping amplitude of the previous cycle is 2 nodes and the overlapping amplitude of the current cycle is 1 node, the difference is -1 node. The polarity of the difference indicates the stability trend. The polarity refers to whether the difference is positive or negative. If the difference is positive, it means that the overlapping amplitude is increasing and the anonymity trend is improving; if the difference is negative, it means that the overlapping amplitude is decreasing and the anonymity trend is deteriorating; if the difference is zero, it means that the overlapping amplitude is stable. Combined with the penetration adjustment stability threshold, which is a preset value used to judge the degree of convergence of the overlapping change, for example, the penetration adjustment stability threshold is set to 0.5 nodes. When the absolute value of the overlapping amplitude difference of 5 consecutive cycles is less than 0.5 nodes, the change is considered to have converged. It is judged whether the overlapping change has converged into a single trend, that is, whether the polarity of the overlapping amplitude difference remains consistent in multiple consecutive cycles and its absolute value gradually decreases and falls below the stability threshold, thus obtaining the penetration fusion stable trend value.

[0143] S513: Based on the penetration fusion stable trend value, identify the coupling relationship between the routing distribution characteristics and the dynamic error of the transit node within the stable period, and output the anonymity maintenance state label by combining the anonymous path consistency length ratio.

[0144] Based on the penetration-fusion stability trend value, the coupling relationship between routing distribution characteristics and relay node dynamic errors within the stable period is identified. The stable period refers to a continuous time period during which the penetration-fusion stability trend value is "stable". The routing distribution characteristics refer to the geographical distribution, hop count distribution, and node type distribution of anonymous paths in the network within this stable period. For example, within the stable period, anonymous paths are mainly distributed in Europe and North America, with an average hop count of 4, and are mainly composed of high-bandwidth relay nodes. The coupling relationship of relay node dynamic errors refers to the degree and pattern of mutual influence among dynamic errors such as routing errors, load errors, and latency jitter of relay nodes within this stable period. For example, it is observed that within the stable period, when the routing error rate of a certain relay node increases, its corresponding load error also increases synchronously, indicating that there is a strong coupling relationship between the two. Combined with the anonymous path consistency length ratio, which is the ratio of the actual anonymous path overlap length to the ideal anonymous path length, for example, if the ideal path length is 5 hops and the actual overlap length is 4 hops, then the consistency length ratio is 0.8. Anonymity maintenance status label is output, which is a classification identifier used to describe the current anonymity maintenance status of the Tor network.

[0145] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A method for automated implementation of TOR network penetration and anonymous tracing based on computer models, characterized in that, The method comprises the following steps: S1: acquiring TOR network node distribution data, monitoring network link delay characteristics and transit node load state information in real time, generating link optimization parameters by analyzing the differential characteristics of link delay fluctuation events and combining the multi-path transmission delay ratio index; S2: based on the link optimization parameters, extracting the network penetration demand direction, optimizing the path by screening the dynamic routing capability of the transit node, and obtaining the penetration control command; S3: according to the penetration control command, performing sequence parallel optimization of distributed transit node constraints, analyzing the routing error distribution characteristics in the penetration process, combining the limiting protection algorithm to dynamically compensate the routing error, and obtaining the limiting optimized routing penetration control instruction; S4: calling the limiting optimized routing penetration control instruction, comparing the consistency of the real-time penetration path and the target anonymous path, if there is deviation, adjusting the penetration control parameter, correcting the transit node connection feedback, and obtaining the routing penetration control command set.

2. The method of claim 1, wherein the computer model based automated implementation of a TOR network penetration and anonymous traceability method is characterized by, The link optimization parameters include link delay fluctuation range, multi-path delay ratio threshold, and link fluctuation frequency range. The penetration control command includes penetration demand direction, dynamic routing optimization path, and routing distribution consistency. The limiting optimized routing penetration control instruction includes routing error distribution characteristics, distributed transit node constraint conditions, and limiting protection algorithm optimization results. The routing penetration control command set includes penetration control parameter adjustment amount, transit node feedback correction amount, and anonymous path consistency correction amount.

3. The method of claim 1, wherein the computer model based automated implementation of a TOR network penetration and anonymous traceability method is characterized by, The acquisition step of the link optimization parameters is specifically: S111: acquiring TOR network node distribution data, extracting link delay characteristics and transit node load state information, generating link configuration optimization index by analyzing the differential characteristics of link delay fluctuation events and combining the multi-path transmission delay ratio index; S112: based on the link configuration optimization index, detecting link delay events, analyzing the influence law of differential link delay on penetration efficiency, identifying the trend of key delay characteristics, and generating link initial parameter group; S113: according to the link initial parameter group, combining the matching relationship between delay characteristics and transit node load capacity, evaluating the penetration demand direction, and generating the link optimization parameters.

4. The computer model based automated implementation of a TOR network penetration and anonymous traceback method according to claim 3, wherein, The acquisition step of the penetration control command is specifically: S211: based on the link optimization parameters, extracting the instantaneous routing response curve of the transit node, identifying the penetration demand direction, combining the delay characteristic identification reference to determine the penetration interval, and obtaining the penetration mapping value group; S212: calling the penetration mapping value group, monitoring the response characteristics of the penetration demand node, analyzing the load loss distribution and routing stability performance of the transit node in the penetration process, selecting the optimal penetration node through the load loss amplitude, and obtaining the penetration steady-state node set; S213: according to the penetration steady-state node set, extracting the load distribution characteristics and routing transmission efficiency consistency sample of the penetration node, combining the load loss difference and routing expansion degree of the corresponding node, calculating the routing response optimization index, and obtaining the penetration control command.

5. The computer model based automated implementation of a TOR network penetration and anonymous traceback method according to claim 4, wherein, The acquisition step of the limiting optimized routing penetration control instruction is specifically: S311: According to the penetration control command, the distributed relay node data on the path is extracted and converted into a relay node sequence matrix, and it is judged whether the node route jump is within the limit range according to the routing interval to obtain a routing constraint balance value; S312: The routing constraint balance value is called to extract the dynamic error distribution characteristics of the relay node, and the routing and load error are normalized respectively to identify the error fluctuation interval and obtain a dynamic offset distribution coefficient array; S313: According to the dynamic offset distribution coefficient array, the routing coupling value, load adjustment amplitude and control update frequency of the node are extracted in combination with the distributed relay node constraint sequence, the node routing adjustment interval is identified, the routing control strength index is calculated, the node with the optimal value is configured as the control basis, and the amplitude-limited optimized routing penetration control instruction is generated.

6. The computer model based automated implementation of a TOR network penetration and anonymous traceback method according to claim 5, wherein, The acquisition step of the routing penetration control command set is specifically: S411: The amplitude-limited optimized routing penetration control instruction is called, the consistency of the current penetration path and the target anonymous path is compared, the consistency deviation value of the anonymous path is obtained by judging the main route angle deviation, the secondary route amplitude offset and the route graph trend deviation; S412: According to the anonymous path consistency deviation value, the current penetration control parameter group is adjusted, the routing output amplitude and load loss difference under the differential parameter adjustment result are analyzed, the routing deviation adjustment metric value is calculated, the control parameter combination is selected, and the optimal penetration control adjustment amount is obtained; S413: The optimal penetration control adjustment amount is called in combination with the current relay node feedback vector to correct the matching relationship between the differential routing component amplitude and the controller response, optimize the penetration adjustment sequence, and obtain the routing penetration control command set.

7. The computer model based automated implementation of a TOR network penetration and anonymous traceback method according to claim 1, wherein, The method further comprises the S5 step: S5: Based on the routing penetration control command set, the consistency change of the penetration path and the target anonymous path in the continuous period is collected, whether the change trend is in a convergence state is analyzed, the stability period after penetration execution is determined, and an anonymous maintenance state label is output; The anonymous maintenance state label comprises an anonymous path consistency, a penetration adjustment stability identifier and a load loss consistency.

8. The computer model based automated implementation of a TOR network penetration and anonymous traceback method according to claim 7, wherein, The acquisition step of the anonymous maintenance state label is specifically: S511: Based on the routing penetration control command set, the link end penetration path change and the anonymous path change output by the instantaneous state model of the relay node are collected, the anonymous path consistency and the overlap length are extracted, and the change amplitude of the overlap length in the period is analyzed to generate a period overlap amplitude sequence; S512: The period overlap amplitude sequence is called to extract the overlap amplitude difference sequence in the continuous period, the stable trend is identified according to the difference value change polarity, the penetration adjustment stability threshold is combined, it is judged whether the overlap change converges to a single trend, and a penetration fusion stable trend value is obtained; S513: According to the penetration fusion stable trend value, the routing distribution characteristics and the dynamic error coupling relationship of the relay node in the stable period are identified, and the anonymous path consistency length ratio is combined to output the anonymous maintenance state label.

Images