Intelligent network operation and maintenance method and system with global self-sensing, causal inference and autonomous evolution capabilities
By deploying enhanced telemetry agents and causal field models in the network, the problems of insufficient global awareness and the inability of strategies to evolve autonomously in traditional network operations and maintenance are solved. This enables the network to achieve global self-awareness, causal inference, and autonomous evolution, thereby improving operational efficiency and network stability.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- UNICOM (HENAN) IND INTERNET CO LTD
- Filing Date
- 2026-03-09
- Publication Date
- 2026-04-24
AI Technical Summary
Traditional network operations and maintenance suffer from insufficient global awareness, lack of causal inference in anomaly identification, and inability to autonomously evolve strategies, resulting in untimely handling of network faults, low operational efficiency, and poor network stability.
By deploying enhanced telemetry agents to collect network data across all dimensions, and combining standardized coding, regional aggregation and global fusion processing, a real-time updated global network state view is formed. Dynamic analysis is performed using a causal field model to reconstruct the network state context, generate optimization strategies, and implement configuration adjustments through multi-objective optimization algorithms and distributed execution mechanisms, while updating causal field model parameters and strategy libraries in real time.
It enables comprehensive, real-time, and accurate perception of network operating status, accurately identifies abnormal states, improves operation and maintenance efficiency, enhances network environment adaptability and stability, and reduces operation and maintenance costs.
Smart Images
Figure CN121923988A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of network operation and maintenance technology, and discloses an intelligent network operation and maintenance method and system with global self-awareness, causal inference and autonomous evolution capabilities. Background Technology
[0002] With the rapid development of network technology, network scale is constantly expanding, network topology is becoming increasingly complex, and the diversity and real-time requirements of network services are continuously increasing. Traditional manual and passive operation and maintenance methods are no longer suitable for the current network operation and management needs. Traditional network operation and maintenance often suffers from problems such as single data collection dimensions and insufficient global status perception capabilities. It can only monitor local network data and cannot form a complete global network status view. At the same time, in the network anomaly identification stage, it relies more on correlation analysis rather than causal relationship inference, making it difficult to accurately locate the root cause of the anomaly and its evolution pattern. In addition, the policy library of traditional operation and maintenance systems is mostly statically configured and cannot be autonomously optimized and evolved according to the actual network operation. Faced with a dynamically changing network environment, the adaptability and effectiveness of operation and maintenance policies are greatly reduced, which can easily lead to problems such as untimely network fault handling, low operation and maintenance efficiency, and poor network operation stability. Summary of the Invention
[0003] This invention addresses the problems of insufficient global awareness, lack of causal inference in anomaly identification, and inability of strategies to evolve autonomously in traditional network operations and maintenance. Therefore, this invention provides an intelligent network operations and maintenance method and system with global self-awareness, causal inference, and autonomous evolution capabilities.
[0004] To achieve the above-mentioned technical effects, the technical solution adopted by this invention is: an intelligent network operation and maintenance method with global self-awareness, causal inference, and autonomous evolution capabilities, comprising the following steps:
[0005] S1: Network data is captured in real time by enhanced telemetry agents deployed throughout the network, and the network data is standardized and encoded to obtain enhanced telemetry data; S2: Perform regional aggregation and global fusion processing on the enhanced telemetry data to obtain a real-time updated global network status view; S3: Based on the global state view of the network, dynamic analysis is performed using a causal field model to identify abnormal network states; S4: Based on the spatiotemporal coordinates corresponding to the network anomaly state, reconstruct the network state context at the time of the anomaly by retrieving historical data; S5: Generate a network optimization strategy using a multi-objective optimization algorithm based on the network global state view, the network abnormal state, and the network state context; S6: Based on the network optimization strategy, network configuration adjustments are implemented through a distributed execution mechanism to obtain optimized network status data; S7: Update the causal field model parameters based on the optimized network state data, and optimize the system policy library through the policy evolution mechanism.
[0006] As a preferred embodiment, the standardization coding process in step S1 includes: S101: Parse the header of the data packets in the network data to obtain basic metadata; S102: Based on the basic metadata, perform feature extraction calculations to obtain transmission delay, path hash, and state representation code; S103: Based on the feature calculation results obtained in step S102, perform encoding mapping processing to obtain standardized data blocks; S104: Perform verification and encapsulation processing on the standardized data block to obtain complete enhanced telemetry data.
[0007] As a preferred embodiment, the expression for the causal field model in step S3 is: ; in, For scalar fields in network topology space coordinates x and time coordinate t, The causal field value at the network topology spatial coordinates x′ and time coordinate t′; Let be the partial derivative of the causal field with respect to time, and α be the normalization parameter. For phase parameters, Let λ represent the causal influence strength of the states at different locations x′ and different times t′ in the network on the current location x and the current time t, and let λ be the coupling strength coefficient.
[0008] As a preferred embodiment, the network state context reconstruction in step S4 includes: S401: Retrieve the corresponding hologram encoded data from the distributed spatiotemporal storage module according to the spatiotemporal coordinates; S402: The hologram encoded data is decoded using the inverse Fourier transform algorithm to reconstruct the historical network event sequence; S403: Compare and analyze the historical network event sequence with the network abnormal state to obtain a complete fault evolution view; The hologram-encoded data in step S401 is generated in the following manner: ; in, For spatiotemporal frequency domain holographic coding, Let x, y, z be the three-dimensional physical space coordinates (x, y, z) and time t be the network event occurring at time t, where x, y, z are the three-dimensional physical space coordinates, t is the time when the network event occurs, and k is the time when the network event occurs. x k y k z ω represents the spatial frequency component, and ω represents the time angular frequency.
[0009] As a preferred embodiment, the expression for the multi-objective optimization algorithm in step S5 is: ; Where U(S) is the comprehensive evaluation value of network state, S is the network state vector, n≥5, and s i Let w be the original value of the i-th key performance indicator. i The weight coefficient of the i-th indicator, f i Let be the normalization function for the i-th index.
[0010] This invention also proposes an intelligent network operation and maintenance system with global self-awareness, causal inference, and autonomous evolution capabilities, comprising: An enhanced telemetry data acquisition module is used to capture network data in real time through enhanced telemetry agents deployed in various parts of the network, and to perform standardized encoding processing on the network data to obtain enhanced telemetry data; The network global status view generation module is used to perform regional aggregation and global fusion processing on the enhanced telemetry data to obtain a real-time updated network global status view. The network anomaly state identification module is used to identify network anomalies by performing dynamic analysis based on the global network state view and a causal field model. The network state context reconstruction module is used to reconstruct the network state context at the time of the anomaly by retrieving historical data based on the spatiotemporal coordinates corresponding to the network anomaly state. The network optimization strategy generation module is used to generate a network optimization strategy based on the network global state view, the network abnormal state, and the network state context, using a multi-objective optimization algorithm. The network configuration adjustment execution module is used to implement network configuration adjustment through a distributed execution mechanism according to the network optimization strategy, and obtain optimized network status data. The system strategy optimization module is used to update the causal field model parameters based on the optimized network state data, and to optimize the system strategy library through a strategy evolution mechanism.
[0011] As a preferred embodiment, the enhanced telemetry data acquisition module includes: The basic metadata parsing unit is used to parse the packet headers in the network data to obtain basic metadata; The feature extraction calculation unit is used to perform feature extraction calculations based on the basic metadata to obtain transmission delay, path hash and state characterization code; The encoding mapping processing unit is used to perform encoding mapping processing based on the feature calculation results obtained by the feature extraction calculation unit to obtain standardized data blocks; The verification and encapsulation processing unit is used to perform verification and encapsulation processing on the standardized data block to obtain complete enhanced telemetry data.
[0012] As a preferred embodiment, the expression of the causal field model in the network anomaly identification module is: ; in, For scalar fields in network topology space coordinates x and time coordinate t, The causal field value at the network topology spatial coordinates x′ and time coordinate t′; Let be the partial derivative of the causal field with respect to time, and α be the normalization parameter. For phase parameters, Let λ represent the causal influence strength of the states at different locations x′ and different times t′ in the network on the current location x and the current time t, and let λ be the coupling strength coefficient.
[0013] As a preferred embodiment, the network state context reconstruction module includes: The hologram encoded data retrieval unit is used to retrieve the corresponding hologram encoded data from the distributed spatiotemporal storage module according to the spatiotemporal coordinates; The historical network event sequence reconstruction unit is used to decode the hologram encoded data using an inverse Fourier transform algorithm to reconstruct the historical network event sequence. The fault evolution view generation unit is used to compare and analyze the historical network event sequence with the network abnormal state to obtain a complete fault evolution view. The hologram encoded data is generated in the following manner: ; in, For spatiotemporal frequency domain holographic coding, Let x, y, z be the three-dimensional physical space coordinates (x, y, z) and time t be the network event occurring at time t, where x, y, z are the three-dimensional physical space coordinates, t is the time when the network event occurs, and k is the time when the network event occurs. x k y k z ω represents the spatial frequency component, and ω represents the time angular frequency.
[0014] As a preferred embodiment, the expression of the multi-objective optimization algorithm in the network optimization strategy generation module is: ; Where U(S) is the comprehensive evaluation value of network state, S is the network state vector, n≥5, and s i Let w be the original value of the i-th key performance indicator. i The weight coefficient of the i-th indicator, f i Let be the normalization function for the i-th index.
[0015] Compared with existing technologies, this invention achieves full-dimensional network data collection by deploying enhanced telemetry agents throughout the network. Combined with standardized coding, regional aggregation and global fusion processing, it forms a real-time updated global network status view, solving the problem of local perception in traditional operation and maintenance, and realizing full-domain, real-time and accurate perception of network operation status. Based on the causal field model, the global state of the network is dynamically analyzed, which breaks through the limitations of traditional correlation analysis. It can quantify the causal influence relationship of different spatiotemporal locations of the network, accurately identify abnormal network states, and reconstruct the network state context through holographic encoding and inverse Fourier transform to form a fault evolution view, thereby realizing the root cause location and evolution law inference of network anomalies. Based on the optimized status data after network configuration adjustment, the parameters of the causal field model are updated in real time, and the system policy library is optimized through the policy evolution mechanism, so that the operation and maintenance policies can be autonomously iterated and upgraded with the dynamic changes of the network environment, thereby improving the adaptability and coping ability of the operation and maintenance system to complex network environments. By generating optimal network optimization strategies through multi-objective optimization algorithms and implementing configuration adjustments based on distributed execution mechanisms, the entire process of network fault identification, tracing, and handling is automated, significantly improving network operation and maintenance efficiency, reducing manual intervention, lowering operation and maintenance costs, and effectively ensuring the stable and reliable operation of the network. Attached Figure Description
[0016] Figure 1 This is a logical flowchart of the intelligent network operation and maintenance method of the present invention, which has the capabilities of global self-sensing, causal inference, and autonomous evolution. Figure 2 This is a schematic diagram of the modules of the intelligent network operation and maintenance system of the present invention, which has the capabilities of global self-sensing, causal inference, and autonomous evolution. Figure 3 This is a schematic diagram of the enhanced telemetry data acquisition module of the intelligent network operation and maintenance system of the present invention, which has the capabilities of global self-sensing, causal inference, and autonomous evolution. Figure 4 This is a schematic diagram of the network state context reconstruction module of the intelligent network operation and maintenance system of the present invention, which has the capabilities of global self-awareness, causal inference, and autonomous evolution. Explanation of reference numerals in the attached figures: 1. Enhanced telemetry data acquisition module; 11. Basic metadata parsing unit; 12. Feature extraction and calculation unit; 13. Encoding and mapping processing unit; 14. Verification and encapsulation processing unit. 2. Network global status view generation module; 3. Network anomaly identification module; 4. Network State Context Reconstruction Module; 41. Holographic Encoded Data Retrieval Unit; 42. Historical Network Event Sequence Reconstruction Unit; 43. Fault Evolution View Generation Unit. 5. Network optimization strategy generation module; 6. Network configuration adjustment execution module; 7. System strategy optimization module. Detailed Implementation
[0017] The present invention will now be described in further detail with reference to the embodiments and accompanying drawings. However, this should not be construed as limiting the scope of the above-described subject matter of the present invention to the following embodiments; all technologies implemented based on the content of the present invention fall within the scope of the present invention.
[0018] refer to Figure 1 Example 1: This invention provides an intelligent network operation and maintenance method with global self-awareness, causal inference, and autonomous evolution capabilities, comprising the following steps: Step 1: Acquire Enhanced Telemetry Data Enhanced telemetry agents are deployed at key locations in the network. These agents capture various data in the network in real time, such as information related to data packets.
[0019] The operation follows the standardized encoding processing steps. In S101, the header of the captured network data packets is parsed to extract basic metadata, such as the source address, destination address, and protocol type of the data packets.
[0020] In S102, based on the extracted basic metadata, feature extraction calculation is performed through a specific algorithm to obtain the transmission delay, that is, the time taken for the data packet to travel from the source to the destination; the path hash, which is used to identify the path traversed by the data packet; and the state characterization code, which reflects the current state characteristics of the network.
[0021] In S103, based on the previously obtained transmission delay, path hash, and state representation code, the calculation results are converted into standardized data blocks through encoding mapping rules for subsequent unified processing.
[0022] In S104, the standardized data blocks are verified and encapsulated. By adding verification information, the integrity and accuracy of the data are ensured, thus obtaining complete enhanced telemetry data.
[0023] Step 2: Generate a global network state view The acquired enhanced telemetry data is aggregated by region, and the data within the same region is integrated and processed. Then, global fusion is performed to summarize the data from each region. After a series of processing steps, a real-time updated global network status view is obtained, which reflects the current status of the entire network.
[0024] Step 3: Identify abnormal network conditions Based on the generated global state view of the network, it is substituted into the causal field model, and the causal field model expression is as follows: ; in, For scalar fields in network topology space coordinates x and time coordinate t, The causal field value at the network topology spatial coordinates x′ and time coordinate t′; Let be the partial derivative of the causal field with respect to time, and α be the normalization parameter. For phase parameters, Let λ represent the causal influence strength of the states at different locations x′ and different times t′ in the network on the current location x and the current time t, and let λ be the coupling strength coefficient.
[0025] The expression for the causal influence strength of the states at different locations x′ and at different times t′ on the current location x and the current time t: ; Where, d phys (x, x′) represents the topological distance between nodes, the shortest path hop count or physical distance from node x to node x′ in the network, and d logic (x, x′) represents the logical topological distance, defined as the number of hops along the shortest path from position x to position x′ in the network; α1 is the nonlocal time delay influence weight coefficient, with a value range of (0,1]; β is the local instantaneous influence weight coefficient, with a value range of (0,1]; and Δ is the time delay parameter. The variance parameter of the physical distance term, For time decay constant, This is the variance parameter for the logical distance term.
[0026] Step 4: Reconstruct the network state context Once an abnormal network state is identified, the spatiotemporal coordinates corresponding to that abnormal state are determined, i.e., the specific location and time when the abnormality occurred.
[0027] Based on these spatiotemporal coordinates, the corresponding hologram encoded data is retrieved from the distributed spatiotemporal storage module; Holographic encoded data is generated in the following way: ; in, For spatiotemporal frequency domain holographic coding, Let x, y, z be the three-dimensional physical space coordinates (x, y, z) and time t be the network event occurring at time t, where x, y, z are the three-dimensional physical space coordinates, t is the time when the network event occurs, and k is the time when the network event occurs. x k y k z ω represents the spatial frequency component, and ω represents the time angular frequency.
[0028] The retrieved holographic encoded data is decoded using the inverse Fourier transform algorithm to restore the encoded data to a historical network event sequence, that is, a series of events that occurred in the network in chronological order.
[0029] By comparing and analyzing the obtained historical network event sequences with the identified network anomalies, the changes in network events before and after the anomalies occur can be identified, thus obtaining a complete fault evolution view that clearly shows how the fault occurs and develops.
[0030] Step 5: Generate network optimization strategy Taking into account the global network state view, abnormal network states, and network state context, and using this information as input, the multi-objective optimization algorithm is expressed as follows: ; Where U(S) is the comprehensive evaluation value of network state, S is the network state vector, n≥5, and s i Let w be the original value of the i-th key performance indicator. i The weight coefficient of the i-th indicator, f i Let be the normalization function for the i-th index.
[0031] This algorithm comprehensively evaluates and optimizes multiple key performance indicators, such as network bandwidth utilization and latency, to generate network optimization strategies that can improve network performance and resolve network anomalies.
[0032] Step Six: Implement Network Configuration Adjustments Based on the generated network optimization strategy, a distributed execution mechanism is used to implement network configuration adjustments on various relevant devices in the network, such as adjusting the routing policies of routers and the port configurations of switches.
[0033] After completing the configuration adjustment, obtain the optimized network status data, which reflects the new state of the network after the optimization configuration.
[0034] Step 7: Optimize the system policy library Based on the optimized network state data, the parameters of the causal field model are updated so that the model can more accurately reflect the changes in the network.
[0035] At the same time, the system policy library is optimized through a policy evolution mechanism, incorporating the experience and new optimization strategies from this network operation and maintenance process into the policy library, so that similar situations can be handled more quickly and effectively in the future, achieving the system's autonomous evolution.
[0036] refer to Figure 2 , Figure 3 , Figure 4 Example 2: An intelligent network operation and maintenance system with global self-awareness, causal inference, and autonomous evolution capabilities, comprising: The enhanced telemetry data acquisition module 1, in which the basic metadata parsing unit 11, feature extraction and calculation unit 12, encoding and mapping processing unit 13 and verification and encapsulation processing unit 14 work together to complete the entire process from network data capture to enhanced telemetry data generation, providing basic data for subsequent network operation and maintenance.
[0037] The network global status view generation module 2 receives enhanced telemetry data provided by the enhanced telemetry data acquisition module 1, performs regional aggregation and global fusion processing, and generates a real-time updated network global status view, providing comprehensive network status information for subsequent steps such as network anomaly identification.
[0038] The network anomaly state identification module 3, based on the network global state view provided by the network global state view generation module 2, uses a causal field model for dynamic analysis to identify network anomalies, providing key information for subsequent reconstruction of the network state context and generation of network optimization strategies.
[0039] The network state context reconstruction module 4, based on the spatiotemporal coordinates corresponding to the network anomaly state identified by the network anomaly state identification module 3, reconstructs the network state context at the time of the anomaly through the collaborative work of the holographic encoded data retrieval unit 41, the historical network event sequence reconstruction unit 42, and the fault evolution view generation unit 43, providing detailed background information for accurately analyzing the cause of the network anomaly.
[0040] The network optimization strategy generation module 5 integrates information provided by the network global state view generation module 2, the network anomaly state identification module 3, and the network state context reconstruction module 4. It generates network optimization strategies through a multi-objective optimization algorithm, providing specific solutions for improving network performance and resolving network anomaly issues.
[0041] The network configuration adjustment execution module 6 implements network configuration adjustments in the network through a distributed execution mechanism based on the network optimization strategy generated by the network optimization strategy generation module 5, obtains optimized network status data, implements the optimization measures in the actual network, and provides feedback on the optimized network status.
[0042] The system policy optimization module 7 adjusts the optimized network status data provided by the execution module 6 based on the network configuration, updates the causal field model parameters, and optimizes the system policy library through the policy evolution mechanism, thereby realizing the system's autonomous evolution and continuous optimization, and improving the overall network operation and maintenance capabilities and adaptability.
[0043] Through the above-mentioned intelligent network operation and maintenance methods and the collaborative work of each step and module in the system, it is possible to achieve global self-awareness, causal inference and autonomous evolution of the network, effectively improve the efficiency and quality of network operation and maintenance, and ensure the stable operation of the network.
[0044] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.
Claims
1. An intelligent network operation and maintenance method with global self-awareness, causal inference, and autonomous evolution capabilities, characterized in that: Includes the following steps: S1: Network data is captured in real time by enhanced telemetry agents deployed throughout the network, and the network data is standardized and encoded to obtain enhanced telemetry data; S2: Perform regional aggregation and global fusion processing on the enhanced telemetry data to obtain a real-time updated global network status view; S3: Based on the global state view of the network, dynamic analysis is performed using a causal field model to identify abnormal network states; S4: Based on the spatiotemporal coordinates corresponding to the network anomaly state, reconstruct the network state context at the time of the anomaly by retrieving historical data; S5: Generate a network optimization strategy using a multi-objective optimization algorithm based on the network global state view, the network abnormal state, and the network state context; S6: Based on the network optimization strategy, network configuration adjustments are implemented through a distributed execution mechanism to obtain optimized network status data; S7: Update the causal field model parameters based on the optimized network state data, and optimize the system policy library through the policy evolution mechanism.
2. The intelligent network operation and maintenance method with global self-awareness, causal inference, and autonomous evolution capabilities as described in claim 1, characterized in that, The standardization coding process in step S1 includes: S101: Parse the header of the data packets in the network data to obtain basic metadata; S102: Based on the basic metadata, perform feature extraction calculations to obtain transmission delay, path hash, and state representation code; S103: Based on the feature calculation results obtained in step S102, perform encoding mapping processing to obtain standardized data blocks; S104: Perform verification and encapsulation processing on the standardized data block to obtain complete enhanced telemetry data.
3. The intelligent network operation and maintenance method with global self-awareness, causal inference, and autonomous evolution capabilities as described in claim 1, characterized in that, The expression for the causal field model in step S3 is as follows: ； in, For scalar fields in network topology space coordinates x and time coordinate t, The causal field value at the network topology spatial coordinates x′ and time coordinate t′; Let be the partial derivative of the causal field with respect to time, and α be the normalization parameter. For phase parameters, Let λ represent the causal influence strength of the states at different locations x′ and different times t′ in the network on the current location x and the current time t, and let λ be the coupling strength coefficient.
4. The intelligent network operation and maintenance method with global self-awareness, causal inference, and autonomous evolution capabilities as described in claim 1, characterized in that, The network state context reconstruction in step S4 includes: S401: Retrieve the corresponding hologram encoded data from the distributed spatiotemporal storage module according to the spatiotemporal coordinates; S402: The hologram encoded data is decoded using the inverse Fourier transform algorithm to reconstruct the historical network event sequence; S403: Compare and analyze the historical network event sequence with the network abnormal state to obtain a complete fault evolution view; The hologram-encoded data in step S401 is generated in the following manner: ； in, For spatiotemporal frequency domain holographic coding, Let x, y, z be the three-dimensional physical space coordinates (x, y, z) and time t be the network event occurring at time t, where x, y, z are the three-dimensional physical space coordinates, t is the time when the network event occurs, and k is the time when the network event occurs. x k y k z ω represents the spatial frequency component, and ω represents the time angular frequency.
5. The intelligent network operation and maintenance method with global self-awareness, causal inference, and autonomous evolution capabilities as described in claim 1, characterized in that, The expression for the multi-objective optimization algorithm in step S5 is as follows: ； Where U(S) is the comprehensive evaluation value of network state, S is the network state vector, n≥5, and s i Let w be the original value of the i-th key performance indicator. i The weight coefficient of the i-th indicator, f i Let be the normalization function for the i-th index.
6. An intelligent network operation and maintenance system with global self-awareness, causal inference, and autonomous evolution capabilities, characterized in that: include: The enhanced telemetry data acquisition module (1) is used to capture network data in real time through enhanced telemetry agents deployed in various parts of the network, and to perform standardized encoding processing on the network data to obtain enhanced telemetry data; The network global status view generation module (2) is used to perform regional aggregation and global fusion processing on the enhanced telemetry data to obtain a real-time updated network global status view; The network anomaly identification module (3) is used to identify network anomalies by performing dynamic analysis based on the global network state view and a causal field model. The network state context reconstruction module (4) is used to reconstruct the network state context when the anomaly occurred by retrieving historical data based on the spatiotemporal coordinates corresponding to the network anomaly state. The network optimization strategy generation module (5) is used to generate a network optimization strategy based on the network global state view, the network abnormal state and the network state context, using a multi-objective optimization algorithm. The network configuration adjustment execution module (6) is used to implement network configuration adjustment through a distributed execution mechanism according to the network optimization strategy to obtain optimized network status data; The system strategy optimization module (7) is used to update the causal field model parameters based on the optimized network state data and optimize the system strategy library through the strategy evolution mechanism.
7. The intelligent network operation and maintenance system with global self-awareness, causal inference, and autonomous evolution capabilities as described in claim 6, characterized in that, The enhanced telemetry data acquisition module (1) includes: The basic metadata parsing unit (11) is used to parse the header of the data packets in the network data to obtain basic metadata; The feature extraction calculation unit (12) is used to perform feature extraction calculation based on the basic metadata to obtain transmission delay, path hash and state characterization code; The encoding mapping processing unit (13) is used to perform encoding mapping processing based on the feature calculation results obtained by the feature extraction calculation unit (12) to obtain a standardized data block; The verification and encapsulation processing unit (14) is used to perform verification and encapsulation processing on the standardized data block to obtain complete enhanced telemetry data.
8. The intelligent network operation and maintenance system with global self-awareness, causal inference, and autonomous evolution capabilities as described in claim 6, characterized in that, The expression for the causal field model in the network anomaly identification module (3) is: ； in, For scalar fields in network topology space coordinates x and time coordinate t, The causal field value at the network topology spatial coordinates x′ and time coordinate t′; Let be the partial derivative of the causal field with respect to time, and α be the normalization parameter. For phase parameters, Let λ represent the causal influence strength of the states at different locations x′ and different times t′ in the network on the current location x and the current time t, and let λ be the coupling strength coefficient.
9. The intelligent network operation and maintenance system with global self-awareness, causal inference, and autonomous evolution capabilities as described in claim 6, characterized in that, The network state context reconstruction module (4) includes: The hologram encoded data retrieval unit (41) is used to retrieve the corresponding hologram encoded data from the distributed spatiotemporal storage module according to the spatiotemporal coordinates; The historical network event sequence reconstruction unit (42) is used to decode the hologram encoded data using the inverse Fourier transform algorithm to reconstruct the historical network event sequence; The fault evolution view generation unit (43) is used to compare and analyze the historical network event sequence with the network abnormal state to obtain a complete fault evolution view; The hologram encoded data is generated in the following manner: ； in, For spatiotemporal frequency domain holographic coding, Let x, y, z be the three-dimensional physical space coordinates (x, y, z) and time t be the network event occurring at time t, where x, y, z are the three-dimensional physical space coordinates, t is the time when the network event occurs, and k is the time when the network event occurs. x k y k z ω represents the spatial frequency component, and ω represents the time angular frequency.
10. The intelligent network operation and maintenance system with global self-awareness, causal inference, and autonomous evolution capabilities according to claim 6, characterized in that, The expression for the multi-objective optimization algorithm in the network optimization strategy generation module (5) is as follows: ； Where U(S) is the comprehensive evaluation value of network state, S is the network state vector, n≥5, and s i Let w be the original value of the i-th key performance indicator. i The weight coefficient of the i-th indicator, f i Let be the normalization function for the i-th index.