Multi-network integration data processing method, device, equipment, medium and program product

Abstract

The invention provides a multi-network integration data processing method, device and equipment, a medium and a program product. The method comprises the following steps: acquiring data traffic information in a plurality of systems, and carrying out preprocessing and time-space alignment on the data traffic information to obtain a plurality of data sequences for the plurality of systems; acquiring traffic statistical feature vectors of the plurality of data sequences, and performing feature fusion on the traffic statistical feature vectors of the plurality of systems based on an attention weighting algorithm to obtain a fused feature vector and a network state recognition result; generating a first network bandwidth resource allocation strategy based on the fusion feature vector, the corresponding network state recognition result and the flow feature vectors of the multiple systems; and executing at least one of bandwidth allocation, routing adjustment and link switching on the network bandwidth resources in the plurality of systems based on the first network bandwidth resource allocation strategy. According to the invention, dynamic resource allocation in a cross-network fusion scene can be realized, and the network awareness, intelligent scheduling and remote management capabilities of the system are significantly improved.

Description

Technical Field

[0001] This disclosure relates to the field of computer technology, and in particular to a multi-network converged data processing method, apparatus, device, medium, and program product. Background Technology

[0002] With the rapid development of information technology, telecommunications networks, broadcast television networks, and the Internet are gradually converging at the access and service layers, forming a complex network environment of "multiple networks coexisting and multiple services running in parallel." While this convergence provides users with a wealth of multimedia communication and information services, it also brings problems such as increased network complexity, resource allocation conflicts, and greater difficulty in operation and maintenance management. Different network architectures use different protocol standards and data interfaces, resulting in low interconnection efficiency, scattered data collection points, inconsistent formats, and difficulties for maintenance personnel in timely understanding the overall operational status.

[0003] Most existing network management systems rely on static configuration and manual policy adjustments, lacking the ability to analyze multi-source dynamic traffic in real time. When sudden network congestion or link delays occur, traditional systems cannot perform cross-network collaborative scheduling, easily leading to bandwidth waste and service quality degradation. Summary of the Invention

[0004] In view of this, the purpose of this disclosure is to provide a multi-network converged data processing method, apparatus, equipment, medium and program product to solve or partially solve the above problems.

[0005] To achieve the above objectives, the first aspect of this disclosure provides a multi-network converged data processing method, comprising: Data traffic information from multiple systems is acquired, and the data traffic information is preprocessed and spatiotemporally aligned to obtain multiple data sequences for the multiple systems. Obtain the traffic statistics feature vectors of the multiple data sequences, and perform feature fusion on the traffic statistics feature vectors of the multiple systems based on the attention weighting algorithm to obtain the fused feature vector and network state identification result; Based on the fused feature vector and the corresponding network state identification result, and the traffic feature vector of the multiple systems, a first network bandwidth resource allocation strategy is generated. Based on the first network bandwidth resource allocation strategy, at least one of bandwidth allocation, routing adjustment, and link switching is performed on the network bandwidth resources of the multiple systems.

[0006] In some embodiments, the preprocessing and spatiotemporal alignment of the data traffic information includes: Perform at least one of the following on the data traffic information: invalid data filtering, format conversion, cleaning, and standardization; Data alignment is performed on the data traffic information of the multiple systems using time as an index.

[0007] In some embodiments, the feature fusion of the traffic statistics feature vectors of the multiple systems based on the attention weighting algorithm to obtain a fused feature vector includes: Get the context vector; Determine the dot product result between the context vector and the traffic statistics feature vector of each system; The weight of each system is determined based on the ratio of the dot product result of each system to the sum of the dot product results of the multiple systems. The fusion feature vector is obtained by weighting the system based on its corresponding weight and the traffic statistics feature vector.

[0008] In some embodiments, generating a first network bandwidth resource allocation strategy based on the fused feature vector and the corresponding network state identification result, and the traffic statistics feature vectors of the multiple systems includes: The fused feature vector and the corresponding network state identification result, as well as the traffic statistics feature vector of the multiple systems, are input into the first model, and the first network bandwidth resource allocation strategy is determined based on the output information of the first model. The update process of the first model includes: The state vector of the first model is constructed based on the fused feature vector, the corresponding network state identification result, and the traffic statistics feature vector of the multiple systems. Based on the state vector, a dynamic vector of the first model is constructed, and the dynamic vector represents the second network bandwidth resource allocation strategy for the multiple systems. The second network bandwidth resource allocation strategy corresponding to the state vector is executed, and the execution effect of the second network bandwidth resource allocation strategy is evaluated by the reward function of the first model, wherein the reward function is configured to aim at at least one of maximizing bandwidth utilization, minimizing latency, and minimizing packet loss rate. The parameters of the first model are updated based on the evaluation results.

[0009] In some embodiments, the method further includes: executing the second network bandwidth resource allocation strategy corresponding to the state vector based on preset constraints; The constraints include at least one of the following: In response to the network latency of the first system in the plurality of systems exceeding a first preset duration, the first connection being executed in the first system is migrated to another system other than the first system; In response to the network latency of the first system lasting less than a second preset duration and remaining stable for a third preset duration, the first link is migrated back to the first system, where the second preset duration is less than the first preset duration. Set the proportion of network bandwidth resources adjusted in each bandwidth allocation execution to be less than or equal to the first preset value; In response to the real-time load rate of the second system exceeding a second preset value among the multiple systems, the priority of the second system in executing the network bandwidth resource allocation strategy is increased.

[0010] In some embodiments, the step of performing at least one of bandwidth allocation, routing adjustment, and link switching on network bandwidth resources in the plurality of systems based on the first network bandwidth resource allocation strategy further includes: User types are determined based on the user's business behavior characteristics, priority, and network usage in the multiple systems. Determine the service priority type based on the service type of the data traffic information in the multiple systems; Based on the first network bandwidth resource allocation strategy and the user type and / or the service priority type, a third network bandwidth resource allocation strategy matching the user is determined. Based on the third network bandwidth resource allocation strategy, at least one of bandwidth allocation, routing adjustment, and link switching is performed on the network bandwidth resources of the multiple systems.

[0011] A second aspect of this disclosure provides a multi-network converged data processing apparatus, comprising: The system service module is configured to: acquire data traffic information from multiple systems, preprocess and align the data traffic information in time and space to obtain multiple data sequences for the multiple systems; The behavior analysis module is configured to: acquire traffic statistical feature vectors of the multiple data sequences, perform feature fusion on the traffic statistical feature vectors of the multiple systems based on an attention weighting algorithm, and obtain fused feature vectors and network state recognition results; The strategy scheduling module is configured to generate a first network bandwidth resource allocation strategy based on the fused feature vector and the corresponding network state identification result, and the traffic feature vector of the multiple systems. The policy execution module is configured to: perform at least one of bandwidth allocation, routing adjustment and link switching on the network bandwidth resources of the multiple systems based on the first network bandwidth resource allocation policy.

[0012] A third aspect of this disclosure provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the multi-network converged data processing method as described in the first aspect.

[0013] A fourth aspect of this disclosure provides a non-transitory computer-readable storage medium storing computer instructions for causing the computer to perform the multi-network converged data processing method described in the first aspect.

[0014] The fifth aspect of this disclosure provides a computer program product, including computer program instructions that, when executed on a computer, cause the computer to perform the multi-network converged data processing method as described in the first aspect.

[0015] As can be seen from the above, the multi-network convergence data processing method, apparatus, device, medium, and program product provided in this disclosure obtains multiple data sequences for multiple systems by acquiring data traffic information from multiple network systems and performing preprocessing and spatiotemporal alignment. It then obtains traffic statistical feature vectors for the multiple data sequences, performs feature fusion on the traffic statistical feature vectors of the multiple systems based on an attention weighting algorithm, and obtains a fused feature vector and network status identification results. Based on the fused feature vector, the corresponding network status identification results, and the traffic feature vectors of the multiple systems, a first network bandwidth resource allocation strategy is generated. Based on the first network bandwidth resource allocation strategy, at least one of bandwidth allocation, routing adjustment, and link switching is performed on the network bandwidth resources of the multiple systems, realizing dynamic resource allocation and self-learning optimization in cross-network convergence scenarios. This significantly improves the system's network awareness, intelligent scheduling, and remote management capabilities. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in this disclosure or related technologies, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 A schematic diagram of an exemplary system provided by an embodiment of this disclosure is shown.

[0018] Figure 2 A flowchart illustrating an exemplary method provided by an embodiment of this disclosure is shown.

[0019] Figure 3 A schematic diagram of an exemplary data acquisition method provided by an embodiment of this disclosure is shown.

[0020] Figure 4 A schematic diagram of an exemplary page provided by an embodiment of this disclosure is shown.

[0021] Figure 5 A schematic diagram of another exemplary page provided by an embodiment of this disclosure is shown.

[0022] Figure 6 A schematic diagram of an exemplary apparatus provided by an embodiment of the present disclosure is shown.

[0023] Figure 7 A schematic diagram of the hardware structure of an exemplary computer device provided in an embodiment of this disclosure is shown. Detailed Implementation

[0024] It is understood that before using the technical solutions of the various embodiments in this disclosure, users will be informed of the type, scope of use, and usage scenarios of the personal information involved in an appropriate manner, and user authorization will be obtained.

[0025] For example, upon receiving a user's active request, a prompt message is sent to the user to explicitly inform them that the requested operation will require the acquisition and use of the user's personal information. This allows the user to independently choose, based on the prompt message, whether to provide personal information to the software or hardware such as electronic devices, applications, servers, or storage media performing the operations of this disclosed technical solution.

[0026] As an optional but not limited implementation, in response to a user's active request, sending a prompt message to the user can be done via a pop-up window, where the prompt message can be presented in text format. Furthermore, the pop-up window can also include a selection control allowing the user to choose "agree" or "disagree" to provide personal information to the electronic device.

[0027] It is understood that the above notification and user authorization process are merely illustrative and do not constitute a limitation on the implementation of this disclosure. Other methods that comply with relevant laws and regulations may also be applied to the implementation of this disclosure.

[0028] It is understood that the data involved in this technical solution (including but not limited to the data itself, the acquisition or use of the data) shall comply with the requirements of relevant laws, regulations and related provisions.

[0029] To make the objectives, technical solutions, and advantages of this disclosure clearer, the principles and spirit of this disclosure will be described below with reference to several exemplary embodiments. It should be understood that these embodiments are provided merely to enable those skilled in the art to better understand and implement this disclosure, and are not intended to limit the scope of this disclosure in any way. Rather, these embodiments are provided to make this disclosure more thorough and complete, and to fully convey the scope of this disclosure to those skilled in the art.

[0030] In this article, it is important to understand that any number of elements in the accompanying figures is for illustrative purposes and not for limitation, and any naming is for distinction only and has no limiting meaning.

[0031] It should be noted that, unless otherwise defined, the technical or scientific terms used in the embodiments of this disclosure should have the ordinary meaning understood by one of ordinary skill in the art to which this disclosure pertains. The terms "first," "second," and similar words used in the embodiments of this disclosure do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly. The article "a" or "an" preceding an element does not exclude the existence of multiple such elements.

[0032] The principles and spirit of this disclosure will be explained in detail below with reference to several representative embodiments.

[0033] As described in the background section, most existing network management systems rely on static configuration and manual policy adjustments, lacking the ability to analyze multi-source dynamic traffic in real time. When sudden network congestion or link delays occur, traditional systems cannot perform cross-network collaborative scheduling, easily leading to bandwidth waste and service quality degradation.

[0034] In view of this, embodiments of this disclosure provide a multi-network converged data processing method, comprising: acquiring data traffic information from multiple systems; preprocessing and spatiotemporally aligning the data traffic information to obtain multiple data sequences for the multiple systems; acquiring traffic statistical features of the multiple data sequences; fusing the traffic feature vectors of the multiple systems based on an attention weighting algorithm to obtain a fused feature vector and a network state identification result; generating a first network bandwidth resource allocation strategy based on the fused feature vector, the corresponding network state identification result, and the traffic feature vectors of the multiple systems; and performing at least one of bandwidth allocation, routing adjustment, and link switching on the network bandwidth resources of the multiple systems based on the first network bandwidth resource allocation strategy to solve the above problems.

[0035] Meanwhile, the existing technology also has the following problems: users' business needs vary significantly, voice, video and data services have different sensitivities to bandwidth and latency, and traditional unified management mechanisms cannot flexibly allocate resources according to business priorities; in addition, the alarm and monitoring functions of existing systems are often limited to a single network, lacking a unified visual management and automated alarm mechanism for multi-network converged environments, resulting in long problem location and processing cycles and delayed responses.

[0036] This application combines service type and user priority with the first network bandwidth resource allocation strategy to generate a personalized bandwidth allocation strategy; at the same time, it provides a unified visualization of data in a multi-network converged environment to solve the above problems.

[0037] After introducing the basic principles of this disclosure, various non-limiting embodiments of this disclosure will be described in detail below.

[0038] Figure 1 A schematic diagram of an exemplary system 100 provided in an embodiment of this disclosure is shown.

[0039] like Figure 1 As shown, system 100 can be used to implement data processing functions based on multi-network convergence, and may include terminal devices 102A and 102B, server 106, and database server 108. The terminal devices 102A and 102B may include a medium (e.g., a network) providing a communication link with server 106 and database server 108. This network may include various connection types, such as wired, wireless communication links, or fiber optic cables, etc.

[0040] The terminal devices 102A and 102B can be equipped with various applications (APPs) or software, such as project management applications or software, collaborative office applications or software, image processing applications or software, video conferencing applications or software, reading applications or software, video applications or software, social applications or software, payment applications or software, web browsers, and instant messaging tools. In some embodiments, these applications or software can be used for multi-protocol interface communication based on a microservice architecture.

[0041] The terminal devices 102A and 102B here can be either hardware or software. When terminal devices 102A and 102B are hardware, they can be various electronic devices with displays, including but not limited to smartphones, tablets, e-book readers, MP3 players, laptops, and desktop computers (PCs). When terminal devices 102A and 102B are software, they can be installed in the electronic devices listed above. They can be implemented as multiple software programs or software modules (e.g., to provide distributed services) or as a single software program or software module. No specific limitations are set here.

[0042] Server 106 can be a server that provides various services, such as a backend server that supports various applications displayed on terminal devices 102A and 102B. Database server 108 can also be a database server that provides various services. It is understood that if server 106 can implement the relevant functions of database server 108, database server 108 may not need to be set up in system 100.

[0043] The server 106 and database server 108 here can be either hardware or software. When they are hardware, they can be implemented as a distributed server cluster consisting of multiple servers, or as a single server. When they are software, they can be implemented as multiple software programs or software modules (for example, to provide distributed services), or as a single software program or software module. No specific limitations are made here.

[0044] It should be noted that the information processing method provided in this embodiment can be executed by server 106. It should be understood that... Figure 1 The number of terminal devices, users, servers, and database servers shown is merely illustrative. Depending on implementation needs, there can be any number of terminal devices, users, servers, and database servers.

[0045] The following is combined with Figure 1The above application scenarios are used to describe the methods according to exemplary embodiments of this disclosure. It should be noted that the above application scenarios are shown only to facilitate understanding of the spirit and principles of this disclosure, and the embodiments of this disclosure are not limited in any way. Rather, the embodiments of this disclosure can be applied to any applicable scenario.

[0046] like Figure 2 As shown in the figure, this disclosure provides a multi-network converged data processing method, including: Step S101: Obtain data traffic information from multiple systems, preprocess and align the data traffic information in time and space to obtain multiple data sequences for the multiple systems.

[0047] The multiple systems may include telecommunications network systems, broadcast television network systems, and Internet network systems, and may also include other types of network systems; this embodiment does not limit this.

[0048] like Figure 3 As shown, multi-network converged data processing can deploy data acquisition nodes at the convergence layer of telecommunications network systems, broadcast television network systems, and Internet network systems. Each data acquisition node is linked to the core switch or service server and collects real-time data streams and obtains data traffic information through mirroring ports and packet capture mechanisms.

[0049] In this embodiment, the mirror port is configured in full-traffic mirroring mode to ensure that the data collected at any time can reflect the real business load.

[0050] In this embodiment, after receiving data streams from different network systems, the timestamps are first calibrated. In this embodiment, the data acquisition node employs a high-precision time synchronization module, whose internal clock maintains an event deviation of less than or equal to 1ms with the upstream Network Time Protocol (NTP) or Precision Time Protocol (PTP) service to ensure time consistency of data sampling across multiple network systems.

[0051] The data acquisition nodes are equipped with protocol identification and classification capabilities. The system utilizes Deep Packet Inspection (DPI) technology to identify data traffic information based on protocols such as Transmission Control Protocol (TCP), User Datagram Protocol (UDP), Real-time Transport Protocol (RTP), Hypertext Transfer Protocol (HTTP), Real-Time Streaming Protocol (RTSP), and Quick UDP Internet Connections (QUIC), and tags the data traffic information according to the protocol header fields.

[0052] The collected data is then sampled and cached.

[0053] The sampling period is set to Δt=1s, and the sampling fields include at least one of the following: timestamp ts, source address src, destination address dst, protocol type proto, data length len, and service identifier tag. Data from each sampling point is packaged and transmitted in JSON format. In some embodiments, the sampled data traffic information can be uploaded in batches and transmitted in segments.

[0054] In some embodiments, to prevent sampling buffer overflow, the sampled data traffic information can be cached in a tiered manner. The first-level cache resides in the memory of the acquisition node, while the second-level cache is stored on the local disk. The first-level cache is refreshed every 5 seconds. When the cache queue length Q > Q_max, the system automatically transfers the data to the second-level cache, where Q_max represents the maximum data length that the first-level cache can store.

[0055] In some embodiments, when a sudden congestion occurs on the link, data traffic information can be compressed. In this embodiment, a sliding window algorithm can be used to reduce the storage of redundant packets while retaining statistical information such as packet count and average latency, thereby reducing data transmission pressure without losing key features.

[0056] In this embodiment, real-time data streams are collected through port mirroring and packet capture mechanisms, and data packets of different protocol types are automatically identified, invalid traffic is filtered out, and continuous sampling is achieved at the second level, providing a high-quality data source for subsequent feature analysis.

[0057] Step S103: Obtain the traffic statistics feature vectors of the multiple data sequences, and perform feature fusion on the traffic statistics feature vectors of the multiple systems based on the attention weighting algorithm to obtain the fused feature vector and network state identification result.

[0058] In this embodiment, traffic statistics are extracted for the data sequence corresponding to each system. That is, the traffic statistics features of the telecommunications network system, the broadcast television network system and the Internet network system are extracted respectively to obtain the traffic statistics features corresponding to each system.

[0059] The traffic statistics features include at least one of the following: bandwidth utilization, average latency, packet loss rate, retransmission rate, and connection rate.

[0060] In this embodiment, after obtaining the traffic statistical features corresponding to each system, the statistical features extracted from each system are combined to construct the traffic statistical feature vector of the network system. Where i ∈ {telecommunications network, broadcasting network, Internet}. In this embodiment, the potential correlation degree of each traffic statistical feature can be calculated by a multi-layer sensing unit to obtain the traffic statistical feature vector of each system.

[0061] In this embodiment, after obtaining the traffic statistics feature vectors corresponding to each system, the traffic statistics feature vectors of the multiple systems are fused using an attention weighting algorithm to obtain a fused feature vector; then, network state identification is performed based on the fused feature vector to obtain the network state identification result corresponding to the fused feature vector. The network state identification result may include a network state label and a corresponding confidence value.

[0062] Step S105: Based on the fused feature vector and the corresponding network state identification result, and the traffic statistics feature vector of the multiple systems, a first network bandwidth resource allocation strategy is generated.

[0063] In this embodiment, based on the fused feature vector And the corresponding network status identification results, as well as traffic statistics feature vectors of multiple network systems. , and generate strategies.

[0064] Among them, the fused feature vector Used to determine the overall network trend, while the traffic statistics feature vectors of each network system It is used to detect local congestion or anomalies.

[0065] In this embodiment, based on the traffic statistics feature vector A state vector for the currently managed network is generated, which includes the bandwidth utilization, average latency, packet loss rate, number of users, and real-time load metrics for each network. By comparing the historical state sequence with the current state trend, it is determined whether there are persistent uneven bandwidth usage, congestion, or link anomalies, providing input for subsequent scheduling decisions.

[0066] Step S107: Based on the first network bandwidth resource allocation strategy, perform at least one of bandwidth allocation, route adjustment, and link switching on the network bandwidth resources of the multiple systems.

[0067] In this embodiment, the generated first network bandwidth resource allocation strategy is sent to the execution unit of the network control layer through a message queue. The control unit parses the scheduling instructions and implements at least one of the corresponding bandwidth allocation, route adjustment and link switching.

[0068] In this embodiment, data traffic information from multiple network systems is acquired, preprocessed, and spatiotemporally aligned to obtain multiple data sequences for each system. Traffic statistical feature vectors from these multiple data sequences are then obtained. An attention-weighted algorithm is used to fuse the traffic statistical feature vectors of the multiple systems, resulting in a fused feature vector and a network state identification result. Based on the fused feature vector, the corresponding network state identification result, and the traffic feature vectors from the multiple systems, a first network bandwidth resource allocation strategy is generated. Based on this strategy, at least one of bandwidth allocation, routing adjustment, and link switching is performed on the network bandwidth resources of the multiple systems. This achieves dynamic resource allocation and self-learning optimization in cross-network convergence scenarios, significantly improving the system's network awareness, intelligent scheduling, and remote management capabilities.

[0069] In some embodiments, the preprocessing and spatiotemporal alignment of the data traffic information in step S101 includes: Step S201 involves performing at least one of the following processes on the data traffic information: invalid data filtering, format conversion, cleaning, and standardization.

[0070] In this embodiment, invalid data in the data traffic information can be filtered.

[0071] The system can use preset filtering rules to filter out voice, video, and HTTP data streams, and filter out invalid data such as heartbeat packets, ARP broadcast packets, and idle traffic. The preset filtering rules may include: capturing all signaling (SIP) used to establish a call, streaming media (RTP) during the call, and web browsing traffic (UDP port 5060 or RTP) or (TCP port 80 or 443).

[0072] In this embodiment, the packet loss rate is calculated after each acquisition cycle: .

[0073] like If a preset value is set, such as 1%, a resampling mechanism is triggered to dynamically adjust the sampling interval in order to restore the continuity of traffic sampling.

[0074] All collected data is encrypted before being uploaded using a preset encryption algorithm to prevent data interception or tampering at the data link layer. The preset encryption algorithm may include AES-256, etc.

[0075] In this embodiment, a preset verification algorithm can be used to verify the integrity of the data packets to ensure the authenticity and consistency of the uploaded data. The preset verification algorithm may include an MD5 checksum.

[0076] In some embodiments, the data acquisition node can perform self-diagnosis. For example, if time drift exceeding 5ms or sampling loss exceeding 2% is detected twice consecutively, the system automatically issues a maintenance alarm and reports the status to the central management terminal.

[0077] In this embodiment, the data traffic information can be format converted, including format alignment of the data traffic information.

[0078] In this embodiment, the data formats of data traffic information from different network systems are standardized. Data from telecommunications network systems is output in Simple Network Management Protocol (SNMP) or Network Flow (NetFlow) format; data from broadcasting network systems is in Moving Picture Experts Group (MPEG) bitstream format; and data from internet network systems is in HTTP / TCP structure. In some embodiments, a custom parsing template can be used to uniformly convert various data types into a standard field structure, such as {ts, src, dst, proto, len, tag}, thereby ensuring consistency across multiple input sources.

[0079] In this embodiment, when the received data packet has missing fields or incorrect timestamps, the data can be corrected. The correction logic includes field padding, time reordering, and packet merging operations to eliminate time drift caused by different network sampling delays.

[0080] In this embodiment, data cleaning and anomaly repair can be performed on data traffic information.

[0081] In this embodiment, data traffic information can be cleaned to remove redundant fields and duplicate samples. Specifically, a hash fingerprint matching mechanism can be used to detect duplicate packets. When two data packets are completely identical in five key fields (source address, destination address, protocol, length, and timestamp), the duplicate data is deleted, and only one sample is retained.

[0082] In this embodiment, an interpolation algorithm can be used to repair missing data. Here, it is assumed that there are continuous sampling points t. If valid data exists at time 1 and t+1, but missing at time t, then linear interpolation is used for calculation: The repaired sequences are labeled as dummy samples and their weights are reduced by a preset value in subsequent analyses, for example, by 10%.

[0083] In this embodiment, when the deviation between the data flow information and the average value of the sliding window exceeds 3σ, the sample is marked as an outlier and smoothed using a local weighted regression method.

[0084] In this embodiment, data traffic information can be standardized.

[0085] The z-score standardization method can be used to map feature values ​​of different dimensions to the same statistical interval. The calculation formula is as follows: , where μ and σ are the mean and standard deviation within the sliding window, respectively.

[0086] Step S203: Data alignment of the data traffic information of the multiple systems is performed using time as an index.

[0087] In this embodiment, the standardized data traffic information is used to generate a time-aligned time series matrix { using time as the index}. ,in It is a time-series matrix that contains only data traffic information from telecommunications network systems. It is a time-series matrix that contains only data traffic information from the cable television network system. It is a time-series matrix that contains only data traffic information from Internet network systems.

[0088] The three time series matrices mentioned above share the same sampling reference on the time axis, but their data are not mixed. Traffic samples from different sources are aligned on the same time reference.

[0089] In some embodiments, the data refresh cycle is dynamically adjusted based on system resource load. The default refresh cycle is 60 seconds, which is automatically extended to 90 seconds when the data volume surges to ensure stability. The length of the aforementioned time window is consistent with the data refresh cycle and is adjusted synchronously with it.

[0090] The completed time series matrix is ​​stored in a temporary Redis cache. The cache space adopts a ring structure to ensure that the data is updated in chronological order.

[0091] In some embodiments, the step S103, which involves fusing the traffic statistics feature vectors of the multiple systems based on an attention-weighted algorithm to obtain a fused feature vector, includes: Step S301: Obtain the context vector.

[0092] In this embodiment, the context vector is obtained. .in, It is a trainable context vector used to evaluate the importance of each network feature to the global state.

[0093] Step S303: Determine the dot product result of the context vector and the traffic statistics feature vector of each system.

[0094] In this embodiment, a shared context vector is used. Each network system's traffic statistics feature vector is compared with the traffic statistics feature vector. Performing the dot product operation yields an importance scalar fraction. .

[0095] In some embodiments, the Softmax function can be used to normalize the importance scalar score.

[0096] Step S305: Determine the weight of each system based on the ratio of the dot product result of each system to the sum of the dot product results of the multiple systems.

[0097] In this embodiment, the formulas for the weights of each system are as follows: , in, Represents feature weights, For the extracted traffic statistics feature vector of the i-th network system, This is an iteration of the traffic statistics feature vectors for all network systems.

[0098] In this embodiment, feature weights The system dynamically adjusts based on network operating status. The formula calculates that this calculation process itself achieves dynamic adjustment based on network operating status. Specifically, due to the traffic statistics feature vector... The extracted traffic statistics features, such as bandwidth and latency, are already included. When the bandwidth utilization or latency of a network fluctuates drastically, its traffic statistics feature vector will show... The value will change significantly, causing its calculated value to change according to the formula. The dot product score increases accordingly, thus increasing its feature weights. It is also adaptively enhanced. For example, when the bandwidth utilization or latency fluctuation of a certain network exceeds 20%, the system automatically increases the weight of that network's features in the fusion vector to enhance the sensitivity of anomaly detection.

[0099] Step S307: Perform weighted calculation based on the weights of each system and the traffic statistics feature vector to obtain the fused feature vector.

[0100] In this embodiment, the final fused feature vector It is the traffic statistics feature vector of all network systems. The weighted average sum is used to automatically focus on the network features at the current time. Among these, the fused feature vectors... The formula is as follows:

[0101] In this embodiment, the traffic statistics feature vectors of the three network systems—telecommunications network system, broadcast television network system, and Internet network system—are fused into a single, high-dimensional global fused feature using an attention weighting algorithm. Fusion features It serves as the global state input for the generation of subsequent policy scheduling.

[0102] In some embodiments, the fusion feature Simultaneously, timestamps, data source identifiers, and fusion weights are recorded for subsequent backtracking and model retraining.

[0103] In some embodiments, after obtaining the fused feature vector, network state identification is performed based on the fused feature vector to obtain the network state identification result corresponding to the fused feature vector.

[0104] In some embodiments, a support vector machine (SVM) classifier is used to identify the network state of the fused feature vectors to obtain the network state identification result.

[0105] In some embodiments, the network state identification result includes at least one of normal, mild congestion, severe congestion, abnormal fluctuations, and link loss. Abnormal fluctuations refer to an unstable network state learned through SVM model training that does not conform to the normal congestion pattern and where latency or packet loss rate oscillates drastically within a short period.

[0106] In some embodiments, when the network status identification result shows abnormal fluctuation tags three times consecutively, it is determined that the network is in a state of persistent instability. At this time, emergency pre-regulation is performed, including link switching and load balancing based on preset conditions, including at least one of the following: in response to the network latency duration of the first system in the plurality of systems exceeding a first preset duration, the first link being executed in the first system is migrated to another system other than the first system; in response to the network latency duration of the first system being less than a second preset duration and remaining stable for a third preset duration, the first link is migrated back to the first system, where the second preset duration is less than the first preset duration; the proportion of network bandwidth resources adjusted in each bandwidth allocation is set to be less than or equal to a first preset value; in response to the real-time load rate of the second system in the plurality of systems exceeding a second preset value, the priority of the second system in executing the network bandwidth resource allocation strategy is increased.

[0107] In some embodiments, the model may perform adaptive updates and drift detection.

[0108] In this embodiment, the current distribution can be calculated periodically. Historical distribution KL divergence: , in, This refers to the probability distribution of traffic statistics features extracted within the current time window. When the divergence value exceeds 0.05, the system triggers model retraining. Model retraining uses a micro-batch incremental update mechanism, loading a data window of nearly one hour each time as training samples.

[0109] After training is complete, the new model parameters replace the old model, and the old model parameters are automatically put into a cold standby state to prevent analysis interruption due to update failure. The module records the time, reason, and validation results of each model update, forming a model version library to support subsequent version backtracking and validation of the system.

[0110] In some embodiments, the step S105 of generating a first network bandwidth resource allocation strategy based on the fused feature vector and the corresponding network state identification result and the traffic statistics feature vector of the multiple systems includes: inputting the fused feature vector and the corresponding network state identification result and the traffic statistics feature vector of the multiple systems into a first model, and determining the first network bandwidth resource allocation strategy based on the output information of the first model.

[0111] The update process of the first model includes: Step S401: Construct the state vector of the first model based on the fused feature vector, the corresponding network state identification result, and the traffic statistics feature vector of the multiple systems.

[0112] In this embodiment, the first model can be a state-action-feedback model. This first model can generate a dynamic resource allocation strategy, i.e., a first network bandwidth resource allocation strategy, based on reinforcement learning methods.

[0113] The state vector is composed of the fused feature vector. Each network system's own traffic statistics feature vector Composition, based on fused feature vectors Each network system's own traffic statistics feature vector The state vectors of the first model can be constructed.

[0114] Step S403: Based on the state vector, construct the dynamic vector of the first model, wherein the dynamic vector represents the second network bandwidth resource allocation strategy for the multiple systems.

[0115] In this embodiment, the dynamic vector of the first model is constructed based on the state vector.

[0116] The action vector represents the second network bandwidth resource allocation strategy in each network system. This second network bandwidth resource allocation strategy may include a bandwidth resource allocation ratio.

[0117] Step S405: Execute the second network bandwidth resource allocation strategy corresponding to the state vector, and evaluate the execution effect of the second network bandwidth resource allocation strategy through the reward function of the first model, wherein the reward function is configured to aim at at least one of maximizing bandwidth utilization, minimizing latency, and minimizing packet loss rate.

[0118] In this embodiment, a reward function is used. The effectiveness of each strategy execution is evaluated, with the goal of maximizing bandwidth utilization, minimizing latency, and minimizing packet loss rate.

[0119] Step S407: Update the parameters of the first model based on the evaluation results.

[0120] In some embodiments, the Proximal Policy Optimization (PPO) algorithm can be used for online learning, with an update cycle of 2 seconds, and the model parameters are updated based on feedback data after each policy execution.

[0121] When the system performance improvement is detected to be less than a preset threshold, the model automatically enters exploration mode and adjusts the strategy search range to avoid getting trapped in local optima.

[0122] In some embodiments, the method further includes: executing the second network bandwidth resource allocation strategy corresponding to the state vector based on preset constraints; wherein the constraints include at least one of the following: In response to a network latency in the first system exceeding a first preset duration, the first link currently in operation in the first system is migrated to another system besides the first system. For example, when a link latency in a network system is detected... If the handover continues for more than 120 milliseconds, the cross-network handover process is initiated. The handover process employs a buffer migration mechanism: first, the user sessions of the current connection are frozen, and the data is cached in a temporary buffer area. Then, the connection is rebuilt on the backup link and the service flow is restored, thereby ensuring that the service is not interrupted during the handover.

[0123] In response to the network latency of the first system lasting less than a second preset duration and remaining stable for a third preset duration, the first link is migrated back to the first system, where the second preset duration is less than the first preset duration. For example, if the latency is detected to drop below 100 milliseconds and remain stable for more than 10 seconds, the system performs a switchback operation, migrating traffic back to the main link.

[0124] The proportion of network bandwidth resources adjusted in each bandwidth allocation is set to be less than or equal to a first preset value. For example, based on a hysteresis control strategy, the bandwidth adjustment proportion is kept below 20% during each switch, thereby preventing resource fluctuations caused by frequent switches and achieving smooth transition and stable scheduling.

[0125] In response to the real-time load rate of the second system exceeding a second preset value among the multiple systems, the priority of the second system in executing the network bandwidth resource allocation strategy is increased. For example, by monitoring the real-time load rate of each network, when the utilization rate of any network exceeds 85%, the system prioritizes scheduling idle links to share some traffic, achieving the effect of network-wide load balancing.

[0126] After generating the first network bandwidth resource allocation strategy, the generated first network bandwidth resource allocation strategy is sent to the execution unit of the network control layer through a message queue. The control unit parses the scheduling instructions and implements at least one of the corresponding bandwidth allocation, route adjustment and link switching.

[0127] In some embodiments, after each round of policy execution, key indicators such as changes in bandwidth utilization, latency fluctuations, and packet loss rate are recorded, and the performance improvement rate is calculated.

[0128] In some embodiments, if the strategy does not produce significant performance improvement after being executed for a preset number of consecutive times, the first model is automatically retrained and its parameters are refitted. The preset number of times may include three, four, five, etc., and this embodiment does not limit this.

[0129] All execution records are stored in the policy database for subsequent analysis and model tracing. Through this mechanism, self-learning and self-optimization control based on real-time feedback is achieved, forming a dynamic scheduling closed loop within the system.

[0130] In some embodiments, step S107, which involves performing at least one of bandwidth allocation, routing adjustment, and link switching on network bandwidth resources in the plurality of systems based on the first network bandwidth resource allocation strategy, further includes: Step S501: Determine the user type based on the user's business behavior characteristics, priority, and network usage in the multiple systems.

[0131] In this embodiment, user types are determined based on user business behavior characteristics, priority, and network usage. For example, a behavioral profile can be created for each terminal user, obtaining statistical characteristics including voice call duration, video viewing ratio, download speed, and peak service volume. Through multidimensional clustering analysis of user business behavior characteristics, users are divided into three types: ordinary, priority, and critical. Ordinary users prioritize stable transmission, priority users are more sensitive to latency, while critical users typically correspond to enterprise or government services and have the highest requirements for bandwidth and stability.

[0132] In this embodiment, basic resource shares can be allocated first based on the classification results of user types, and state data can be shared with the first model to achieve collaborative optimization between the user layer and the network layer.

[0133] Step S503: Determine the service priority type based on the service type of the data traffic information in the multiple systems.

[0134] In this embodiment, the service priority type is determined based on the service type of the data traffic information in each network system, and each service priority type corresponds to a different level of service level model.

[0135] In this embodiment, a multi-level service level model is adopted to achieve differentiated service control. The service levels include three tiers: guaranteed service (B1), standard service (B2), and resilient service (B3).

[0136] B1 Guarantee Service (B1) corresponds to the high priority type and is used for high priority services, such as video conferencing and voice calls, to ensure bandwidth priority and low latency.

[0137] Standard service (B2), corresponding to the normal priority type, is used for ordinary data services, such as web page access and file download.

[0138] Elastic services (B3) correspond to low-priority types and are low-priority services, such as system updates or background transmissions.

[0139] In this embodiment, the service level can be automatically determined based on the user's service category and real-time load, and a corresponding service policy table can be generated. The policy table includes the service rate limit, burst traffic threshold, scheduling priority, and policy number. All policies are stored in template form, and the system can load and switch them in milliseconds to adapt to dynamic network environments.

[0140] Step S505: Based on the first network bandwidth resource allocation strategy and the user type and / or the service priority type, determine a third network bandwidth resource allocation strategy that matches the user.

[0141] In this embodiment, a second network bandwidth resource allocation strategy matching the user is determined based on the first network bandwidth resource allocation strategy and the user type and / or the service priority type. Specifically, based on the first network bandwidth resource allocation strategy (a macro-level bandwidth allocation strategy), a micro-level, personalized QoS guarantee is executed through a token bucket mechanism according to the user type and / or service priority type.

[0142] For example, the first network bandwidth resource allocation strategy can be "total bandwidth of the telecommunications network is 3Gbps". In this embodiment, under the constraint of the first network bandwidth resource allocation strategy "total bandwidth of the telecommunications network is 3Gbps", a strategy that conforms to the user type and / or service priority type can be generated according to the user type and / or service priority type, that is, the third network bandwidth resource allocation strategy.

[0143] Step S507: Based on the third network bandwidth resource allocation strategy, perform at least one of bandwidth allocation, route adjustment, and link switching on the network bandwidth resources of the multiple systems.

[0144] In this embodiment, the generated third network bandwidth resource allocation strategy is sent to the execution unit of the network control layer through a message queue. The control unit parses the scheduling instructions and implements at least one of the corresponding bandwidth allocation, route adjustment and link switching.

[0145] In some real-time scenarios, the method also includes dynamic adjustment and adaptive optimization.

[0146] In this embodiment, the actual bandwidth usage of each user is continuously monitored. When a user's bandwidth utilization exceeds a preset threshold for multiple consecutive sampling periods, their service level is automatically upgraded or policy parameters are adjusted. For critical business users, when latency or packet loss rate rises to a warning level, a priority protection signal is immediately issued to allocate more resources. When idle or low-activity business is detected, the system will automatically reduce the corresponding priority to achieve resource reclamation and reallocation. In some embodiments, group scheduling can also be performed, and policy parameters can be uniformly adjusted for similar user groups to achieve batch management and resource fairness.

[0147] In some embodiments, all adjustment records are written to the policy log for subsequent performance analysis and auditing.

[0148] In some embodiments, the method further includes performing visual monitoring, including: In this embodiment, the visual monitoring module integrates a multi-layered security mechanism to verify the identity of users logging into the management system and to implement hierarchical access control.

[0149] This can be achieved using a token-based authentication mechanism, where each user generates a unique access token upon login. The system determines the system functions and data scope that a user can access based on their role, thereby ensuring the legitimacy and controllability of user policies and preventing bandwidth abuse caused by unauthorized access.

[0150] like Figure 4 As shown, this is the system login interface. Users need to complete identity verification through a token authentication mechanism before entering the main system interface, thus implementing role-based access control.

[0151] In some embodiments, visual monitoring includes: Step S601: Real-time monitoring and indicator collection.

[0152] In this embodiment, a front-end and back-end separation design is implemented based on a B / S architecture. The front-end uses the ECharts graphics engine and the WebSocket communication protocol to achieve millisecond-level data refresh.

[0153] like Figure 5The image shows the main interface of the system in this embodiment of the disclosure. The interface displays real-time traffic curves, latency distribution, and an alarm list. Maintenance personnel can view the system's operating status and implement policy interventions on this interface. The module monitors bandwidth utilization, average latency, packet loss rate, network connection count, user activity, and policy execution status in the three-network environment in real time. The system defaults to a five-second refresh cycle, which is dynamically adjusted based on data pressure. To ensure smooth display, an incremental update mechanism is used to refresh only the changed data, thereby significantly reducing server load. All monitoring data is temporarily stored in a cache queue to ensure stable display on the front end even under network fluctuations.

[0154] Step S603, alarm mechanism and automatic response.

[0155] In this embodiment, a multi-level alarm mechanism can be set, including three categories: mild, severe, and emergency. When the packet loss rate exceeds 1% or the average latency is higher than 120 milliseconds, the system generates a mild alarm; when the link is continuously abnormal or the latency of a critical node exceeds 150 milliseconds, a severe alarm is generated; if a service interruption or the master control node goes offline, an emergency alarm is triggered immediately.

[0156] In this embodiment, alarm information is pushed to the operation and maintenance management terminal via a message queue. The alarm information includes the event type, scope of impact, timestamp, and handling suggestions. This embodiment can aggregate alarm information; for example, it can automatically merge similar events and output them in chronological order to prevent duplicate notifications from causing information congestion. Operation and maintenance personnel can view, confirm, or disable alarms through the terminal interface and can directly issue adjustment commands to complete rapid recovery.

[0157] Step S605, Historical Log Management and Intelligent Analysis.

[0158] In this embodiment, historical records can be managed, and system operation data and alarm records for the past seven days can be saved.

[0159] In this embodiment, multi-dimensional queries and trend analysis can also be performed on the data. Operation and maintenance personnel can filter and statistically analyze the data by time interval, network type, or user group.

[0160] In some embodiments, a built-in intelligent analysis algorithm can automatically calculate the frequency and recovery time of abnormal events and generate a performance evaluation report. This report is used for model retraining and system parameter optimization. The system also has self-learning capabilities, enabling it to automatically adjust alarm thresholds based on historical false alarms and missed alarms, thereby improving alarm accuracy and response efficiency.

[0161] Step S607, Management terminal and remote control.

[0162] In this embodiment, the visual monitoring module maintains real-time communication with the management terminal, enabling remote operation and policy intervention by maintenance personnel. When the system detects a critical alarm event, it can automatically enter emergency mode, suspending non-critical tasks and prioritizing core business links. The terminal interface uses a hierarchical permission display, allowing users with different roles to view information at different levels. The module supports both mobile and desktop access, allowing maintenance personnel to view system status, handle alarms, or adjust policies at any time via mobile devices, achieving truly remote intelligent operation and maintenance.

[0163] Testing and verification have shown that the system described in this embodiment operates stably in a typical triple-play environment, improving bandwidth utilization by approximately 20%, reducing average latency by 25%, and achieving an alarm response time of less than 3 seconds. It has also operated continuously for 1000 hours without failure. This system integrates data acquisition, intelligent analysis, strategy scheduling, and visual control, demonstrating high reliability, scalability, and engineering application value.

[0164] It should be noted that the method of this disclosure embodiment can be executed by a single device, such as a computer or server. The method of this embodiment can also be applied to a distributed scenario, where multiple devices cooperate to complete the task. In such a distributed scenario, one of these devices may execute only one or more steps of the method of this disclosure embodiment, and the multiple devices will interact with each other to complete the method described.

[0165] It should be noted that the above description describes some embodiments of this disclosure. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in a different order than that shown in the above embodiments and still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0166] Based on the same inventive concept, corresponding to any of the above embodiments, this disclosure also provides a multi-network converged data processing device.

[0167] refer to Figure 6 The device includes: System service module 11 is configured to: acquire data traffic information from multiple systems, preprocess and align the data traffic information in time and space to obtain multiple data sequences for the multiple systems; Behavior analysis module 13 is configured to: acquire traffic statistical feature vectors of the multiple data sequences, perform feature fusion on the traffic statistical feature vectors of the multiple systems based on attention weighting algorithm, and obtain fused feature vectors and network state recognition results; The strategy scheduling module 15 is configured to generate a first network bandwidth resource allocation strategy based on the fused feature vector and the corresponding network state identification result, and the traffic feature vector of the multiple systems. The policy execution module 17 is configured to: perform at least one of bandwidth allocation, routing adjustment and link switching on the network bandwidth resources of the plurality of systems based on the first network bandwidth resource allocation policy.

[0168] In some embodiments, the system service module 11 is further configured to: Perform at least one of the following on the data traffic information: invalid data filtering, format conversion, cleaning, and standardization; Data alignment is performed on the data traffic information of the multiple systems using time as an index.

[0169] In some embodiments, the behavior analysis module 13 is further configured to: Get the context vector; Determine the dot product result between the context vector and the traffic statistics feature vector of each system; The weight of each system is determined based on the ratio of the dot product result of each system to the sum of the dot product results of the multiple systems. The fusion feature vector is obtained by weighting the system based on its corresponding weight and the traffic statistics feature vector.

[0170] In some embodiments, the policy scheduling module 15 is further configured to: The fused feature vector and the corresponding network state identification result, as well as the traffic statistics feature vector of the multiple systems, are input into the first model, and the first network bandwidth resource allocation strategy is determined based on the output information of the first model. The update process of the first model includes: The state vector of the first model is constructed based on the fused feature vector, the corresponding network state identification result, and the traffic statistics feature vector of the multiple systems. Based on the state vector, a dynamic vector of the first model is constructed, and the dynamic vector represents the second network bandwidth resource allocation strategy for the multiple systems. The second network bandwidth resource allocation strategy corresponding to the state vector is executed, and the execution effect of the second network bandwidth resource allocation strategy is evaluated by the reward function of the first model, wherein the reward function is configured to aim at at least one of maximizing bandwidth utilization, minimizing latency, and minimizing packet loss rate. The parameters of the first model are updated based on the evaluation results.

[0171] In some embodiments, the apparatus is further configured to: execute the second network bandwidth resource allocation strategy corresponding to the state vector based on preset constraints; The constraints include at least one of the following: In response to the network latency of the first system in the plurality of systems exceeding a first preset duration, the first connection being executed in the first system is migrated to another system other than the first system; In response to the network latency of the first system lasting less than a second preset duration and remaining stable for a third preset duration, the first link is migrated back to the first system, where the second preset duration is less than the first preset duration. Set the proportion of network bandwidth resources adjusted in each bandwidth allocation execution to be less than or equal to the first preset value; In response to the real-time load rate of the second system exceeding a second preset value among the multiple systems, the priority of the second system in executing the network bandwidth resource allocation strategy is increased.

[0172] In some embodiments, the policy execution module 17 is further configured to: User types are determined based on the user's business behavior characteristics, priority, and network usage in the multiple systems. Determine the service priority type based on the service type of the data traffic information in the multiple systems; Based on the first network bandwidth resource allocation strategy and the user type and / or the service priority type, a third network bandwidth resource allocation strategy matching the user is determined. Based on the third network bandwidth resource allocation strategy, at least one of bandwidth allocation, routing adjustment, and link switching is performed on the network bandwidth resources of the multiple systems.

[0173] For ease of description, the above apparatus is described in terms of its functions, divided into various modules. Of course, in implementing this disclosure, the functions of each module can be implemented in one or more software and / or hardware.

[0174] The apparatus described above is used to implement the corresponding multi-network convergence data processing method in any of the foregoing embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0175] Based on the same inventive concept, corresponding to the methods of any of the above embodiments, this disclosure also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the multi-network converged data processing method described in any of the above embodiments.

[0176] Figure 7 This embodiment illustrates a more specific hardware structure of an electronic device. The device may include a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, memory 1020, input / output interface 1030, and communication interface 1040 are interconnected internally via the bus 1050.

[0177] The processor 1010 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.

[0178] The memory 1020 can be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 1020 can store the operating system and other applications. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program code is stored in the memory 1020 and is called and executed by the processor 1010.

[0179] The input / output interface 1030 is used to connect input / output modules to realize information input and output. Input / output modules can be configured as components within the device (not shown in the figure) or externally connected to the device to provide corresponding functions. Input devices may include keyboards, mice, touchscreens, microphones, various sensors, etc., while output devices may include displays, speakers, vibrators, indicator lights, etc.

[0180] The communication interface 1040 is used to connect a communication module (not shown in the figure) to enable communication between this device and other devices. The communication module can communicate via wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).

[0181] Bus 1050 includes a pathway for transmitting information between various components of the device, such as processor 1010, memory 1020, input / output interface 1030, and communication interface 1040.

[0182] It should be noted that although the above-described device only shows the processor 1010, memory 1020, input / output interface 1030, communication interface 1040, and bus 1050, in specific implementations, the device may also include other components necessary for normal operation. Furthermore, those skilled in the art will understand that the above-described device may only include the components necessary for implementing the embodiments of this specification, and not necessarily all the components shown in the figures.

[0183] The electronic devices described above are used to implement the corresponding multi-network convergence data processing methods in any of the foregoing embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0184] Based on the same inventive concept, corresponding to the methods of any of the above embodiments, this disclosure also provides a non-transitory computer-readable storage medium storing computer instructions for causing the computer to execute the multi-network converged data processing method as described in any of the above embodiments.

[0185] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transfer medium that can be used to store information accessible by a computing device.

[0186] The computer instructions stored in the storage medium of the above embodiments are used to cause the computer to execute the multi-network converged data processing method as described in any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0187] Based on the same inventive concept, corresponding to the methods of any of the above embodiments, this disclosure also provides a computer program product, which includes a computer program. In some embodiments, the computer program is executable by one or more processors to cause the processors to perform the multi-network converged data processing method. Corresponding to the execution entity for each step in each embodiment of the method, the processor executing the corresponding step may belong to the corresponding execution entity.

[0188] The computer program product of the above embodiments is used to cause the processor to execute the multi-network converged data processing method as described in any of the above embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0189] Those skilled in the art will recognize that embodiments of this disclosure can be implemented as a system, method, or computer program product. Therefore, this disclosure can be implemented as entirely hardware, entirely software (including firmware, resident software, microcode, etc.), or a combination of hardware and software, generally referred to herein as a "circuit," "module," or "system." Furthermore, in some embodiments, this disclosure can also be implemented as a computer program product contained in one or more computer-readable media, which includes computer-readable program code.

[0190] Any combination of one or more computer-readable media may be used. A computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. A computer-readable storage medium can be, for example,, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (not exhaustive) of a computer-readable storage medium may include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in connection with an instruction execution system, apparatus, or device.

[0191] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, capable of sending, propagating, or transmitting programs for use by or in connection with an instruction execution system, apparatus, or device.

[0192] Program code contained on a computer-readable medium may be transmitted using any suitable medium, including but not limited to wireless, wire, optical fiber, RF, etc., or any suitable combination thereof.

[0193] Computer program code for performing the operations of this disclosure can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0194] It should be understood that each block of a flowchart and / or block diagram, as well as combinations of blocks in a flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device to produce a machine that, when executed by a computer or other programmable data processing device, creates means for implementing the functions / operations specified in the blocks of the flowchart and / or block diagram.

[0195] These computer program instructions may also be stored in a computer-readable medium that enables a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable medium produce a product comprising an instruction apparatus that implements the functions / operations specified in the boxes of a flowchart and / or block diagram.

[0196] Computer program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, such that the instructions that execute on the computer or other programmable apparatus can provide a process for implementing the functions / operations specified in the boxes of a flowchart and / or block diagram.

[0197] Furthermore, although the operations of the methods of this disclosure are described in a specific order in the accompanying drawings, this does not require or imply that these operations must be performed in that specific order, or that all of the operations shown must be performed to achieve the desired result. Rather, the steps depicted in the flowcharts may be executed in a different order. Additionally or alternatively, certain steps may be omitted, multiple steps may be combined into one step, and / or one step may be broken down into multiple steps.

[0198] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. Each block in a flowchart or block diagram may represent a module, segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0199] It should be noted that although several modules or units for the device used to perform actions have been mentioned in the detailed description above, this division is not mandatory. In fact, according to the embodiments of this application, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0200] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of this disclosure (including the claims) is limited to these examples; within the framework of this disclosure, the technical features of the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations of different aspects of the embodiments of this disclosure as described above, which are not provided in detail for the sake of brevity.

[0201] Additionally, to simplify the description and discussion, and to avoid obscuring the embodiments of this disclosure, the provided drawings may or may not show well-known power / ground connections to integrated circuit (IC) chips and other components. Furthermore, the apparatus may be shown in block diagram form to avoid obscuring the embodiments of this disclosure, and this also takes into account the fact that the details of implementation of these block diagram apparatuses are highly dependent on the platform on which the embodiments of this disclosure will be implemented (i.e., these details should be fully understood by those skilled in the art). While specific details (e.g., circuits) have been set forth to describe exemplary embodiments of this disclosure, it will be apparent to those skilled in the art that the embodiments of this disclosure can be implemented without these specific details or with variations thereof. Therefore, these descriptions should be considered illustrative rather than restrictive.

[0202] Although this disclosure has been described in conjunction with specific embodiments thereof, many substitutions, modifications, and variations of these embodiments will be apparent to those skilled in the art from the foregoing description. For example, other memory architectures (e.g., dynamic RAM (DRAM)) may be used with the embodiments discussed.

[0203] This disclosure is intended to cover all such substitutions, modifications, and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A multi-network converged data processing method, comprising: Data traffic information from multiple systems is acquired, and the data traffic information is preprocessed and spatiotemporally aligned to obtain multiple data sequences for the multiple systems. Obtain the traffic statistics feature vectors of the multiple data sequences, and perform feature fusion on the traffic statistics feature vectors of the multiple systems based on the attention weighting algorithm to obtain the fused feature vector and network state identification result; Based on the fused feature vector and the corresponding network state identification result, and the traffic feature vector of the multiple systems, a first network bandwidth resource allocation strategy is generated. Based on the first network bandwidth resource allocation strategy, at least one of bandwidth allocation, routing adjustment, and link switching is performed on the network bandwidth resources of the multiple systems.

2. The method of claim 1, wherein, The preprocessing and spatiotemporal alignment of the data traffic information includes: Perform at least one of the following on the data traffic information: invalid data filtering, format conversion, cleaning, and standardization; Data alignment is performed on the data traffic information of the multiple systems using time as an index.

3. The method of claim 1, wherein, The attention-weighted algorithm is used to fuse the traffic statistics feature vectors of the multiple systems to obtain a fused feature vector, including: Get the context vector; Determine the dot product result between the context vector and the traffic statistics feature vector of each system; The weight of each system is determined based on the ratio of the dot product result of each system to the sum of the dot product results of the multiple systems. The fusion feature vector is obtained by weighting the system based on its corresponding weight and the traffic statistics feature vector.

4. The method of claim 1, wherein, The step of generating a first network bandwidth resource allocation strategy based on the fused feature vector, the corresponding network state identification result, and the traffic statistics feature vectors of the multiple systems includes: The fused feature vector and the corresponding network state identification result, as well as the traffic statistics feature vector of the multiple systems, are input into the first model, and the first network bandwidth resource allocation strategy is determined based on the output information of the first model. The update process of the first model includes: The state vector of the first model is constructed based on the fused feature vector, the corresponding network state identification result, and the traffic statistics feature vector of the multiple systems. Based on the state vector, a dynamic vector of the first model is constructed, and the dynamic vector represents the second network bandwidth resource allocation strategy for the multiple systems. The second network bandwidth resource allocation strategy corresponding to the state vector is executed, and the execution effect of the second network bandwidth resource allocation strategy is evaluated by the reward function of the first model, wherein the reward function is configured to aim at at least one of maximizing bandwidth utilization, minimizing latency, and minimizing packet loss rate. The parameters of the first model are updated based on the evaluation results.

5. The method of claim 4, further comprising: The second network bandwidth resource allocation strategy corresponding to the state vector is executed based on preset constraints. The constraints include at least one of the following: In response to the network latency of the first system in the plurality of systems exceeding a first preset duration, the first connection being executed in the first system is migrated to another system other than the first system; In response to the network latency of the first system lasting less than a second preset duration and remaining stable for a third preset duration, the first link is migrated back to the first system, where the second preset duration is less than the first preset duration. Set the proportion of network bandwidth resources adjusted in each bandwidth allocation execution to be less than or equal to the first preset value; In response to the real-time load rate of the second system exceeding a second preset value among the multiple systems, the priority of the second system in executing the network bandwidth resource allocation strategy is increased.

6. The method of claim 4, wherein, The step of performing at least one of bandwidth allocation, routing adjustment, and link switching on network bandwidth resources in the multiple systems based on the first network bandwidth resource allocation strategy further includes: User types are determined based on the user's business behavior characteristics, priority, and network usage in the multiple systems. Determine the service priority type based on the service type of the data traffic information in the multiple systems; Based on the first network bandwidth resource allocation strategy and the user type and / or the service priority type, a third network bandwidth resource allocation strategy matching the user is determined. Based on the third network bandwidth resource allocation strategy, at least one of bandwidth allocation, routing adjustment, and link switching is performed on the network bandwidth resources of the multiple systems.

7. A multi-network converged data processing device, comprising: The system service module is configured to: acquire data traffic information from multiple systems, preprocess and align the data traffic information in time and space to obtain multiple data sequences for the multiple systems; The behavior analysis module is configured to: acquire traffic statistical feature vectors of the multiple data sequences, perform feature fusion on the traffic statistical feature vectors of the multiple systems based on an attention weighting algorithm, and obtain fused feature vectors and network state recognition results; The strategy scheduling module is configured to generate a first network bandwidth resource allocation strategy based on the fused feature vector and the corresponding network state identification result, and the traffic feature vector of the multiple systems. The policy execution module is configured to: perform at least one of bandwidth allocation, routing adjustment and link switching on the network bandwidth resources of the multiple systems based on the first network bandwidth resource allocation policy.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the multi-network converged data processing method as described in any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium storing computer instructions for causing the computer to perform the multi-network converged data processing method according to any one of claims 1 to 6.

10. A computer program product, characterised in that, It includes computer program instructions that, when executed on a computer, cause the computer to perform the multi-network converged data processing method as described in any one of claims 1 to 6.

Images