Application acceleration method, apparatus, device, and storage medium

Abstract

The application discloses an application acceleration method, device, equipment and storage medium, obtains to-be-identified traffic data, identifies the to-be-identified traffic data, identifies the application corresponding to the to-be-identified traffic data, obtains a quality difference analysis index corresponding to an application type to which a current application belongs, and performs quality difference analysis on traffic of the current application according to the quality difference analysis index; for to-be-accelerated traffic which needs to be accelerated according to a quality difference analysis result, the route of the to-be-accelerated traffic is colored, and the to-be-accelerated traffic is accelerated by using an application acceleration router. According to the embodiment of the application, the application can be accurately identified, the application can be accelerated in time, and the user experience is improved.

Description

Technical Field

[0001] This invention relates to the field of data communication, and more particularly to an application acceleration method, apparatus, device, and storage medium. Background Technology

[0002] In the context of the gigabit era, the demands for latency and bandwidth have given rise to many new applications, including: Cloud VR, smart homes, live streaming social networking, cloud desktops, gaming, smart cities, enterprise cloud migration, online education, telemedicine, smart manufacturing, and more. As network scale increases, it's crucial to better identify users' internet quality and deliver appropriate services to those who need them, creating differentiated service capabilities. To improve users' internet quality, application identification is necessary, followed by acceleration for each application. Currently, application deployment and maintenance typically rely on static policy routing and manual hop-by-hop deployment. Traditional methods result in complex, labor-intensive, and slow-moving network-wide implementations, making it difficult to respond promptly to large-scale and multi-service dynamic deployments, thus hindering timely application acceleration. Summary of the Invention

[0003] The purpose of this invention is to provide an application acceleration method, apparatus, device, and storage medium that can accurately identify applications and accelerate them in a timely manner to improve user experience.

[0004] To achieve the above objectives, embodiments of the present invention provide an application acceleration method, comprising:

[0005] Acquire traffic data to be identified and identify the application corresponding to the traffic data to be identified;

[0006] Obtain the quality difference analysis index corresponding to the application type of the current application, and perform quality difference analysis on the traffic of the current application based on the quality difference analysis index;

[0007] For traffic that needs acceleration according to the poor quality analysis results, the routes of the traffic to be accelerated are colored, and the application acceleration router is used to accelerate the traffic to be accelerated.

[0008] As an improvement to the above solution, the identification of the traffic data to be identified includes:

[0009] Extract feature data from the traffic data to be identified;

[0010] The feature data is input into a pre-trained media stream recognition model to output the application recognition result of the traffic data to be identified.

[0011] As an improvement to the above solution, the identification of the traffic data to be identified includes:

[0012] Extract the access requests of the traffic data to be identified;

[0013] The IP address and domain name corresponding to the access request are resolved using DNS packets;

[0014] Based on the IP address and the domain name, the application identification result of the traffic data to be identified is obtained.

[0015] As an improvement to the above solution, the identification of the traffic data to be identified includes:

[0016] Extract the feature parameters from the traffic data to be identified;

[0017] When the feature parameter matches the standard parameter of any target application pre-stored in the database, the target application is used as the application identification result of the traffic data to be identified.

[0018] As an improvement to the above scheme, the application types include game applications, video applications, educational applications, and other applications besides the above three types. Different types of applications correspond to different quality difference analysis indicators.

[0019] As an improvement to the above scheme, when the direction of the traffic data to be identified is from the user terminal to the server terminal, the step of accelerating the traffic to be accelerated using the application acceleration router includes:

[0020] The broadband remote access server diverts the traffic to be accelerated sent by the user to the independent network.

[0021] The standalone network redirects the next-hop address in the virtual private network and iterates the route to the specified SRv6 tunnel;

[0022] The application acceleration router matches the specified SRv6 tunnel based on the differential service code point in the packet, and forwards the traffic to be accelerated to the server through the specified SRv6 tunnel within the virtual private network.

[0023] As an improvement to the above scheme, when the direction of the traffic data to be identified is from the server to the user, the step of accelerating the traffic to be accelerated using the application acceleration router includes:

[0024] The traffic to be accelerated sent by the server returns to the application acceleration router;

[0025] The application acceleration router determines the next-hop broadband remote access server based on the routes published in the virtual private network and iteratively enters the specified SRv6 tunnel.

[0026] The broadband remote access server diverts the traffic to be accelerated to the standalone network and forwards it to the user through a designated SRv6 tunnel.

[0027] To achieve the above objectives, embodiments of the present invention also provide an application acceleration device, comprising:

[0028] An application identification module is used to acquire traffic data to be identified and to identify the application corresponding to the traffic data to be identified.

[0029] The quality difference analysis module is used to obtain quality difference analysis indicators corresponding to the application type to which the current application belongs, and to perform quality difference analysis on the traffic of the current application based on the quality difference analysis indicators.

[0030] The acceleration module is used to color the routes of traffic to be accelerated for traffic that needs acceleration according to the quality analysis results, and to accelerate the traffic to be accelerated using the application acceleration router.

[0031] To achieve the above objectives, embodiments of the present invention also provide an application acceleration device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement the application acceleration method as described in any of the above embodiments.

[0032] To achieve the above objectives, embodiments of the present invention also provide a computer-readable storage medium, the computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform the application acceleration method as described in any of the above embodiments.

[0033] Compared with existing technologies, the application acceleration method, apparatus, device, and storage medium disclosed in this invention have the following beneficial effects:

[0034] 1. By acquiring traffic data to be identified and identifying the traffic data to be identified, the application corresponding to the traffic data to be identified is identified. This application proposes three methods for identifying applications: fingerprint recognition with encrypted applications, DNS association recognition, and recognition by combination of data flow features and behaviors. Through heuristic rules, time series analysis, probabilistic graphical models, etc., the parameters are continuously trained to more accurately identify applications.

[0035] 2. For different types of applications, there are corresponding quality improvement analysis indicators, which allow users to analyze their application usage more comprehensively without being aware of it, provide a full range of quality improvement conclusions, locate different applications and give differentiated optimization effects, and obtain accurate positioning before accelerating media streams.

[0036] 3. For traffic that needs acceleration according to the poor quality analysis results, add it into the tunnel by coloring it with routing policies. Color the destination route with a specific color, then configure the next destination with redirection, and iterate the route into the specified SRv6 tunnel. The traffic to be accelerated is forwarded through the specified SRv6 tunnel, which can improve the user experience in the network. Attached Figure Description

[0037] Figure 1 This is a flowchart of an application acceleration method provided in an embodiment of the present invention;

[0038] Figure 2 This is a schematic diagram of a TLS fingerprint identification example provided in an embodiment of the present invention;

[0039] Figure 3 This is the traffic identification method for the Skype application provided in the embodiments of the present invention;

[0040] Figure 4 This is a schematic diagram illustrating the acceleration of traffic provided in an embodiment of the present invention;

[0041] Figure 5 This is a structural block diagram of an application acceleration device provided in an embodiment of the present invention;

[0042] Figure 6 This is a structural block diagram of an application acceleration device provided in an embodiment of the present invention. Detailed Implementation

[0043] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0044] See Figure 1 , Figure 1 This is a flowchart of an application acceleration method provided in an embodiment of the present invention, the application acceleration method comprising:

[0045] S1. Obtain traffic data to be identified, and identify the traffic data to be identified in order to identify the application corresponding to the traffic data to be identified;

[0046] S2. Obtain the quality difference analysis index corresponding to the application type of the current application, and perform quality difference analysis on the traffic of the current application based on the quality difference analysis index;

[0047] S3. For traffic that needs to be accelerated according to the quality analysis results, color the route of the traffic to be accelerated, and use the application acceleration router to accelerate the traffic to be accelerated.

[0048] Specifically, in step S1, traditional application identification technology mainly relies on the identification of IP / TCP / UDP packet headers, identifying applications based on fields such as IP address, port, and DSCP (Differentiated Services Code Point). However, with various VPNs and deceptive applications exploiting vulnerabilities in various protocol designs to forge traffic, obtaining plaintext features such as SNI (Server Name Indication) that traditional traffic identification technologies depend on is no longer feasible, causing traditional application identification technologies to fail to accurately identify applications. Therefore, this embodiment of the invention provides a new application identification method, mainly employing a multi-dimensional fusion encryption combination identification and parsing approach. It identifies applications by matching features in the packet payload, roughly divided into the following three modes: Mode 1, identification combined with encrypted application fingerprints; Mode 2, identification combined with DNS association; and Mode 3, identification combined with data flow features and behaviors.

[0049] Specifically, for Mode 1, the identification of the traffic data to be identified includes: extracting feature data from the traffic data to be identified; and inputting the feature data into a pre-trained media stream identification model to output the application identification result of the traffic data to be identified.

[0050] For example, see Figure 2 , Figure 2 This is a schematic diagram illustrating an example of TLS (Transport Layer Security) fingerprinting technology identification provided in this embodiment of the invention. By employing TLS fingerprinting technology, encrypted traffic lacking obvious characteristics, such as HTTP and map software, can be identified. TLS fingerprinting is a technology used to identify and verify TLS communication. It determines the characteristics of TLS communication by examining information such as the cipher suites, protocol versions, and encryption algorithms used during the TLS handshake process, thereby obtaining the characteristic data. This characteristic data includes interval time, characteristic packet length, the number of observed ACK (Acknowledgement) packets, the number of retransmissions, and round-trip time.

[0051] In this embodiment of the invention, a model is pre-built based on the feature data of the traffic, and then machine learning and model training are performed. Heuristic rules (a combination of simplified virtual machine and simplified behavior judgment engine, used for updating feature value recognition technology and solving the defects of single feature code comparison), time series analysis (time series refers to a set of numbers arranged in chronological order, and time series analysis uses this set of numbers to process it with mathematical statistics methods to predict the development of future events), probabilistic graphical models (probabilistic graphical models are a branch of machine learning that use probability distributions to describe the real world and make valuable predictions about it) are used to continuously train parameters, thereby obtaining a trained media stream recognition model, and accurately identifying media streams based on this media stream recognition model.

[0052] Specifically, for Mode 2, the identification of the traffic data to be identified includes: extracting the access request of the traffic data to be identified; resolving the IP address and domain name corresponding to the access request using DNS packets; and obtaining the application identification result of the traffic data to be identified based on the IP address and the domain name.

[0053] For example, most domain names are publicly available information. Knowing the domain name allows us to understand the protocol and purpose of the traffic data it sends. Therefore, if we find the domain name corresponding to the source IP address, we can identify which application sent the traffic data through the domain name. For instance, after a browser parses a URL and generates an HTTP data structure, it needs to look up the IP address corresponding to the domain name before sending it to the server. This is because the operating system must provide the IP address of the communication object when sending a message. DNS is specifically used to store the mapping between domain names and IP addresses, and it's the server responsible for resolving domain names. An IP address can have multiple domain names; when accessing a website, including the host field allows the server to know which website is being accessed.

[0054] Suppose there are two websites, A and B. Website A has the domain name http: / / aaa.com, and website B has the domain name http: / / bbb.com. Both domain names resolve to IP addresses with the domain name xxx.xxx.x.xxx. When configuring a web server, create a first website with the host header "webA", IP address xxx.xxx.x.xxx, and domain name http: / / aaa.com. Create a second website with the host header "webB", IP address xxx.xxx.x.xxx, and domain name http: / / bbb.com. When a user accesses http: / / aaa.com, an access request is sent. First, the DNS server resolves the IP address corresponding to the access request to xxx.xxx.x.xxx. Then, it sends an HTTP request to this server. The server determines that it is website webA based on the domain name being accessed, thus completing application identification.

[0055] Specifically, for mode three, the identification of the traffic data to be identified includes: extracting feature parameters from the traffic data to be identified; when the feature parameters match the standard parameters of any target application pre-stored in the database, the target application is taken as the application identification result of the traffic data to be identified.

[0056] For example, based on multi-packet association and multi-flow analysis, application identification is performed according to characteristic parameters, including: packet length, transmission direction, arrival time, packet transmission interval, etc. These characteristic parameters differ significantly between different applications, thus allowing us to determine which application initiated the attack. In this process, it is necessary to collect standard parameters for these applications beforehand to build a database, which can then be used for real-time application identification. See also... Figure 3 For example, Skype (an instant messaging software) can be identified based on port and traffic behavior. For ports commonly used by Skype, such as UDP ports 12340, 12350, 80, and 433, the traffic of the Skype application can be identified by combining feature association recognition and behavior analysis methods.

[0057] In this embodiment of the invention, the application identification methods described above, employing the three modes described above, significantly improve accuracy compared to traditional application identification. Furthermore, it effectively identifies applications with encrypted protocols that cannot be identified using feature keywords. Simultaneously, the background application knowledge base is continuously updated, allowing for timely comparison and faster, more effective identification after application identification. In addition, this invention optimizes traditional identification methods by identifying more features in the media stream, matching them with user behavior, and comparing them with background applications, thereby achieving more accurate application identification.

[0058] Specifically, in step S2, traditional quality degradation analysis methods typically identify and rectify issues based on characteristics such as weak ONU light, RADIUS disconnections, excessively high splitting ratios, overloaded PON ports, and overloaded links. However, these methods have a low degree of consistency with actual user perception. Many network metrics lag behind user experience; users may have already clearly perceived a deterioration in internet quality while the network and equipment appear normal. Furthermore, traditional quality degradation analysis methods do not differentiate between service types, and identifying services based on threshold exceedances is inherently biased. This makes it difficult to assess the quality of diverse internet services using traditional evaluation metrics. Fault localization only begins after a large number of user complaints, resulting in a reactive approach that consumes significant manpower and negatively impacts the user's internet experience.

[0059] To address the issues of lag, high manpower consumption, and poor internet experience associated with traditional quality defect analysis methods, this invention employs KQI (Key Quality Indicators) modeling and a quality defect indicator system to conduct quality defect analysis on applications. A quality defect indicator system strategy is proposed for the business layer, TCP layer, and protocol layer, as follows:

[0060] 1. At the business layer, focus on four dimensions: business awareness modeling, HTTP, video, and games. Specifically: ① Business awareness modeling focuses on lag and latency; ② HTTP focuses on HTTP GET method strategies and server response latency; ③ Video focuses on effective download speed and response latency; ④ Games focus on uplink and downlink data transmission latency.

[0061] 2. At the TCP layer, focus on four dimensions: user-side metrics, server-side metrics, DNS metrics, and traffic metrics. Specifically: ① The user-side dimension focuses on user-side TCP latency, downlink latency, and packet loss rate; ② The server-side dimension focuses on server-side TCP latency, uplink and downlink performance, and packet loss rate; ③ The DNS dimension focuses on DNS success rate and response latency; ④ The traffic dimension focuses on the flow of various types of user traffic.

[0062] 3. At the protocol layer, focus on two dimensions: PPPoE (Point-to-Point Protocol over Ethernet) and AAA (Authentication, Authorization, Accounting) authentication. Specifically: ① The PPPoE dimension focuses on authentication results and heartbeat latency; ② The AAA dimension focuses on AAA authentication records, online / offline records, and abnormal disconnections. AAA authentication records, online / offline records, and abnormal disconnections can be applied to quality control analysis in scenarios such as OTT video, web access, video conferencing, gaming, and downloads.

[0063] Specifically, the application types include game applications, video applications, educational applications, and other applications besides these three types. Different application types correspond to different quality deviation analysis indicators. For example, after implementing a quality deviation indicator strategy system, different detailed quality deviation mechanisms are formulated based on the application to evaluate and analyze the four major categories of applications: games, videos, education, and other applications. The analysis indicators are as follows:

[0064] 1. Quality analysis metrics for game applications include uplink and downlink data transmission counts and server latency for the user end, and downlink packet loss rate and battle heartbeat latency (heartbeat latency generated during character battles) for the resource end. For example, for the user end, if the uplink and downlink data transmission counts are greater than 12 times within 5 minutes, and the server latency is greater than 100ms per upload, the quality is poor; for the resource end, if the downlink packet loss rate is greater than 1% and the battle heartbeat latency is greater than 50ms per upload, the quality is poor.

[0065] 2. For users, video application quality analysis metrics include viewing time, buffering rate, and latency; for resources, they include downlink packet loss rate and server response latency. For example, for users, viewing time greater than 30 minutes, buffering rate greater than 3%, and latency greater than 50ms per upload are considered quality issues; for resources, downlink packet loss rate greater than 1% and server response latency greater than 400ms per upload are considered quality issues.

[0066] For video media streaming applications, this invention also proposes a "reservoir principle" quality difference analysis method. For video portal applications, this patent compares the video buffer to a reservoir, with downloading as the inlet pipe and playback as the outlet pipe. Initially, the reservoir is empty. When a user clicks play, downloading begins, and water enters through the inlet pipe. When the water level reaches a certain threshold, playback begins, and water flows out through the outlet pipe. If the inflow rate is consistently slower than the outflow rate (i.e., video bitrate), at some point, the reservoir will empty, the outlet pipe will close, and video playback will pause. Water continues to enter through the inlet pipe, and when the water level reaches a certain threshold, the outlet pipe will open, ending the pause and resuming playback. For video KPQ judgment, in this embodiment, the recovery threshold, initialization threshold, and buffering threshold are pre-set; then, the download traffic is calculated, and the remaining buffer amount is analyzed. The condition for no buffering is: viewing time < (download time - minimum buffer); the condition for no buffering is: viewing time = download time - minimum buffer.

[0067] 3. Quality degradation analysis indicators for educational applications include downlink data transmission latency and downlink packet loss rate for the user end, and downlink packet loss rate and latency for the resource end. For example, for the user end, a downlink data transmission latency greater than 50ms and a downlink packet loss rate greater than 0.5% within 5 minutes constitutes a quality degradation issue; for the resource end server, a downlink packet loss greater than 1% and a latency greater than 100ms constitute a quality degradation issue.

[0068] 4. Other application quality analysis metrics include downlink data transmission latency for the user end and downlink packet loss rate and HTTP GET success rate for the resource end. For example, a poor upload quality occurs when the user end traffic is greater than 100M and the downlink data transmission latency is greater than 100ms; a poor upload quality occurs when the resource end server packet loss is greater than 1% and the HTTP GET success rate is less than 95%.

[0069] In this invention, corresponding quality improvement analysis indicators are provided for different types of applications, allowing users to analyze their application usage more comprehensively without being aware of it, providing a full range of quality improvement conclusions, identifying different applications and providing differentiated optimization effects, and achieving precise positioning before accelerating media streams.

[0070] Specifically, in step S3, traditional application acceleration mainly targets the access side, accelerating applications by increasing their priority. In this embodiment of the invention, acceleration is performed from the interconnection side. For traffic that needs acceleration based on the quality analysis result (e.g., if the video application was detected to have uploaded a low-quality video once, then the traffic of the video application at this time is the traffic to be accelerated), after adding an intelligent engine, the routing policy is colored and added into the tunnel, coloring the destination route with a specific color; then the next destination is configured using redirection, the route iterates into the tunnel, and then the DSCP value in the packet is matched, and mapped to the SRV6 Policy color according to the configuration before entering the tunnel.

[0071] For example, see Figure 4 , Figure 4 This is a schematic diagram illustrating traffic acceleration provided in this embodiment of the invention. CGN (Carrier-grade NAT) is mainly deployed in carrier networks to complete network address translation. BRAS (Broadband Remote Access Server) is a new type of access gateway for broadband network applications. SA is an abbreviation for Standalone, meaning independent networking, and is a type of 5G networking. In SA standalone networking mode, accessing 5G base stations and the 5G core network can better leverage the advantages of 5G, such as ultra-low latency. CR (Core Router) is the core router. UCL (Uniform Content Locator) is a standardized vector metadata that describes the multi-dimensional and comprehensive characteristics of information resources, developed based on the Uniform Resource Locator (URL). It indexes various content data through processes such as Chinese news information collection, processing, transmission, and aggregation, enabling support for content information aggregation, dissemination, management, and personalized services.

[0072] SRv6 (Segment Routing IPv6) is simply SR (Segment Routing) + IPv6, a next-generation IP bearer protocol. It utilizes existing IPv6 forwarding technology and achieves network programmability through flexible IPv6 extension headers. Within traditional MPLS TE VPN scenarios, some services require forwarding on low-latency paths, while others require forwarding on high-bandwidth paths, leading to a lack of fine-grained service segmentation. SRv6 Policy tunneling technology, based on SRv6, can solve this problem. SRv6 networks primarily employ two routing technologies: SRv6 BE and SRv6 Policy. SRv6 BE uses IGP convergence to find the shortest path, but services cannot be forwarded along a specified path. SRv6 Policy, on the other hand, allows for path planning between any nodes in the network. Compared to SRv6 BE, using SRv6 Policy not only meets the differentiated requirements of user networks in terms of latency, bandwidth, jitter, and reliability, but also improves network bandwidth utilization and saves investment through fine-grained control of latency and bandwidth. An SRv6 Policy path is represented as a list of segments, also known as a Segment ID list. Each SID is an explicit path from source to destination. The SID list instructs devices in the network to follow the specified path, rather than the shortest path calculated by the IGP. If a data packet is imported into an SRv6 Policy, the node that generated the SRv6 Policy adds the SID list to the packet, and the remaining devices in the network execute the path specified in the SRv6 Policy SID list.

[0073] An SRv6 Policy consists of the following triples: HeadEnd, representing the node from which the SRv6 Policy is generated (no additional configuration is required); Endpoint, representing the destination address of the SRv6 Policy; and Color, an extended community attribute. Colors can identify paths with different latency or bandwidth, thus distinguishing different SRv6 Policies from the same headend to the endpoint. SRv6 Policies require color and endpoint configuration at the headend. The headend matches the corresponding SRv6 Policy using the color attribute and next-hop information carried in the route, enabling service packet forwarding.

[0074] SRv6 Policy supports importing routes into SRv6 Policy using colors. Routes need to be colored before importing. Route coloring refers to adding color-extended community attributes to routes through routing policies. Routes carrying color attributes can then be iterated into SRv6 Policy based on the color attribute and the next-hop address.

[0075] Specifically, when the direction of the traffic data to be identified is from the user terminal to the server terminal, the acceleration of the traffic to be accelerated using the application acceleration router includes: the broadband remote access server diverting the traffic to be accelerated sent by the user terminal to the standalone network; the standalone network redirecting the next-hop address in the virtual private network and routing iteratively to the designated SRv6 tunnel; the application acceleration router matching the designated SRv6 tunnel according to the differential service code point in the packet, and forwarding the traffic to be accelerated to the server terminal through the designated SRv6 tunnel within the virtual private network.

[0076] For example, ① the Broadband Remote Access Server (BRAS) UCL redirects traffic to the Standalone Network (SA). ② The SA identifies accelerated traffic, redirects the next-hop address (IP) in the Virtual Private Network (VPN), and iteratively routes it to the designated SRv6 tunnel. For non-accelerated traffic, the SA forwards it to the CR by default. ③ The application acceleration router terminates the SRv6 tunnel. The application acceleration router matches the designated SRv6 tunnel based on the Differential Service Code Point (DSC) in the packet and forwards the traffic to the accelerated access point (server) within the VPN through the designated SRv6 tunnel. The application acceleration router and the accelerated access point are connected to the VPN via their respective network ports.

[0077] It's worth noting that VPN refers to the technology of establishing a private network over a public network, also known as a virtual network. This is mainly because the connection between any two nodes in a VPN network does not have the end-to-end physical link required by traditional private networks. Instead, it's a logical network built on a network platform provided by a public network service provider, such as the Internet or ATM, with user data transmitted over this logical link. SRv6 Policy can be based on DSCP for traffic redirection. Specifically, the redirection method involves matching the tunnel binding policy based on the packet's destination address, associating it with an SRv6 Policy group, finding its associated Color attribute based on the packet's DSCP value, and then matching the Color attribute to a specific SRv6 Policy within the SRv6 Policy group. In this way, packets carrying a specified DSCP can be forwarded through the specified SRv6 Policy.

[0078] Specifically, when the direction of the traffic data to be identified is from the server to the user, the acceleration of the traffic to be accelerated using the application acceleration router includes: the traffic to be accelerated sent by the server returns to the application acceleration router; the application acceleration router determines the next-hop broadband remote access server according to the route published in the virtual private network, and iteratively enters the designated SRv6 tunnel; the broadband remote access server diverts the traffic to be accelerated to the standalone network and forwards it to the user through the designated SRv6 tunnel.

[0079] For example, ① the accelerated traffic returns to the application acceleration router. ② the application acceleration router determines the next-hop BRAS based on the route advertised by the BRAS in the VPN and iteratively enters the SRv6 tunnel. The BRAS UCL diverts the traffic to be accelerated to the SA and forwards it to the user through the designated SRv6 tunnel.

[0080] In this embodiment of the invention, differentiated scheduling service capabilities are provided. Terminals accessing the bearer network with different SLAs (Service-Level Agreements) select the corresponding tunnel SRv6 Policy for bearer, forming an end-to-end differentiated bearer scheme based on application flows. The SRv6 Policy enables low-latency, high-value channels. By introducing protocols such as BGP-LS and TWAMP, a real-time network digital map is constructed, featuring path orchestration and monitoring capabilities. It supports multi-factor routing calculations based on latency, bandwidth, and other factors, realizing the service-oriented nature of network capabilities. Rapid service provisioning capabilities are provided. Specific service traffic is routed with low latency to the optimal edge computing node for path orchestration and optimization, enabling rapid service provisioning. Intelligent network operation and maintenance capabilities are provided. Based on application-based flow detection technology, real-time visibility of the entire service SLA is provided, proactively identifying poor-quality services, improving high-precision service quality perception and management capabilities, and enhancing the intelligent operation and maintenance capabilities of the IP network.

[0081] Compared with existing technologies, the application acceleration method disclosed in this invention has the following beneficial effects:

[0082] 1. By acquiring traffic data to be identified and identifying the traffic data to be identified, the application corresponding to the traffic data to be identified is identified. This application proposes three methods for identifying applications: fingerprint recognition with encrypted applications, DNS association recognition, and recognition by combination of data flow features and behaviors. Through heuristic rules, time series analysis, probabilistic graphical models, etc., the parameters are continuously trained to more accurately identify applications.

[0083] 2. For different types of applications, there are corresponding quality improvement analysis indicators, which allow users to analyze their application usage more comprehensively without being aware of it, provide a full range of quality improvement conclusions, locate different applications and give differentiated optimization effects, and obtain accurate positioning before accelerating media streams.

[0084] 3. For traffic that needs acceleration according to the poor quality analysis results, add it into the tunnel by coloring it with routing policies. Color the destination route with a specific color, then configure the next destination with redirection, and iterate the route into the specified SRv6 tunnel. The traffic to be accelerated is forwarded through the specified SRv6 tunnel, which can improve the user experience in the network.

[0085] See Figure 5 , Figure 5 This is a structural block diagram of an application acceleration device 100 provided in an embodiment of the present invention. The application acceleration device 100 includes:

[0086] Application identification module 11 is used to acquire traffic data to be identified and to identify the application corresponding to the traffic data to be identified.

[0087] The quality difference analysis module 12 is used to obtain quality difference analysis indicators corresponding to the application type to which the current application belongs, and to perform quality difference analysis on the traffic of the current application based on the quality difference analysis indicators.

[0088] The acceleration module 13 is used to color the routes of traffic to be accelerated for traffic that needs to be accelerated according to the quality analysis results, and to accelerate the traffic to be accelerated using the application acceleration router.

[0089] Specifically, the application identification module 11 is used to: extract feature data from the traffic data to be identified; input the feature data into a pre-trained media stream identification model to output the application identification result of the traffic data to be identified.

[0090] Specifically, the application identification module 11 is used to: extract the access request of the traffic data to be identified; resolve the IP address and domain name corresponding to the access request using DNS packets; and obtain the application identification result of the traffic data to be identified based on the IP address and the domain name.

[0091] Specifically, the application identification module 11 is used to: extract feature parameters from the traffic data to be identified; when the feature parameters match the standard parameters of any target application pre-stored in the database, the target application is used as the application identification result of the traffic data to be identified.

[0092] Specifically, the application types include game applications, video applications, educational applications, and other applications besides the above three types. Different types of applications correspond to different quality difference analysis indicators.

[0093] Specifically, the acceleration module 13 includes a broadband remote access server and an application acceleration router located at the user end and the server end.

[0094] Specifically, when the direction of the traffic data to be identified is from the user terminal to the server terminal, the broadband remote access server diverts the traffic to be accelerated sent by the user terminal to the standalone network; the standalone network redirects the next-hop address in the virtual private network and routes iteratively to the designated SRv6 tunnel; the application acceleration router matches the designated SRv6 tunnel according to the differential service code point in the packet, and forwards the traffic to be accelerated to the server terminal through the designated SRv6 tunnel within the virtual private network.

[0095] Specifically, when the direction of the traffic data to be identified is from the server to the user, the traffic to be accelerated sent by the server returns to the application acceleration router; the application acceleration router determines the next-hop broadband remote access server according to the route published in the virtual private network, and iteratively enters the designated SRv6 tunnel; the broadband remote access server diverts the traffic to be accelerated to the standalone network, and forwards it to the user through the designated SRv6 tunnel.

[0096] Specifically, the acceleration module 13 also includes a core router, and for non-accelerated traffic, the independent network defaults to forwarding the traffic to the core router.

[0097] It is worth noting that the working process of each module in the application acceleration device 100 described in the embodiments of the present invention can refer to the working process of the application acceleration method described in the above embodiments, and will not be repeated here.

[0098] Compared with the prior art, the application acceleration device 100 disclosed in this invention has the following beneficial effects:

[0099] 1. By acquiring traffic data to be identified and identifying the traffic data to be identified, the application corresponding to the traffic data to be identified is identified. This application proposes three methods for identifying applications: fingerprint recognition with encrypted applications, DNS association recognition, and recognition by combination of data flow features and behaviors. Through heuristic rules, time series analysis, probabilistic graphical models, etc., the parameters are continuously trained to more accurately identify applications.

[0100] 2. For different types of applications, there are corresponding quality improvement analysis indicators, which allow users to analyze their application usage more comprehensively without being aware of it, provide a full range of quality improvement conclusions, locate different applications and give differentiated optimization effects, and obtain accurate positioning before accelerating media streams.

[0101] 3. For traffic that needs acceleration according to the poor quality analysis results, add it into the tunnel by coloring it with routing policies. Color the destination route with a specific color, then configure the next destination with redirection, and iterate the route into the specified SRv6 tunnel. The traffic to be accelerated is forwarded through the specified SRv6 tunnel, which can improve the user experience in the network.

[0102] See Figure 6 , Figure 6 This is a structural block diagram of an application acceleration device 200 provided in an embodiment of the present invention. The application acceleration device 200 includes a processor 21, a memory 22, and a computer program stored in the memory 22 and executable on the processor 21. When the processor 21 executes the computer program, it implements the steps in the above-described application acceleration method embodiments, such as steps S1 to S3.

[0103] For example, the computer program may be divided into one or more modules / units, which are stored in the memory 22 and executed by the processor 21 to complete the present invention. The one or more modules / units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the application acceleration device 200.

[0104] The application acceleration device 200 may include, but is not limited to, a processor 21 and a memory 22. Those skilled in the art will understand that the schematic diagram is merely an example of the application acceleration device 200 and does not constitute a limitation on the application acceleration device 200. It may include more or fewer components than illustrated, or combine certain components, or different components. For example, the application acceleration device 200 may also include input / output devices, network access devices, buses, etc.

[0105] The processor 21 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor. The processor 21 is the control center of the application acceleration device 200, connecting all parts of the application acceleration device 200 via various interfaces and lines.

[0106] The memory 22 can be used to store the computer programs and / or modules. The processor 21 implements various functions of the application acceleration device 200 by running or executing the computer programs and / or modules stored in the memory 22 and calling the data stored in the memory 22. The memory 22 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the mobile phone (such as audio data, phonebook, etc.). In addition, the memory 22 may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital card (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0107] Wherein, if the modules / units integrated in the application acceleration device 200 are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by the processor 21, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0108] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. An application acceleration method, characterized in that, include: Acquire traffic data to be identified and identify the application corresponding to the traffic data to be identified; Obtain the quality difference analysis index corresponding to the application type of the current application, and perform quality difference analysis on the traffic of the current application based on the quality difference analysis index; For traffic that needs acceleration according to the poor quality analysis results, the routes of the traffic to be accelerated are colored, and the application acceleration router is used to accelerate the traffic to be accelerated. Wherein, when the direction of the traffic data to be identified is from the user terminal to the server terminal, the acceleration of the traffic to be accelerated using the application acceleration router includes: The broadband remote access server diverts the traffic to be accelerated sent by the user to the independent network. The standalone network redirects the next-hop address in the virtual private network and iterates the route to the specified SRv6 tunnel; The application acceleration router matches the specified SRv6 tunnel based on the differential service code point in the packet, and forwards the traffic to be accelerated to the server through the specified SRv6 tunnel within the virtual private network.

2. The application acceleration method as described in claim 1, characterized in that, Identifying the traffic data to be identified includes: Extract feature data from the traffic data to be identified; The feature data is input into a pre-trained media stream recognition model to output the application recognition result of the traffic data to be identified.

3. The application acceleration method as described in claim 1, characterized in that, Identifying the traffic data to be identified includes: Extract the access requests of the traffic data to be identified; The IP address and domain name corresponding to the access request are resolved using DNS packets; Based on the IP address and the domain name, the application identification result of the traffic data to be identified is obtained.

4. The application acceleration method as described in claim 1, characterized in that, Identifying the traffic data to be identified includes: Extract the feature parameters from the traffic data to be identified; When the feature parameter matches the standard parameter of any target application pre-stored in the database, the target application is used as the application identification result of the traffic data to be identified.

5. The application acceleration method as described in claim 1, characterized in that, The application types include game applications, video applications, educational applications, and other applications besides the above three types. Different types of applications correspond to different quality difference analysis indicators.

6. The application acceleration method as described in claim 1, characterized in that, When the direction of the traffic data to be identified is from the server to the user, the acceleration of the traffic using the application acceleration router includes: The traffic to be accelerated sent by the server returns to the application acceleration router; The application acceleration router determines the next-hop broadband remote access server based on the routes published in the virtual private network and iteratively enters the specified SRv6 tunnel. The broadband remote access server diverts the traffic to be accelerated to the standalone network and forwards it to the user through a designated SRv6 tunnel.

7. An application acceleration device, characterized in that, include: An application identification module is used to acquire traffic data to be identified and to identify the application corresponding to the traffic data to be identified. The quality difference analysis module is used to obtain quality difference analysis indicators corresponding to the application type to which the current application belongs, and to perform quality difference analysis on the traffic of the current application based on the quality difference analysis indicators. An acceleration module is used to color the routes of traffic to be accelerated for traffic that needs acceleration according to the quality analysis results, and to accelerate the traffic to be accelerated using an application acceleration router. Wherein, when the direction of the traffic data to be identified is from the user terminal to the server terminal, the acceleration of the traffic to be accelerated using the application acceleration router includes: The broadband remote access server diverts the traffic to be accelerated sent by the user to the independent network. The standalone network redirects the next-hop address in the virtual private network and iterates the route to the specified SRv6 tunnel; The application acceleration router matches the specified SRv6 tunnel based on the differential service code point in the packet, and forwards the traffic to be accelerated to the server through the specified SRv6 tunnel within the virtual private network.

8. An application acceleration device, characterized in that, It includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor, when executing the computer program, implements the application acceleration method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device on which the computer-readable storage medium is located to perform the application acceleration method as described in any one of claims 1 to 6.

Images