Cpe-based dynamic game acceleration method and system

Abstract

This invention proposes a dynamic game acceleration method and system based on CPE (Customer Premises Equipment). By injecting multi-layer hook functions into the CPE device, it achieves accurate identification and dynamic filtering of data flows between user terminals and mobile networks, dynamically adapting to changes in the game server IP pool. When game data flows are detected, the CPE device initiates a dedicated channel establishment request to the mobile network base station, achieving rapid establishment and targeted routing of end-to-end dedicated bearers. The system continuously monitors transmission quality, dynamically triggers channel reconstruction or resource release, and supports real-time synchronization of the game IP address library through kernel-level communication mechanisms to optimize resource scheduling. This invention solves the problems of rigid resource allocation, low service identification accuracy, and high dedicated channel establishment latency in traditional static QoS schemes, achieving low-latency transmission of game traffic, high bandwidth utilization, and stable experience in multi-service concurrent scenarios. Its modular architecture combines flexibility and scalability, providing effective support for game acceleration applications in 5G / 6G networks.

Description

Technical Field

[0001] This invention relates to the field of network communication technology, and in particular to a dynamic game acceleration method and system based on CPE, specifically used to identify game data streams and establish dedicated transmission channels in a mobile network environment to optimize the network transmission quality of real-time interactive applications. Background Technology

[0002] Current game acceleration technologies in mobile network environments face three major technical bottlenecks. First, traditional Quality of Service (QoS) optimization schemes generally employ static resource allocation mechanisms, failing to dynamically identify and differentiate different types of game services, such as FPS and MOBA. This results in a mismatch between network resource allocation and actual service demands, leading to low overall bandwidth utilization efficiency. Second, existing network equipment generally uses generic traffic scheduling algorithms, lacking the ability to accurately identify game traffic characteristics. This makes it impossible to effectively distinguish real-time game interaction data from ordinary background data streams, causing significant latency fluctuations during game interactions and severely impacting user experience. Third, the lack of a dedicated end-to-end bearer linkage mechanism prevents mobile network base stations from responding to user terminal service demands in real time.

[0003] The deep-seated technical obstacles stem from the design limitations of the existing architecture. On the one hand, the service identification mechanism relies excessively on fixed IP addresses and port numbers, making it unsuitable for the dynamic load balancing strategies and elastic IP resource pool deployment modes commonly used in modern game service architectures. This results in a continuous decline in traffic identification accuracy over time. On the other hand, there is architectural isolation between the network policy enforcement layer and the underlying bearer control layer. The CPE (Customer Premise Equipment) edge cannot coordinate and link with the mobile network base station. When a dedicated transmission channel needs to be established for game services, the signaling interaction process is complex and slow, making it difficult to meet the network agility requirements of real-time interactive services.

[0004] Existing solutions attempt to deploy acceleration functions separately on the base station or terminal side, but neither has solved the core problem of the lack of intelligence in edge networks. Base station-side solutions, due to their distance from user terminals, cannot promptly perceive real-time channel quality changes in the access network segment; terminal-side solutions are limited by device computing power and system permissions, making deep network layer optimization difficult. Especially in 5G network environments, frequent base station handovers and wireless signal fluctuations exacerbate these problems. Traditional static rule-configured game acceleration mechanisms are particularly inadequate in dynamic network environments, urgently requiring an innovative architecture capable of achieving intelligent perception, dynamic decision-making, and resource coordination at the network edge. Summary of the Invention

[0005] Based on the above situation, the present invention aims to provide a dynamic game acceleration method and system based on CPE. By injecting hook functions into the CPE device to accurately identify game data streams, and combining dynamic matching rules with base stations to establish dedicated channels, the present invention enables real-time monitoring of transmission quality and on-demand allocation of resources, thereby solving the problems of high latency and low resource utilization in the prior art.

[0006] The technical solution adopted by this invention to solve its technical problem is as follows:

[0007] In a first aspect, the present invention provides a dynamic game acceleration method based on CPE, comprising the following steps:

[0008] S1, inject hook functions into the CPE device. The hook functions include interface monitoring hooks, which are used to intercept data streams transmitted between the user terminal and the external network through the Modem, WiFi or Ethernet network interface of the CPE device.

[0009] S2, the data stream is filtered according to the matching rules pre-configured on the CPE device. The matching rules can be based on one or more parameters of interface type, data flow direction, IP 5-tuple, and support subnet-level matching through mask;

[0010] S3, when a game data stream that matches the matching rules is detected, the CPE device initiates a dedicated channel establishment request to the mobile network base station based on the identification information of the data stream, carrying the IP five-tuple parameters of the successful match;

[0011] S4. After the mobile network base station confirms that the dedicated channel has been successfully established, the CPE device will route the matching game data stream to the dedicated channel.

[0012] S5, the CPE device continuously monitors the transmission quality of the game data stream. When an abnormal transmission quality is detected, it triggers the reconstruction of the dedicated channel. When the game data stream is detected to be ending, it sends a channel release command to the base station to reclaim the dedicated transmission resources.

[0013] Furthermore, hook functions also include:

[0014] Protocol parsing hooks are used to extract IP 5-tuple features from data streams and identify TCP connection states;

[0015] The business identification hook is used to match the extracted features with preset rules in real time and identify connection establishment and release messages;

[0016] Routing decision hooks are used to dynamically adjust the transmission path of matching data streams based on the channel establishment results;

[0017] The hook functions work together in sequence according to the data processing flow: the interface monitoring hook captures the raw data stream and passes it to the protocol parsing hook. The protocol parsing hook extracts features and then hands them over to the business identification hook for rule matching. After the business identification hook confirms the matching result, it triggers the routing decision hook to perform path adjustment, forming a complete data stream processing chain.

[0018] Furthermore, the matching rules support dynamic updates, and this method also includes the following steps:

[0019] S6, the CPE device maintains real-time synchronization of the game server address database with the service identification hook through a kernel-level communication mechanism, and obtains address change information of the game server cluster;

[0020] S7, the CPE device automatically updates the matching rules based on the address change information, and keeps the data flow identification mechanism synchronized with the game server deployment strategy by dynamically adjusting the IP address range and its mask parameters.

[0021] Furthermore, step S2 specifically includes the following steps:

[0022] S21, a hierarchical matching mechanism is pre-configured in the CPE device. The hierarchical matching mechanism matches the data flow based on the interface type, data flow direction and IP five-tuple in sequence, and supports subnet-level matching through the mask.

[0023] S22, perform game feature identification on the data stream that has been matched hierarchically to confirm that it is a game data stream;

[0024] S23 records the IP 5-tuple information of the confirmed game data stream for use in subsequent dedicated channel establishment requests.

[0025] Furthermore, step S3 specifically includes the following steps:

[0026] S31, When a game data stream that matches the matching rules is detected, the CPE device identifies the application characteristics of the game data stream and determines the corresponding transmission priority;

[0027] S32, the CPE device constructs an enhanced channel establishment request based on the identification result, the request including the successfully matched IP 5-tuple, transmission quality requirement parameters and application priority identifier;

[0028] S33, the CPE device dynamically evaluates the load status and link quality of nearby mobile network base stations and selects the optimal base station that meets the transmission quality requirements;

[0029] S34, Send a request to the selected base station to establish a dedicated channel with encrypted verification;

[0030] S35, if no confirmation response is received from the base station, the CPE device initiates channel establishment requests to other base stations in sequence according to the base station candidate sequence.

[0031] Furthermore, step S4 specifically includes the following steps:

[0032] S41, after receiving the dedicated channel confirmation information returned by the mobile network base station, the CPE device creates a stream mapping table to record the correspondence between the matched game data streams and the dedicated channels;

[0033] S42 performs lossless conversion on each data group in the matched game data stream, adding a dedicated channel identifier while maintaining the integrity of the original content;

[0034] S43, Dynamic priority scheduling is implemented for data groups in the data stream based on the characteristics of game interaction to ensure that highly sensitive interactive data is transmitted through a dedicated channel first;

[0035] S44 establishes a session continuity guarantee mechanism to maintain stable transmission of game sessions during network switching or connection fluctuations.

[0036] S45 monitors the performance differences between dedicated channels and regular transmission paths in real time and dynamically adjusts the routing strategy that matches the game data stream to optimize transmission quality.

[0037] Furthermore, step S5 specifically includes the following steps:

[0038] S51, the CPE device performs real-time quality assessment of game data streams based on multi-dimensional transmission indicators, including latency jitter, packet loss mode and throughput stability.

[0039] S52, when a continuous degradation in transmission quality is detected, the CPE device starts a predictive channel reconstruction mechanism, and then performs traffic switching after the new channel is established to ensure the continuity of the gaming experience;

[0040] S53, CPE devices distinguish between temporary sleep states and actual session termination by analyzing the interaction characteristics of data streams and traffic sleep modes;

[0041] S54, upon confirming the complete termination of the game session, the CPE device sends a hierarchical resource release command to the base station to achieve orderly reclamation of dedicated transmission resources and network load balancing.

[0042] Furthermore, step S5 also includes a step of dynamic service quality adjustment, specifically:

[0043] S55, the CPE device dynamically adjusts the service quality parameters of the dedicated channel based on real-time transmission quality indicators, including bandwidth allocation priority and traffic scheduling strategy;

[0044] S56: When transmission quality is abnormal, service quality parameters are renegotiation is triggered, and the transmission stability of the dedicated channel is optimized through base station collaboration.

[0045] Secondly, the present invention provides a CPE-based dynamic game acceleration system, which is configured in a CPE device and includes:

[0046] The hook function injection unit, configured in the CPE device, is used to intercept the data stream transmitted between the user terminal and the external network through the CPE device's Modem, WiFi, or Ethernet network interface;

[0047] The data stream filtering unit is used to filter the data stream according to pre-configured matching rules. The matching rules are based on one or more parameters of interface type, data flow direction, and IP 5-tuple, and support subnet-level matching through mask.

[0048] The dedicated channel establishment unit is used to initiate a dedicated channel establishment request to the mobile network base station based on the identification information of the game data stream when a game data stream that matches the matching rules is detected, carrying the IP five-tuple parameters of the successfully matched stream.

[0049] The data stream routing unit is used to route the matching game data stream to the dedicated channel after the mobile network base station confirms that the dedicated channel has been successfully established;

[0050] The transmission quality monitoring unit is used to continuously monitor the transmission quality of the game data stream. When an abnormal transmission quality is detected, it triggers the reconstruction of the dedicated channel. When the game data stream is detected to be over, it sends a channel release command to the base station to reclaim the dedicated transmission resources.

[0051] The dynamic rule update unit is used to maintain real-time synchronization of the game server address database with the business identification module through a kernel-level communication mechanism, obtain address change information of the game server cluster, and automatically adjust the matching rules based on the address change information. By updating the IP address range and its mask parameters, the data flow identification mechanism and the game server deployment strategy are kept dynamically synchronized.

[0052] Furthermore, the hook function injection unit includes:

[0053] The interface monitoring module is used to capture the raw data stream;

[0054] The protocol parsing module is used to extract the IP 5-tuple characteristics of the data stream and identify the TCP connection state;

[0055] The service identification module is used to match the extracted features with preset rules in real time and identify connection establishment and release messages;

[0056] The routing decision module is used to dynamically adjust the transmission path of the matching data stream based on the channel establishment results;

[0057] The interface monitoring module, protocol parsing module, service identification module, and routing decision module work together in sequence according to the data processing flow to form a complete data flow processing chain.

[0058] The beneficial effects of this invention include:

[0059] Firstly, this invention injects multi-level hook functions into CPE devices, encompassing interface monitoring, protocol parsing, service identification, and routing decisions, forming a data flow processing chain to achieve accurate identification and dynamic filtering of game data flows. Based on multi-parameter matching rules such as interface type, data flow direction, and IP 5-tuple, as well as subnet-level mask adaptation capabilities, it effectively distinguishes between game and non-game traffic, avoiding resource waste and improving network resource utilization efficiency. Simultaneously, it supports real-time synchronization of the game server address database through a kernel-level communication mechanism, automatically updating matching rules to ensure dynamic consistency between the data flow identification mechanism and the game server deployment strategy, overcoming the shortcomings of traditional static rules that cannot adapt to changes in server clusters.

[0060] Secondly, upon detecting game data streams, the CPE device initiates a dedicated channel establishment request to the mobile network base station as needed, carrying the successfully matched IP five-tuple parameters to achieve targeted and isolated transmission of service traffic. By evaluating base station load and link quality, the optimal transmission path is selected, and an encryption verification mechanism ensures channel security, reducing the impact of cross-network detours and public network congestion. Furthermore, based on the characteristics of game interaction, a dynamic priority scheduling and session continuity guarantee mechanism are implemented to maintain a low-latency, highly reliable transmission experience during network switching or connection fluctuations, preventing game lag or interruption.

[0061] Thirdly, this invention continuously monitors the transmission quality of game data streams, predicts performance degradation trends based on multi-dimensional indicators such as latency jitter and packet loss patterns, and triggers channel reconstruction in advance to reduce the risk of experience interruption. It achieves orderly reclamation of dedicated transmission resources through tiered resource release commands, avoiding network load imbalance. Simultaneously, it supports maintaining channel reservation when a temporary sleep state is detected, balancing resource consumption and user experience, and solving the problem of rigid resource allocation in traditional static QoS schemes.

[0062] Fourthly, the system of this invention adopts a modular design with functional modules such as a hook function injection unit, a data flow filtering unit, and a dedicated channel establishment unit. Each unit works collaboratively according to the data processing flow to form a complete data flow processing chain. Through the decoupling design of the interface monitoring, protocol parsing, service identification, and routing decision modules, the system's flexibility and maintainability are improved, facilitating subsequent functional expansion or parameter optimization and adapting to changes in the needs of different network environments and business scenarios. Attached Figure Description

[0063] Figure 1This is a flowchart of a dynamic game acceleration method based on CPE provided in an embodiment of the present invention.

[0064] Figure 2 This is a schematic diagram of the structure of a CPE-based dynamic game acceleration system provided in an embodiment of the present invention. Detailed Implementation

[0065] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0066] To address the technical issues of high latency and jitter in mobile network environments, please refer to [reference needed]. Figure 1 This invention provides a dynamic game acceleration method based on CPE, the method comprising the following steps:

[0067] S1, inject hook functions into the CPE device. The hook functions include interface monitoring hooks, which are used to intercept data streams transmitted between the user terminal and the external network through the network interface of the CPE device.

[0068] Specifically, the ability of CPE devices to inject hook functions stems primarily from their open architecture design based on the Linux kernel. Modern CPE devices generally employ operating systems that support dynamic loading of kernel modules, with standardized data processing nodes pre-defined in the kernel network protocol stack. On these nodes, the system allows the registration of custom callback functions, which are automatically triggered when data packets pass through specific processing stages. CPE device manufacturers typically provide kernel development interfaces and network subsystem extension capabilities, enabling third-party functional modules to embed processing logic without modifying the core code. Simultaneously, the device possesses sufficient computing resources and memory space to support additional processing overhead, and kernel-level communication mechanisms ensure that hook functions can efficiently access network data structures. This architectural design maintains system stability while endowing CPE devices with flexible business awareness and decision-making capabilities, enabling them to achieve deep data flow analysis and intelligent routing control functions. The entire hook injection process is secured through digital signature verification, preventing unauthorized code from interfering with normal network services.

[0069] The core working principle of the interface monitoring hook injected into CPE devices is to set interception points on the network data processing path of the device's operating system kernel. When the data flow passes through these key nodes, preset processing functions are automatically triggered. Specifically, the interface monitoring hook function is implanted into the data packet receiving and sending stages of the network protocol stack, forming a transparent monitoring layer that can capture all data traffic passing through the CPE device without interrupting normal communication. When a user terminal communicates with an external network, the data flow is first intercepted by the interface monitoring hook, and then sequentially processed through multiple levels of hooks, including protocol parsing, service identification, and routing decisions. Each level of hook is responsible for a specific analysis task, ultimately forming a complete data flow processing chain. It should be noted that "external network" refers to all network entities connected to the CPE device's wide area network side, specifically including mobile communication network infrastructure such as 4G / 5G base stations and core networks, Internet service nodes such as game server clusters, content delivery networks, and cross-carrier interconnection networks. The interface monitoring hook function intercepts communication traffic that crosses the CPE device's network boundary, excluding local communication between terminals within the same local area network.

[0070] The main function of this hook mechanism is to endow CPE devices with intelligent identification and decision-making capabilities, enabling them to distinguish between ordinary data streams and real-time interactive services such as games, and to optimize transmission paths accordingly. Through hook functions, CPE devices can extract characteristic information of data streams in real time, dynamically adjust routing strategies, and collaborate with mobile network base stations to establish dedicated transmission channels. Compared to the passive forwarding mode of traditional network devices, hook injection achieves proactive intelligent control at the network edge, significantly improving the transmission quality and user experience of real-time services without altering the existing network architecture, while avoiding any modifications to terminal devices and game applications.

[0071] S2, the data stream is filtered according to the matching rules pre-configured on the CPE device. The matching rules can be based on one or more parameters of interface type, data flow direction, IP 5-tuple, and support subnet-level matching through mask.

[0072] Specifically, upon startup, the CPE device loads a pre-configured set of matching rules. These rules are prioritized according to interface type, for example, traffic passing through cellular network interfaces is processed first. When a data stream is captured by a hook function, the system first determines its source interface and transmission direction, then extracts network characteristics such as source and destination addresses, port numbers, and protocol types. The matching engine employs a hierarchical filtering strategy, first performing preliminary screening based on interface type and data flow direction, and then performing fine-grained matching of candidate data streams using IP 5-tuples. For large game service providers, the system utilizes network masking technology to merge multiple consecutive IP addresses into a single rule item, achieving efficient matching at the subnet level, reducing the number of rules and improving processing speed. When a data stream's characteristics match any rule condition, the system marks the stream as potential game traffic and passes it to the next processing stage. Unmatched ordinary data streams continue to be forwarded according to standard routing procedures. The entire filtering process is completed at the CPE device kernel level, completely transparent to users and applications, and does not affect normal communication performance.

[0073] S3. When a game data stream that matches the matching rules is detected, the CPE device initiates a dedicated channel establishment request to the mobile network base station based on the identification information of the data stream, carrying the IP five-tuple parameters of the successful match.

[0074] Specifically, when the CPE device identifies a data stream matching preset game characteristics, the system immediately extracts the complete network identification information of the data stream, including source address, destination address, source port, destination port, and protocol type. These parameters are encapsulated into a structured channel establishment request message and sent to the currently serving mobile network base station via the control plane signaling channel. The request message uses an encrypted signature mechanism to ensure communication security and carries the service quality requirement level of the data stream. After receiving the request, the base station allocates dedicated transmission resources based on the real-time network load and establishes a corresponding data forwarding path. The entire request process uses an asynchronous processing mechanism, which does not affect the user's normal network communication. If the initial request does not receive a response, the system will automatically select a nearby base station for retrying to ensure the reliability of the dedicated channel establishment. This instant-triggered resource allocation method significantly shortens the waiting time of game data streams and provides network quality assurance for real-time interactive services.

[0075] S4. After the mobile network base station confirms the successful establishment of the dedicated channel, the CPE device will route the matching game data stream to the dedicated channel.

[0076] Specifically, after the mobile network base station returns confirmation of successful establishment of the dedicated channel, the CPE device immediately creates a high-priority forwarding rule in the kernel routing table, binding the IP 5-tuple matching the game data stream with the dedicated channel identifier. The system performs lossless encapsulation on the game data stream flowing through the hook function, adding a dedicated channel header while preserving the original content to ensure the base station can identify and prioritize its processing. To ensure a smooth handover process, the CPE device adopts a gradual traffic migration strategy, first importing critical interactive data packets into the dedicated channel, and then gradually transferring all game traffic. During the dedicated channel's operation, the system continuously compares the transmission latency and packet loss rate of the ordinary channel and the dedicated channel, dynamically adjusting the data packet scheduling strategy. When network switching or signal fluctuations are detected, the pre-established backup channel automatically takes over the transmission task, with the entire process seamless for the user. This refined routing management mechanism provides game data with an independent transmission path, effectively isolating it from interference from ordinary data traffic and significantly improving the real-time interactive experience.

[0077] S5, the CPE device continuously monitors the transmission quality of the game data stream. When an abnormal transmission quality is detected, it triggers the reconstruction of the dedicated channel. When the game data stream is detected to be ending, it sends a channel release command to the base station to reclaim the dedicated transmission resources.

[0078] Specifically, after establishing a dedicated channel, the CPE device continuously tracks the transmission performance of the game data stream, evaluating network quality through real-time analysis of multi-dimensional indicators such as packet arrival interval, retransmission frequency, and throughput stability. When the system detects a continuous increase in latency or a packet loss rate exceeding a threshold, it automatically initiates a predictive channel reconstruction process, pre-establishing a new channel in the background and seamlessly switching traffic once ready, avoiding game interruptions. Simultaneously, the device accurately distinguishes between brief pauses and session terminations by recognizing the interaction characteristics and sleep modes of the data stream. When it is confirmed that the game session has truly ended, the CPE device sends a tiered resource release command to the base station, first reducing bandwidth priority and then gradually reclaiming transmission resources to ensure a smooth transition of network load. This intelligent monitoring and dynamic adjustment mechanism ensures the continuity of the gaming experience while avoiding the ineffective occupation of dedicated resources, achieving efficient utilization of network resources.

[0079] It should be noted that in traditional CPE device technology, the application of hook functions is mainly limited to the extension of basic network functions, such as simple traffic statistics, static firewall rule enforcement, or basic QoS marking. These traditional hooks typically work independently, have single functions, and fixed processing logic, making it impossible to achieve deep business awareness and dynamic decision-making. The hook mechanisms of most commercial CPE devices are only used for internal diagnostics or vendor-specific functions, lacking open business identification and collaborative control capabilities.

[0080] The hook functions injected in this solution first construct a four-level collaborative hook architecture, forming a complete data flow processing chain: the interface monitoring hook captures raw traffic, the protocol parsing hook extracts network layer features, the service identification hook performs application-level judgment, and the routing decision hook executes dynamic path adjustment. This pipelined processing mechanism breaks through the functional limitations of traditional single-point hooks. Second, the hook functions have dynamic learning capabilities, enabling real-time synchronization of game server address database changes through kernel-level communication mechanisms and automatic updates to matching rules, solving the identification failure problem caused by dynamic deployment of game services. Most importantly, the hook mechanism in this solution achieves cross-network domain collaborative control, directly triggering the establishment process of a dedicated channel with mobile network base stations, connecting edge intelligence with core network resource scheduling, forming an end-to-end closed-loop optimization system. This deeply integrated hook architecture not only has high identification accuracy but also fast response speed, completing the entire process from traffic identification to dedicated channel establishment within 200 milliseconds, providing network quality assurance for real-time interactive services that traditional CPE devices cannot achieve.

[0081] To enable CPE devices to intelligently identify game data streams, achieve real-time monitoring of transmission quality and dynamic scheduling of dedicated channels, thereby improving the gaming experience and optimizing network resource utilization efficiency, the hook functions also include:

[0082] Protocol parsing hooks are used to extract IP 5-tuple features from data streams and identify TCP connection states.

[0083] Specifically, the protocol parsing hook is deployed on the packet processing path of the CPE device's kernel network stack, and is triggered immediately after the interface monitoring hook captures the raw data stream. This hook deeply parses the network layer and transport layer header information of each packet, accurately extracting the five-tuple features including source address, destination address, source port, destination port, and protocol type, while monitoring TCP flag changes to identify the connection status. When the SYN flag is detected, it is determined to be in the connection establishment phase. The sequence number changes during data transmission are analyzed to determine session activity, and connection termination is identified by the FIN or RST flag. The parsing results are encapsulated into a structured feature vector and passed to the service identification hook in real time for further matching decisions. The entire parsing process is completed in kernel space, using zero-copy technology to ensure processing efficiency, maintaining line-speed processing capability even under high network load, with no perceptible impact on the user's communication experience.

[0084] It's important to note that in online gaming scenarios, when a player exits the game or closes the application, the client and server undergo a standard connection release process via the TCP protocol. Our CPE device leverages this fundamental network characteristic to implement a precise mechanism for reclaiming dedicated channel resources. Specifically, the CPE sets up dedicated interface monitoring hooks and protocol parsing hooks at the kernel driver layer, enabling real-time monitoring of all data flows passing through the device. When a game session ends, the client and server exchange TCP control packets with FIN or RST flags, which are standard connection termination signals. The CPE's hook functions can recognize these special flags, rather than simply relying on fixed timeout mechanisms or application-layer notifications. Once a TCP connection release message is detected, the CPE immediately initiates the resource reclamation process. The system first confirms that this is a genuine session termination rather than a temporary network fluctuation, then sends a standard channel release command to the base station, while simultaneously clearing the corresponding dedicated channel configuration in the local routing table. This process typically completes within 100 milliseconds, far faster than traditional timed reclamation mechanisms.

[0085] Compared to traditional solutions, this method has significant advantages. Ordinary routers or accelerators often employ fixed-duration session persistence strategies; for example, after game traffic stops, resources are only released after 30 seconds or even minutes, during which valuable network channels are inefficiently occupied. This solution, based on the semantics of the TCP protocol itself, can distinguish between game pauses (such as switching to a chat application) and actual game exits, avoiding resource waste or premature release.

[0086] In actual deployment, complex scenarios in mobile network environments have been specifically considered. For example, when players switch networks, such as switching from WiFi to 4G / 5G and exiting the game, or encountering temporary signal interruptions, the system can accurately determine whether the session has truly ended through the connection status tracking table, rather than simply making a decision based on a single data packet. This resource management approach based on protocol semantics ensures the continuity of the gaming experience and improves the overall utilization efficiency of network resources, especially in high-concurrency user scenarios.

[0087] The business identification hook is used to match the extracted features with preset rules in real time and identify connection establishment and release messages.

[0088] Specifically, the service identification hook executes immediately after the protocol parsing hook completes feature extraction, receiving a data stream feature vector containing the IP 5-tuple and connection status. This hook rapidly compares the real-time features with a pre-built game service rule base within the CPE device, which includes address ranges and service port characteristics of mainstream game servers. The matching process employs a multi-level filtering mechanism, first verifying the address and port range, then combining connection behavior characteristics for a comprehensive judgment. Simultaneously, the hook continuously monitors key packet markers in the data stream. When a SYN packet is detected, it is identified as a new connection establishment request, and the session start time is recorded; when a FIN or RST packet sequence is captured, the connection is determined to be entering the release phase. For successfully matched game data streams, the hook generates a service identification tag and marks the connection lifecycle status, promptly transmitting it to the routing decision hook. The entire identification process runs efficiently in kernel space, ensuring millisecond-level response speeds and preventing perceptible network latency for users, providing accurate business decision-making basis for subsequent dedicated channel establishment.

[0089] The routing decision hook is used to dynamically adjust the transmission path of the matching data stream based on the channel establishment result.

[0090] Specifically, the routing decision hook is activated after the service identification hook confirms the game data flow and the dedicated channel is established. This hook receives channel confirmation information from the base station and dynamically updates the forwarding strategy for the corresponding game flow in the kernel routing table. The system creates an independent routing context for the matching data flow, redirecting packets that originally passed through the general forwarding path back to the dedicated channel. During path switching, the hook employs a gradual migration mechanism, prioritizing the forwarding of highly sensitive interactive packets to ensure a seamless experience for the user. When a degradation in dedicated channel quality is detected, the hook automatically initiates a backup path switching process, smoothly migrating traffic once the new channel is ready. If channel reconstruction fails, the system immediately reverts to the normal transmission path and marks the abnormal state. The entire routing decision process is completed in real-time at the kernel level. Combined with a connection state tracking mechanism, it ensures sequential packet transmission and session continuity, thereby improving the priority of game data transmission and maintaining overall network stability.

[0091] Hook functions work collaboratively in the order of data processing flow: after the interface monitoring hook captures the raw data stream, it is passed to the protocol parsing hook. After the protocol parsing hook extracts features, it is handed over to the business identification hook for rule matching. After the business identification hook confirms the matching result, it triggers the routing decision hook to perform path adjustment, forming a complete data stream processing chain.

[0092] Specifically, upon CPE device startup, four levels of hook functions are loaded sequentially and registered to specific processing nodes in the kernel network stack, forming a pipelined data processing chain. When a user terminal initiates network communication, the interface monitoring hook first captures the raw data packet and passes it directly to the protocol parsing hook without any modification. The protocol parsing hook deeply analyzes the packet header information, extracts complete network characteristics and connection status, and passes the structured data to the service identification hook. The service identification hook compares real-time characteristics with a pre-set rule base to confirm whether it is game traffic and determine the session stage, forwarding the identification results and decision instructions to the routing decision hook. Finally, the routing decision hook dynamically adjusts the forwarding path based on the channel status and implements a priority transmission strategy for matching data streams. Communication between hooks at each level is achieved through kernel-level message queues, employing zero-copy technology to ensure efficient data flow. The entire processing is completed within microseconds, with no perceptible impact on the user's network experience. This tightly collaborative chain architecture enables the CPE device to perceive service needs in real time and autonomously optimize transmission paths, realizing an intelligent decision-making closed loop at the network edge.

[0093] It should be noted that in traditional CPE device technology, the application of hook functions is mainly limited to the extension of basic network functions, such as simple traffic statistics, static firewall rule enforcement, or basic QoS marking. These traditional hooks typically work independently, have single functions, and fixed processing logic, making it impossible to achieve deep business awareness and dynamic decision-making. The hook mechanisms of most commercial CPE devices are only used for internal diagnostics or vendor-specific functions, lacking open business identification and collaborative control capabilities.

[0094] The hook functions injected in this solution first construct a four-level collaborative hook architecture, forming a complete data flow processing chain. Second, the hook functions possess dynamic learning capabilities, enabling real-time synchronization of game server address database changes through kernel-level communication mechanisms and automatic updates to matching rules, resolving the identification failure issue caused by dynamic deployment of game services. Most importantly, this solution's hook mechanism achieves cross-network domain collaborative control, directly triggering the establishment process of a dedicated channel with mobile network base stations, connecting edge intelligence with core network resource scheduling to form an end-to-end closed-loop optimization system. This deeply integrated hook architecture not only boasts high identification accuracy but also fast response speed, completing the entire process from traffic identification to dedicated channel establishment within 200 milliseconds, providing network quality assurance for real-time interactive services that traditional CPE devices cannot achieve.

[0095] To ensure accurate matching of data stream identification with the game server's dynamic deployment strategy, address the lag issue of traditional static rules, and improve the real-time performance and accuracy of business identification, this solution uses matching rules that support dynamic updates. The method also includes the following steps:

[0096] S6, the CPE device works with the service identification hook through a kernel-level communication mechanism to maintain real-time synchronization of the game server address database and obtain address change information of the game server cluster.

[0097] Specifically, achieving real-time synchronization of the game server address library in a CPE device involves efficient collaboration between the kernel layer and the user layer. The core process is as follows:

[0098] First, the CPE device's main control program acts as a synchronization hub, periodically calling the API interface provided by the game developer to obtain the latest server address list. This list is transmitted to the CPE user layer in a structured data format, where the main control program parses out the IP address ranges and their associated subnet masks. To ensure data integrity, the main control program compares the old and new address databases and generates incremental update instructions.

[0099] Subsequently, the main control program encapsulates the update command into a message packet of a specific format through the Netlink communication mechanism provided by the operating system kernel, and directly transmits it to the service identification hook module at the kernel layer. Upon receiving the command, the service identification hook immediately parses the IP address range and mask parameters, and dynamically updates the hash table matching rules it maintains. This process ensures that rule changes take effect immediately at the kernel level, without requiring a driver or device restart.

[0100] Ultimately, the business identification hook feeds the updated results back to the main control program via the Netlink channel, forming a closed-loop verification. When new game data flows through the CPE, the business identification hook can directly perform accurate matching based on the latest address database, achieving second-level synchronization from address change awareness to rule effectiveness. This mechanism effectively solves the identification lag problem caused by dynamic adjustments to the game server IP pool, ensuring the continuity and accuracy of the acceleration service.

[0101] S7, the CPE device automatically updates the matching rules based on the address change information, and keeps the data flow identification mechanism synchronized with the game server deployment strategy by dynamically adjusting the IP address range and its mask parameters.

[0102] Specifically, the dynamic updating of matching rules in the CPE device mainly revolves around the collaborative working mechanism between the main control program and the kernel-level service identification module. Its core process is as follows:

[0103] After receiving information about the game server address change, the main control program first parses and verifies the update command, extracting the IP address range to be added or deleted and its corresponding subnet mask parameters. To ensure the integrity and security of the command, the main control program calculates the command's verification and compares it with the original data. Only after successful verification will subsequent operations be executed.

[0104] Subsequently, the main control program encapsulates the verified update command into a message packet of a specific format through the Netlink communication interface provided by the operating system kernel and sends it to the service identification module at the kernel layer. Upon receiving the command, the service identification module immediately parses the IP address range and mask parameters and dynamically updates the hash table matching rules it maintains. This process ensures the atomicity of rule changes through atomic operations, avoiding data inconsistency issues caused by concurrent access.

[0105] After the update is complete, the business identification module feeds back the operation result to the main control program via the Netlink channel, forming a closed-loop verification mechanism. The main control program records the update log and can send notifications to the user layer as needed, indicating that the rules have been synchronized and taken effect. When new game data flows through the CPE, the business identification module can perform accurate matching based on the latest IP address range and mask parameters, achieving real-time synchronization between the data flow identification mechanism and the game server deployment strategy. This solution, through its hardware and software co-design and closed-loop verification mechanism, effectively solves the identification lag problem caused by dynamic changes in the game server IP pool, ensuring the accuracy and continuity of the game acceleration service.

[0106] In order to provide accurate information for the establishment of dedicated channels and effectively avoid misjudgments and waste of resources, step S2 specifically includes the following steps:

[0107] S21, a hierarchical matching mechanism is pre-configured in the CPE device. The hierarchical matching mechanism matches the data flow based on the interface type, data flow direction and IP 5-tuple in sequence, and supports subnet-level matching through the mask.

[0108] Specifically, implementing a hierarchical matching mechanism in CPE devices involves the collaborative design of software configuration and data stream processing procedures. Its core lies in improving the efficiency and accuracy of game data stream identification through a layered filtering strategy.

[0109] First, in the control layer of the CPE device, the main control program receives matching rule configuration instructions from the user via a web or app interface. These rules include interface type, data flow direction, and IP 5-tuple parameters. Interface type covers network interfaces such as Modem, WiFi, and Ethernet; data flow direction specifies whether it is receiving or sending; the IP 5-tuple includes source IP, destination IP, source port, destination port, and transport protocol type. To support subnet-level matching, users can configure the mask parameter for IP address ranges, for example, setting the mask width to 25 to 26 bits to cover the continuous address range in the game server's IP pool.

[0110] After configuration, the traffic_acc driver module of the service identification layer stores these rules in a hash table according to priority. The matching process is executed sequentially: when data flows through the CPE device, the interface monitoring hook first performs primary filtering based on the interface type, retaining only packets from the specified interface; subsequently, the protocol parsing hook further filters according to the data flow direction, distinguishing between uplink and downlink traffic; finally, the service identification hook extracts the IP 5-tuple features and performs subnet matching in conjunction with the mask parameters. For example, if the destination IP address matches the result of a bitwise AND operation with the IP range configured in the rule using the mask, a successful match is determined.

[0111] The entire matching process is optimized for efficiency through hash table lookups, ensuring millisecond-level response times. This hierarchical mechanism effectively reduces interference from irrelevant traffic, while subnet masks adapt to dynamic IP environments, providing accurate input for the establishment of subsequent dedicated channels.

[0112] S22, perform game feature identification on the data stream that has been matched by hierarchical classification, and confirm that it is a game data stream.

[0113] Specifically, game feature recognition in CPE devices primarily relies on the collaborative work of the business recognition layer in deep data stream analysis and rule matching. After the data stream is initially filtered through a hierarchical matching mechanism, the business recognition module drives the business recognition hook to immediately intervene and perform in-depth analysis of the data packet content to confirm its game attributes.

[0114] First, the service identification module extracts transport layer features from the data stream, focusing on the connection status under TCP or UDP protocols. For TCP traffic, the module detects whether it is a three-way handshake initiation message for connection establishment or a FIN message for connection termination; these key nodes typically mark the start or end of a game session. Simultaneously, the module analyzes the payload characteristics of the data packets, such as packet size distribution and packet interval patterns. Game traffic often exhibits a consistent high-frequency characteristic of small packets, unlike the continuous large packets in video streams or the bursty patterns of web browsing.

[0115] The module then compares the extracted features with a pre-defined game rule base in real time. This rule base includes common port ranges, protocol interaction patterns, and application-specific fingerprint information for games. For example, some games use fixed UDP port ranges for real-time data transmission, while others expose their identity through specific ALPN identifiers in the TLS handshake. During the matching process, the module combines IP 5-tuple and subnet mask information to ensure accurate identification and avoid misjudgments.

[0116] Ultimately, when the data stream characteristics closely match the rule base, the business identification module identifies the stream as a game data stream and triggers subsequent processing. This embodiment effectively distinguishes game traffic from ordinary business traffic through multi-dimensional feature cross-validation, providing reliable input for the establishment of dedicated channels.

[0117] S23 records the IP 5-tuple information of the confirmed game data stream for use in subsequent dedicated channel establishment requests.

[0118] Specifically, the recording of IP 5-tuple information for game data streams in CPE devices primarily revolves around the collaborative work of the service identification layer and the control layer. Once the data stream is confirmed through hierarchical matching and game feature identification, the service identification module immediately initiates the recording process. This module first extracts the complete IP 5-tuple characteristics of the data stream, including the source IP address, destination IP address, source port number, destination port number, and transport protocol type, and encapsulates this information into a structured data object. Subsequently, the service identification module transmits the IP 5-tuple information to the main control program in the control layer via kernel shared memory or a message queue mechanism. After receiving the data, the main control program stores it in a dynamic flow mapping table, which is organized using a hash structure to support fast querying and updating. During the recording process, the main control program adds a timestamp and session identifier to each entry to track the lifecycle of the data stream. After recording is complete, the IP 5-tuple information serves as a key parameter for the dedicated channel establishment request. When the control layer detects that the channel establishment conditions are met, it automatically triggers the subsequent AT command assembly and base station interaction process. The entire implementation ensures the real-time and accuracy of information recording through hardware and software collaboration, providing a reliable data foundation for the rapid establishment of dedicated channels.

[0119] In order to provide differentiated transmission services for different games, step S3 specifically includes the following steps:

[0120] S31, when a game data stream that matches the matching rules is detected, the CPE device identifies the application characteristics of the game data stream and determines the corresponding transmission priority.

[0121] Specifically, once the CPE device identifies a game data stream that matches preset rules, the system immediately initiates a deep feature analysis process. This process identifies the specific game type by detecting the protocol interaction mode, transmission frequency characteristics, and payload structure of the data packets, such as distinguishing between real-time competitive games and turn-based games. The system automatically categorizes the game session into high, medium, and low priorities based on the game session's interaction sensitivity and latency tolerance. High priority is suitable for shooting or racing games requiring immediate response, medium priority is suitable for strategy games, and low priority is suitable for background operations such as updates and downloads. This dynamic priority determination mechanism is entirely completed within the device kernel, requiring no user intervention, providing accurate quality of service parameters for subsequent channel establishment, and ensuring that network resources are allocated on demand.

[0122] S32, the CPE device constructs an enhanced channel establishment request based on the identification result. The request includes a successfully matched IP 5-tuple, transmission quality requirement parameters, and application priority identifier.

[0123] Specifically, after the CPE device completes the identification of game application characteristics, the system automatically integrates and analyzes the results to generate an enhanced channel request. This request is constructed in the form of a structured data packet, containing three key pieces of information: first, a network identifier composed of precisely matched source and destination addresses, port numbers, and protocol types; second, transmission quality parameters dynamically determined based on the game type, including maximum allowable latency, packet loss tolerance, and minimum bandwidth guarantee; and finally, an application priority identifier reflecting the real-time sensitivity of different games. All parameters are digitally signed by the kernel security module to prevent tampering. The request construction process is completed in the device's memory, without incurring additional disk read / write overhead, ensuring millisecond-level response times. This enhanced request mechanism enables the base station to accurately understand service requirements, providing a complete decision-making basis for subsequent resource allocation and significantly improving the service quality matching degree of the dedicated channel.

[0124] S33, the CPE device dynamically evaluates the load status and link quality of nearby mobile network base stations and selects the optimal base station that meets the transmission quality requirements.

[0125] Specifically, before establishing a channel request, the CPE device first collects real-time network status information from neighboring base stations. The device periodically acquires load index, signal strength, and interference level data for each base station via control plane signaling, while simultaneously measuring the round-trip latency and throughput of the current connection. Based on preset game transmission quality requirements, the system performs weighted calculations on these indicators to form a comprehensive quality score. The scoring model prioritizes latency-sensitive parameters to ensure the most stable link for competitive games. When multiple base stations meet the minimum quality requirements, the system selects a node with sufficient resource margin and stable signal coverage as the primary target. If the currently serving base station has the highest score, the connection is maintained to avoid handover overhead; otherwise, a list of base stations containing alternative sequences is generated. The entire evaluation process runs continuously in the background, updating the decision results every second to ensure that the channel establishment request always points to the base station with the best network conditions, providing the optimal transmission path foundation for game data streams.

[0126] S34, send a request to the selected base station to establish a dedicated channel with encrypted authentication.

[0127] Specifically, after selecting the optimal base station, the CPE device immediately securely encapsulates the enhanced channel establishment request. The system encrypts the request content using a pre-shared key and generates a digital signature to verify message integrity and source legitimacy. The encrypted request is sent through a dedicated control plane signaling channel, physically isolated from ordinary user data transmission paths, ensuring priority transmission of control commands. After transmission, the device initiates a periodic monitoring mechanism to record the request transmission time and base station identifier. If an unauthorized node is detected attempting to intercept or tamper with the request, the authentication mechanism automatically discards abnormal data packets and triggers a security alarm. The entire encryption and verification process is completed within the kernel security module, without adding any perceptible latency to the user. This end-to-end secure transmission mechanism effectively prevents man-in-the-middle attacks and resource abuse, ensuring the reliability of the dedicated channel establishment process and the accuracy of network resource allocation.

[0128] S35, if no confirmation response is received from the base station, the CPE device initiates channel establishment requests to other base stations in sequence according to the base station candidate sequence.

[0129] Specifically, after the CPE device sends a channel establishment request, it initiates a timeout monitoring mechanism. If no confirmation response is received from the base station within a preset time, the system automatically activates a sequence of backup base stations. This sequence is ranked based on previously evaluated base station quality scores, prioritizing the next best node. The device repackages the original request, adds the previous failure reason identifier, and sends it to the backup base station using the same encryption mechanism. An exponential backoff strategy is used for each retry interval to avoid network congestion. The system records the number of retries and failed base stations, dynamically updating the local base station quality database. When a connection to any base station is successfully established, subsequent attempts are immediately stopped, and that base station is marked as the current serving node. If all backup base stations fail to respond, the device degrades to a normal transmission path and records the abnormal event. The entire failover process is completed in the background within milliseconds, ensuring a seamless gaming experience for the user and providing real-time feedback data for network optimization.

[0130] To ensure optimal transmission quality and stability of game data, step S4 specifically includes the following steps:

[0131] S41, after receiving the dedicated channel confirmation information returned by the mobile network base station, the CPE device creates a stream mapping table to record the correspondence between the matched game data streams and the dedicated channels.

[0132] Specifically, after receiving the dedicated channel confirmation information returned by the mobile network base station, the CPE device initiates the process of creating a flow mapping table. This table, as the core control structure for data flow routing and forwarding, is used to accurately record the binding relationship between the identified game data flow and the dedicated channel allocated by the base station.

[0133] In practice, the main control program first parses the confirmation message returned by the base station, extracting information such as the dedicated channel identifier, allocated resource parameters, and session validity period. Subsequently, the control layer calls the flow table management module to dynamically create or update a hash-indexed flow mapping table in the device's memory. The core fields of this table include the session identifier of the game data stream, the complete IP 5-tuple characteristics, the corresponding dedicated channel identifier, the channel status identifier, and timestamp information.

[0134] To improve query efficiency, the flow mapping table adopts a key-value structure using a combined hash of the session identifier and the IP 5-tuple. When a new data packet arrives, the service identification layer can retrieve the flow mapping table in milliseconds based on its 5-tuple characteristics to determine whether it needs to be forwarded to a dedicated channel. Simultaneously, the system establishes a scheduled maintenance mechanism to periodically clean up timed-out or terminated flow table entries, ensuring that the mapping table is always synchronized with currently active sessions.

[0135] This embodiment achieves precise binding between game data streams and network resources through a stream mapping table, providing underlying support for subsequent lossless forwarding and dynamic scheduling, and effectively ensuring the reliability and efficiency of end-to-end transmission.

[0136] S42 performs lossless conversion on each data group in the matched game data stream, adding a dedicated channel identifier while maintaining the integrity of the original content.

[0137] Specifically, implementing lossless conversion and adding dedicated channel identifiers to matched game data stream packets within the CPE device mainly involves the collaborative work of the service identification layer and the data forwarding module. Its core lies in embedding a dedicated channel identifier into each data packet while maintaining the integrity of the original data content, ensuring that subsequent routing decisions accurately direct traffic to the dedicated channel.

[0138] In practice, once the traffic_acc module of the service identification layer confirms that the data flow matches the game rules, the CPE device initiates the lossless conversion process. Data packets are first passed to the encapsulation module, which uses a lightweight encapsulation protocol, such as adding a custom header structure to the outer layer of the original IP packet. This header structure contains a dedicated channel identifier, which is assigned by the control layer after the dedicated channel is successfully established and corresponds to the channel information in the flow mapping table. The encapsulation process strictly adheres to the lossless principle, meaning that the core content of the original IP packet, such as the source address, destination address, and payload data, remains unchanged; the channel identifier is carried only by adding an extra header field.

[0139] After encapsulation, the data packets are forwarded to the routing decision module. This module, based on the dedicated channel identifier in the header structure, queries the stream mapping table to determine the corresponding dedicated channel path and directs the packets to that channel. The entire processing flow is optimized at the kernel level to ensure line-rate processing and avoid introducing additional latency. This lossless conversion mechanism not only guarantees the end-to-end transmission integrity of the game data stream but also achieves precise binding with the dedicated channel, providing underlying support for a high-quality gaming experience.

[0140] S43, Dynamic priority scheduling is implemented for data groups in the data stream based on the characteristics of game interaction to ensure that highly sensitive interactive data is transmitted through a dedicated channel first.

[0141] Specifically, a concrete implementation of dynamic priority scheduling for data packets in a CPE device is primarily based on the real-time analysis of game interaction characteristics and the intelligent execution of scheduling strategies. The system first deeply analyzes the data stream content through a service identification layer to identify key data packets in game interactions, such as player operation commands, real-time location synchronization information, or skill release signals. These data packets typically exhibit small size, high frequency transmission characteristics, and are extremely sensitive to latency. The identification process relies on the analysis of data packet sequence patterns, interaction timing, and business logic. For example, by detecting changes in TCP connection state or specific identifiers in UDP packet payloads, highly sensitive interactive data is distinguished from background traffic.

[0142] Subsequently, the scheduling module of the control layer assigns the highest priority label to the identified high-sensitivity data and injects it into the priority queue of the dedicated channel. The scheduling algorithm adopts the principle of weighted fair queues to ensure that high-priority data packets are always at the front of the sending queue, while preventing low-priority traffic (such as game background music or non-real-time updated data) from blocking the channel. To achieve seamless switching, the system dynamically adjusts the queue weights and responds in real time to changes in network conditions. For example, when channel congestion is detected, the bandwidth quota for interactive data is temporarily increased. The entire scheduling process maintains millisecond-level response through kernel-level optimization, ensuring the continuity and real-time performance of the game experience.

[0143] S44 establishes a session continuity guarantee mechanism to maintain stable transmission of game sessions during network switching or connection fluctuations.

[0144] Specifically, establishing a session continuity guarantee mechanism in CPE devices mainly ensures the stability of game sessions through three core aspects: real-time monitoring, rapid switching, and state synchronization.

[0145] When game data streams are transmitted through a dedicated channel, the CPE device continuously monitors key indicators such as base station signal strength, link quality, and network latency. Once signal fluctuations or base station handover events are detected, the system immediately initiates a pre-handover process: while maintaining the existing connection, the CPE pre-establishes an auxiliary connection with the target base station through a backup link and synchronizes session state information.

[0146] During the handover execution phase, the CPE employs seamless handover technology to dynamically distribute game data streams to the old and new channels for parallel transmission, avoiding data packet loss. During the handover process, the CPE dynamically manages the transmission queue, prioritizing game command data with high real-time requirements and enabling a data packet buffering mechanism to temporarily store data packets that may arrive late during the handover.

[0147] After the handover is complete, the CPE automatically verifies session integrity, confirms no data loss by comparing sequence numbers, and quickly restores the transmission context based on cached game state information. This entire mechanism ensures that the network-level handover is transparent to the game application through real-time interaction between the control layer and the service identification layer, maintaining a seamless user experience.

[0148] S45, monitor the performance difference between the dedicated channel and the regular transmission path in real time, and dynamically adjust the routing strategy of the matched game data stream to optimize transmission quality.

[0149] Specifically, the performance monitoring and dynamic adjustment of routing strategies for dedicated channels and conventional paths in CPE devices primarily rely on the collaborative working mechanism between the control layer and the service identification layer. The system continuously compares and analyzes key performance indicators of dedicated channels and conventional transmission paths in real time, including parameters such as transmission latency, packet loss rate, and link stability. When the service identification layer detects a matched game data stream, the control layer initiates a parallel monitoring process: on the one hand, it obtains real-time quality data through the bearer status interface of the dedicated channel; on the other hand, it extracts statistical information of the conventional path from the network stack.

[0150] During monitoring, the control layer uses a sliding window algorithm to calculate the dynamic trend of performance differences. If the performance indicators of the dedicated channel consistently outperform those of the regular path, the current routing strategy is maintained. However, if latency fluctuations in the dedicated channel exceed preset tolerances or packet loss rates increase significantly, the control layer immediately triggers the routing decision module. Based on the difference analysis results, this module dynamically adjusts the routing entries in the flow mapping table, for example, diverting some game traffic to the higher-performing regular path or initiating the optimization and reconstruction process for the dedicated channel. The entire adjustment process ensures seamless data flow switching, preventing game session interruptions.

[0151] Furthermore, the system verifies the adjustment effect through a feedback mechanism: after each route change, the service identification layer reassesses the transmission quality and feeds the results back to the control layer, forming a closed-loop optimization. This dynamic routing strategy effectively copes with network fluctuations, ensuring that game data streams are always transmitted through the optimal path, ultimately achieving continuous improvement in transmission quality.

[0152] To ensure game continuity while optimizing network load balancing and avoiding resource waste, step S5 specifically includes the following steps:

[0153] The S51 CPE device performs real-time quality assessment of game data streams based on multi-dimensional transmission metrics, including latency jitter, packet loss patterns, and throughput stability.

[0154] Specifically, after the dedicated channel is activated, the CPE device continuously collects the transmission characteristics of the game data stream. The system calculates the arrival time interval of consecutive data packets in real time, analyzes the latency fluctuation amplitude, statistically analyzes packet loss distribution patterns, distinguishes between random packet loss and burst packet loss types, and monitors the effective data throughput change rate per unit time. These indicators are dynamically weighted and evaluated using a sliding window algorithm to form a comprehensive quality score. When increased latency jitter or continuous packet loss is detected, the system automatically increases the monitoring frequency and judges the network degradation trend by combining traffic interaction characteristics. The evaluation results are recorded in the kernel quality log in real time, providing data support for channel reconstruction decisions. The entire monitoring process is completed at the device level without adding additional transmission overhead, ensuring both evaluation accuracy and real-time performance.

[0155] S52: When a continuous degradation in transmission quality is detected, the CPE device initiates a predictive channel reconstruction mechanism. After the new channel is established, traffic switching is performed to ensure the continuity of the gaming experience.

[0156] Specifically, when the CPE device detects a continuous decline in the transmission quality of the game data stream, the system immediately initiates a predictive reconstruction process. This mechanism silently initializes the new channel establishment procedure in the background while maintaining normal transmission on the existing channel. The device reuses the base station evaluation logic from step S3, selects the optimal candidate node to initiate a channel request, and processes control signaling and user data streams in parallel. Before the new channel is fully established and passes quality verification, all game data continues to be transmitted through the original channel. After confirming that the new channel is ready, the system performs a millisecond-level switch between data packet sequences, directing subsequent traffic to the new path. The switching process employs a state synchronization mechanism to ensure the continuity of TCP sequence numbers and the integrity of the session context. The entire reconstruction process is completed without the user's awareness, avoiding game stuttering or reconnection issues caused by channel interruptions in traditional solutions, effectively ensuring the continuity of the real-time interactive experience.

[0157] S53, the CPE device distinguishes between temporary sleep state and actual session termination by analyzing the interaction characteristics of data streams and traffic sleep modes.

[0158] Specifically, the CPE device continuously monitors the interaction patterns and traffic characteristics of the game data stream, identifying session status by analyzing the temporal distribution patterns of data packets. The system pays particular attention to the interaction frequency of uplink control commands and downlink response data. When a sudden drop in data stream rate is detected, it further analyzes the packet interval distribution characteristics and directionality ratio. If periodic heartbeat packets are maintained or there is a short period of no data but the connection identifier is not released, it is determined to be a temporary sleep state, such as game pause or scene loading. When a complete connection termination sequence is observed, including a FIN handshake packet or an RST abnormal termination flag, combined with the bidirectional silence duration of the data stream and application-layer session markers, it is confirmed that the session has truly ended. The device also incorporates user operation context, such as the correlation between keyboard and mouse activity signals and network traffic, to enhance the accuracy of the judgment. This multi-dimensional analysis mechanism runs in real time at the kernel level, ensuring accurate resource release decisions and avoiding session interruptions or resource waste caused by misjudgments.

[0159] S54, upon confirming the complete termination of the game session, the CPE device sends a hierarchical resource release command to the base station to achieve orderly reclamation of dedicated transmission resources and network load balancing.

[0160] Specifically, once the CPE device confirms the complete termination of the game session, the system does not immediately release all dedicated resources. Instead, it initiates a tiered release process. First, it lowers the transmission bandwidth priority, switching the channel from high-security mode to basic service level, and continuously monitors for several seconds to confirm no data recovery. Then, it sends a first-level release command to reclaim excess bandwidth quotas while preserving the basic connection context. If the system remains silent, a second-level command is sent to release the QoS scheduling policy, and finally, routing table entries and session state memory are released. The interval between each release level is dynamically adjusted based on the current network load, extending the interval during busy periods to avoid sudden load surges and accelerating reclamation during idle periods. After receiving the tiered commands, the base station gradually reintegrates the resources into the public pool, smoothly transitioning them to other user services. This gradual release mechanism avoids drastic fluctuations in network resource utilization, reduces base station rescheduling overhead, and reserves warm-up conditions for possible rapid session recovery, significantly improving overall network resource utilization efficiency and stability.

[0161] To ensure a stable and smooth gaming experience, step S5 also includes a dynamic service quality adjustment step, specifically:

[0162] The S55 CPE device dynamically adjusts the service quality parameters of the dedicated channel based on real-time transmission quality indicators, including bandwidth allocation priority and traffic scheduling strategy.

[0163] Specifically, the dynamic adjustment of service quality parameters in CPE equipment mainly relies on the collaborative working mechanism between the control layer and the service identification layer. Its core lies in building a closed-loop monitoring and adjustment system to continuously optimize the transmission performance of the dedicated channel.

[0164] The control layer of the CPE device collects transmission quality indicators of the dedicated channel in real time through the service identification layer. These indicators include key parameters such as end-to-end latency of data packets, packet loss rate trends, and channel throughput stability. The collection process is performed periodically at fixed time intervals to ensure accurate capture of dynamic fluctuations in network status.

[0165] Upon obtaining real-time quality metrics, the strategy analysis module of the control layer immediately initiates the evaluation process. This module compares the current metrics with preset quality thresholds and simultaneously identifies performance trends by incorporating historical transmission data. When the analysis results indicate signs of transmission quality degradation, the strategy analysis module generates corresponding parameter adjustment plans.

[0166] According to the adjustment plan, the control layer initiates a service quality parameter renegotiation request to the base station through a dedicated configuration interface. The renegotiation mainly involves adjusting the bandwidth allocation ratio and redistributing traffic scheduling weights. For game interaction data with high real-time requirements, the system will appropriately increase its bandwidth guarantee ratio and forwarding priority; while for background traffic with higher tolerance, its resource consumption weight will be reduced accordingly.

[0167] The entire parameter adjustment process employs a gradual implementation mechanism to avoid transmission jitter caused by sudden parameter changes. After adjustment, the system continues to monitor changes in quality indicators, forming a closed-loop control for continuous optimization. This dynamic adjustment mechanism ensures that the dedicated channel can adapt to changes in the network environment and always maintain optimal transmission performance.

[0168] S56: When transmission quality is abnormal, service quality parameters are renegotiation is triggered, and the transmission stability of the dedicated channel is optimized through base station collaboration.

[0169] Specifically, implementing QoS parameter renegotiation in CPE devices mainly revolves around three core aspects: transmission quality anomaly detection, base station collaborative interaction, and dynamic adjustment of channel parameters. Its design goal is to ensure that the dedicated channel can adapt to network fluctuations and maintain high-quality transmission of game data streams.

[0170] When the transmission quality monitoring unit of the CPE device detects anomalies in the transmission indicators of the dedicated channel through continuous sampling, the system immediately initiates a service quality parameter renegotiation process. Anomalies typically manifest as changes in indicators affecting the gaming experience, such as continuously increasing latency, a significant rise in packet loss rate, or drastic fluctuations in throughput. The monitoring unit reports the anomaly status and related quality analysis results to the main control program in the control layer.

[0171] Upon receiving an anomaly report, the main control program first performs a comprehensive assessment of the current channel's operational status to determine whether renegotiation needs to be triggered. Assessment factors include the duration of the anomaly, the deterioration trend, and the actual impact on the gaming service. If renegotiation is deemed necessary, the main control program generates a service quality parameter renegotiation request. This request explicitly includes the parameters to be adjusted, such as the expected guaranteed bit rate, the maximum tolerable latency limit, or the updated traffic scheduling priority.

[0172] Subsequently, the main control program encapsulates the formatted renegotiation request into standard AT command signaling via the device's built-in cellular communication module and sends it to the currently serving mobile network base station. This request is initiated through specific QoS-related instructions, such as extended AT commands used to modify dedicated bearer QoS parameters, and aims to request the base station to update the quality of service attributes of the existing dedicated channel.

[0173] Upon receiving a renegotiation request, the base station's radio resource management module assesses the feasibility of the CPE's request based on the overall network load, radio interface quality, and available resources. If conditions permit, the base station allocates new resources or adjusts existing resource configurations, and sends a renegotiation confirmation message to the CPE via downlink signaling. This message includes the base station's agreement to the adjusted set of quality of service parameters.

[0174] After successfully receiving and parsing the acknowledgment message from the base station, the CPE device's control layer immediately updates its local flow mapping table and related channel status information, making the new quality of service parameters effective. The service forwarding layer then adjusts its data packet scheduling strategy to adapt the game data stream to the optimized channel parameters. Simultaneously, the monitoring unit continues to track the transmission quality under the new parameters, forming a closed-loop control process of "monitoring-evaluation-negotiation-optimization-re-monitoring," thereby effectively improving the transmission stability of the dedicated channel through the coordinated operation of the base station.

[0175] To address key issues in traditional solutions such as low game recognition accuracy, static resource allocation, and session interruptions caused by mobile network switching, and to significantly reduce game latency and jitter while improving the real-time interactive experience in mobile network environments without altering the existing network architecture, please refer to [reference needed]. Figure 2 The present invention also provides a dynamic game acceleration system based on CPE, which is configured in a CPE device and includes:

[0176] The hook function injection unit, configured in the CPE device, is used to intercept data streams transmitted between the user terminal and the external network through the CPE device's Modem, WiFi, or Ethernet network interface.

[0177] Specifically, a specific embodiment of the hook function injection unit involves deploying software hooks in the CPE device driver layer to achieve real-time interception and processing of network data streams. This unit monitors data transmission paths for Modem, WiFi, and Ethernet network interfaces by embedding customized hook functions into the CPE device's network driver stack.

[0178] During implementation, the traffic_acc service identification driver module is first loaded during the CPE device startup phase. During initialization, this driver module registers interface monitoring hook functions with the hw_nat hardware acceleration driver module. The registration process involves binding specific hook function instances to the transmit and receive directions of each network interface, ensuring that data flows through Modem, WiFi, or Ethernet interfaces can be captured in real time.

[0179] When data packets are transmitted through the aforementioned network interface, the hardware acceleration driver invokes the injected hook function. The hook function first intercepts the raw data stream and then passes it to the business identification layer for subsequent parsing and matching. The entire interception process is transparent to upper-layer applications and is optimized at the kernel level to minimize processing latency, providing fundamental data support for game acceleration.

[0180] This embodiment achieves seamless interception of multi-interface data streams through a hardware-software co-design, laying the underlying technical foundation for dynamic game acceleration.

[0181] The data stream filtering unit is used to filter the data stream according to pre-configured matching rules. The matching rules are based on one or more parameters of interface type, data flow direction, and IP 5-tuple, and support subnet-level matching through mask.

[0182] Specifically, the data flow filtering unit involves the coordinated operation of three core components: rule configuration, matching execution, and subnet adaptation. This unit implements its functionality through a service identification-driven module. The main control program first receives matching rule configuration instructions from the user. These rules include parameters such as interface type, data flow direction, and IP 5-tuple. The configuration information is stored in a hash table at the kernel level, where the IP address range supports subnet mask parameter settings to adapt to dynamic changes in the game server's IP pool.

[0183] After the matching process begins, the hardware acceleration driver calls the registered hook functions when data flows through the Modem, WiFi, or Ethernet interface. The service identification driver then intercepts the data packets, performs initial filtering based on interface type and data flow direction, and extracts IP 5-tuple features through protocol parsing hooks. The extracted features are compared with rules in the hash table. When the destination IP address matches the result of a bitwise AND operation with the IP range configured in the rule using a mask, a subnet-level match is determined to be successful.

[0184] For successfully matched data streams, the unit marks them as game traffic and triggers subsequent processing. The entire filtering process optimizes efficiency through a hierarchical matching mechanism, successively reducing the processing overhead of irrelevant traffic, and ultimately achieving accurate identification of game services.

[0185] The dedicated channel establishment unit is used to initiate a dedicated channel establishment request to the mobile network base station based on the identification information of the game data stream when a game data stream that matches the matching rules is detected, carrying the IP five-tuple parameters of the successfully matched stream.

[0186] Specifically, the dedicated channel establishment unit involves the collaborative interaction mechanism between the CPE device and the mobile network base station after recognizing the game data stream. As a core component of the system, this unit is responsible for automatically triggering the dedicated channel establishment process when the service identification layer confirms that the data stream matches the game rules.

[0187] When the traffic_acc driver module of the service identification layer detects a game data stream that matches the matching rules, the dedicated channel establishment unit is immediately activated. The unit first extracts complete IP 5-tuple identification information from the data stream, including the source IP address, destination IP address, source port number, destination port number, and transport protocol type. Subsequently, the unit constructs a dedicated channel establishment request based on these parameters.

[0188] The unit sends a standardized AT command sequence to the mobile network base station through the cellular communication module of the CPE device. Specifically, this includes defining dedicated bearer parameters using the CGDCONT instruction, activating the bearer using the CGACT instruction, and setting quality of service requirements using the C5GQOS instruction. A key step is submitting a traffic template to the base station via the CGTFT instruction, explicitly carrying the IP 5-tuple parameters of the game data stream, to request the base station to allocate dedicated channel resources.

[0189] To optimize connection success rate, the unit dynamically assesses the load status and link quality of nearby base stations, prioritizing base stations with stable signals as the target for requests. If the primary base station does not respond in time, the unit automatically switches to the backup base station sequence to retransmit the request. The entire process ensures that the dedicated channel is established within milliseconds, providing low-latency transmission guarantees for game data streams.

[0190] The data flow routing unit is used to route the matching game data flow to the dedicated channel after the mobile network base station confirms that the dedicated channel has been successfully established.

[0191] Specifically, the data flow routing unit involves the efficient redirection of game data flows by the CPE device after a dedicated channel is successfully established. When the mobile network base station returns dedicated channel confirmation information, this unit immediately initiates the route redirection process. The unit first parses the base station confirmation message, extracts the dedicated channel identifier and related resource parameters, updates the internally maintained flow mapping table, and establishes the binding relationship between the IP 5-tuple characteristics of the game data flow and the dedicated channel.

[0192] Subsequently, for each data packet flowing through the CPE device, the data flow routing unit retrieves the flow mapping table in real time through the service identification layer. If the IP 5-tuple of the data packet matches an entry in the table, the unit dynamically modifies its routing policy, forcing the data packet to be transmitted through a dedicated channel. This process is implemented through routing decision hooks at the kernel layer, ensuring seamless switching of the data flow to a dedicated path while maintaining the integrity of the original content. The entire routing redirection mechanism is based on a hardware-software co-engineering architecture, achieving low-latency, highly reliable data transmission and ensuring the continuity of the gaming experience.

[0193] The transmission quality monitoring unit is used to continuously monitor the transmission quality of the game data stream. When an abnormal transmission quality is detected, it triggers the reconstruction of the dedicated channel. When the game data stream is detected to be over, it sends a channel release command to the base station to reclaim the dedicated transmission resources.

[0194] Specifically, the transmission quality monitoring unit involves a real-time assessment and intelligent response mechanism for the transmission status of game data streams within the CPE device. This unit periodically collects key quality indicators through a built-in sampling module, including the fluctuation trend of transmission latency, the regularity of data packet loss, and the stability changes in throughput. The collection process analyzes the dynamics of the indicators based on a sliding window algorithm. When the quality degradation exceeds a preset threshold within multiple consecutive sampling periods, the unit immediately determines it as a transmission anomaly.

[0195] Upon triggering an anomaly, the unit initiates a predictive dedicated channel reconstruction process, prioritizing negotiation with the base station to establish a new channel. Once the new channel is ready, game traffic is seamlessly switched over, ensuring a completely unnoticed user experience. Simultaneously, the unit accurately distinguishes between temporary hibernation and permanent session termination by deeply analyzing the interaction characteristics and traffic quiescent modes of the data stream. When it is confirmed that the game data stream has completely ended, the unit generates a tiered resource release command, notifying the base station via standardized signaling to reclaim dedicated transmission resources, achieving balanced network load optimization. This entire monitoring and response mechanism forms a closed-loop control, effectively ensuring the continuity and reliability of the game acceleration service.

[0196] The dynamic rule update unit is used to keep the game server address database synchronized in real time with the business identification module through a kernel-level communication mechanism, obtain address change information of the game server cluster, and automatically adjust the matching rules based on the address change information. By updating the IP address range and its mask parameters, the data flow identification mechanism and the game server deployment strategy are kept dynamically synchronized.

[0197] Specifically, the dynamic rule update unit involves a closed-loop synchronization mechanism between the game server address database and data flow identification rules. This unit establishes a kernel-level communication channel and a bidirectional data flow with the business identification module to achieve real-time synchronization of game server cluster address information and dynamic adjustment of rules.

[0198] During the initialization phase, the unit establishes a communication link with the business identification module through a shared memory area and loads the complete configuration of the current game server address database. The address database adopts a hierarchical storage structure, containing parameters for the main server cluster, backup nodes, and load balancing strategies. The business identification module uses an asynchronous event notification mechanism to convert address change operations into standardized protocol messages and send them to the dynamic rule update unit.

[0199] During real-time monitoring, the unit employs an event-driven model to handle address change information. When a server address is added or removed, or the mask parameter is adjusted, the parsing module automatically extracts the IP address range, subnet mask, and weight parameters from the changed entry. After validating the address format using a regular expression engine, a new rule template is generated and injected into the in-memory database. This process utilizes a read-write lock mechanism to ensure the safety of concurrent operations and prevent data flow interruptions during rule updates.

[0200] During the rule update phase, the unit reconstructs the matching rule set based on the changed address database. An incremental update strategy is adopted, replacing rules only for affected IP address ranges while retaining unchanged parts to reduce processing overhead. New rules are written to persistent storage via atomic operations, simultaneously triggering the rule version verification process of the business identification module. After successful verification, the updated rule set is quickly loaded into the data stream processing pipeline via shared memory.

[0201] To ensure the reliability of rule synchronization, the unit integrates a two-way confirmation mechanism. After applying a new rule, the business identification module returns a confirmation response, which the dynamic rule update unit uses to update its local state machine. If three consecutive confirmations fail, a rollback operation is automatically triggered, restoring the most recent valid rule version. The entire synchronization process is controlled by timestamps and version numbers to ensure the temporal consistency and traceability of rule changes.

[0202] The exception handling module continuously monitors the communication link status. When a kernel-level communication interruption is detected, it automatically switches to a backup channel and activates local caching rules to maintain basic functionality. Simultaneously, a heartbeat detection mechanism identifies the liveness of the business identification module, triggering a degradation strategy in case of anomalies to ensure that core data flow identification functions are unaffected. This mechanism, through a layered fault-tolerant design, achieves strong consistency and high availability for dynamic rule updates.

[0203] To form a complete data stream processing chain and achieve accurate identification and intelligent routing of game data streams, the hook function injection unit includes:

[0204] The interface monitoring module is used to capture raw data streams.

[0205] Specifically, the interface monitoring module is automatically loaded into the kernel network stack when the CPE device starts up, deployed on the data reception and transmission paths of the three major network interfaces: Modem, WiFi, and Ethernet. This module, by registering kernel callback functions, transparently captures all data packets crossing the CPE device boundary without interrupting normal communication. When the user terminal interacts with the external network, the module intercepts raw data frames in real time, retaining complete header and payload information, while recording interface type and transmission direction metadata. The captured data stream is directly transmitted to the protocol parsing module via a zero-copy memory mechanism, avoiding latency caused by redundant buffer copying. The entire monitoring process adopts a lightweight design, maintaining line-speed processing capabilities even under high traffic loads, with no perceptible impact on the user's network experience, providing a complete raw data foundation for subsequent game traffic identification.

[0206] The protocol parsing module is used to extract the IP 5-tuple characteristics of the data stream and identify the TCP connection state.

[0207] Specifically, the protocol parsing module is activated immediately after the interface monitoring module captures the data stream and is deployed in the deep processing layer of the CPE device's kernel network protocol stack. This module parses the network and transport layer headers of each data packet, accurately extracting the complete five-tuple information consisting of the source address, destination address, source port, destination port, and protocol type. Simultaneously, the module analyzes TCP flag changes in real time, identifying connection establishment, data transmission, and session termination states by monitoring the SYN, ACK, and FIN control bit sequences. The parsing process employs a flow state tracking mechanism, associating discrete data packets with the same session context to ensure the continuity of feature extraction. The processing results are output in a structured data format, including an encrypted checksum to ensure integrity, and are efficiently transmitted to the business identification module via a kernel-level message queue. The entire parsing process uses memory pool optimization technology to avoid frequent allocation and release operations, maintaining microsecond-level processing latency even under high network load scenarios, laying the foundation for accurate identification of subsequent game traffic.

[0208] The service identification module is used to match the extracted features with preset rules in real time and identify connection establishment and release messages.

[0209] Specifically, the service identification module runs in the CPE device kernel space and receives structured feature data transmitted by the protocol parsing module. This module maintains a dynamically loaded game service rule base, including server address ranges, feature port ranges, and interaction mode templates. The matching process employs a two-level filtering mechanism: first, a coarse screening is performed using address and port ranges, followed by a finer judgment based on packet timing characteristics. The module pays particular attention to TCP handshakes and termination sequences; when a SYN packet is detected, it is marked as a new connection establishment; when a complete FIN four-way handshake sequence is observed, it is determined as a normal release; and when an RST abnormal termination flag is encountered, the session abnormal state is recorded. Successfully matched game data streams are assigned a unique service identifier and accompanied by a connection lifecycle marker. The identification results are communicated to the routing decision module in real time via a kernel event mechanism, while simultaneously updating the flow state table. The entire processing flow uses memory mapping technology to accelerate rule queries, avoid disk I / O latency, and ensure microsecond-level response speeds under high-concurrency scenarios without affecting normal network transmission performance.

[0210] The routing decision module is used to dynamically adjust the transmission path of the matching data stream based on the channel establishment results.

[0211] Specifically, the routing decision module is deployed in the CPE device's kernel forwarding engine, receiving real-time feedback on the establishment status of the dedicated channel. When the base station returns channel confirmation information, the module creates a high-priority flow table entry in the routing table, binding the five-tuple matching the game data flow with the dedicated channel identifier. When a data packet arrives, the module determines the forwarding path through a fast hash lookup, adds a dedicated marker header to the game traffic, and directs it to the optimized channel. During path switching, a gradual migration strategy is adopted, forwarding highly sensitive data such as control commands first, then transferring ordinary game data to ensure a continuous user experience. When a channel quality degradation is detected, the module automatically activates a pre-established backup path, performing a seamless millisecond-level switchover once the new channel is ready. If channel reconstruction fails, it immediately reverts to the default route and marks the abnormal state. All decisions are based on real-time updates to the connection state table, working in conjunction with the quality monitoring unit through a kernel-level notification mechanism to ensure orderly data packet transmission and uninterrupted sessions. The entire process is completed in microseconds and is completely transparent to the user.

[0212] The interface monitoring module, protocol parsing module, service identification module, and routing decision module work together in sequence according to the data processing flow to form a complete data flow processing chain.

[0213] Specifically, upon CPE device startup, four processing modules are sequentially loaded and registered to the kernel network stack, forming a pipelined data processing architecture. When a user initiates network communication, the interface monitoring module first captures the raw data stream at the physical interface layer and directly transmits it to the protocol parsing module without any modification. The protocol parsing module deeply disassembles the data packet structure, extracts complete network characteristics and connection status information, and sends the structured data to the service identification module. The service identification module compares real-time characteristics with the rule base, confirms the game traffic attributes and marks the connection lifecycle, and forwards the decision result to the routing decision module. The routing decision module dynamically updates the forwarding strategy based on the channel status and prioritizes the transmission of matching data streams. The modules communicate with each other through a kernel-level zero-copy message queue, using an event-triggered mechanism to achieve seamless connection and avoid latency caused by repeated buffer copying. The entire processing chain executes continuously in kernel mode, and the entire process from data packet capture to routing decision is completed in microseconds, with no perceptible impact on the user's network experience. This tightly collaborative architecture enables the CPE device to perceive service needs in real time and autonomously optimize the transmission path, transforming the traditional passive forwarding device into a network edge node with intelligent decision-making capabilities.

[0214] The following is another specific embodiment of the dynamic game acceleration method and system based on CPE. This embodiment only matches IP flows of interest, and matching can be performed based on one or more parameters, such as interface Modem / WiFi / Ethernet, IP data flow direction RX / TX, and IP 5-tuple (source IP / destination IP / source port / destination port / protocol type). When matching source IP / destination IP is required, subnet matching can also be performed based on the subnet mask, satisfying the existing situation where a domain name corresponds to a set of subnet IPs, achieving flexibility and efficiency to meet the needs of different business scenarios.

[0215] 1. System architecture, including:

[0216] (1) User Layer: Located on the CPE device, the user layer provides a user interface. Users can establish a connection with the control layer via WEB or APP. The user layer and the control layer configure acceleration functions through custom message interaction, including: whether to enable the acceleration function; the five-tuple of the IP channel to be accelerated (source IP / destination IP / source port / destination port / protocol type), interface name, data flow direction, and subnet mask. The subnet mask consists of the actual subnet mask plus several bits of the host address, used to match multiple IPs in a flow, ensuring that the matched game service IP is allocated in the IP pool, improving the flexibility of matching; the successfully matched IP flow; the user layer also provides monitoring functions, that is, to perform statistical analysis on the latency and sent and received data packets of the successfully matched IP flow.

[0217] (2) Control layer: Implement the main control function of game acceleration, including sending the whitelist IP five-tuple of WEB users to the kernel driver module, establishing or releasing the IP link matched by the kernel module, and establishing or releasing a dedicated data channel between CPE, base station and core network through AT commands.

[0218] (3) Service identification layer: Based on the matching rules of the IP 5-tuple configured by the user, the data stream of Modem / WiFi / Ethernet driver RX / TX is filtered, the data packets are parsed, the matching is performed according to the rules, and the information of the successfully matched IP stream - the IP 5-tuple is sent to the control layer application.

[0219] 2. Core methodology and process, including:

[0220] Step 1, configuration and initialization, including: (1) Loading the game acceleration driver (traffic_acc, hereinafter the same), traffic_acc installs Modem / WiFi / Ethernet device driver hook functions for the hardware acceleration driver (hw_nat, hereinafter the same); (2) WEB / APP sends IP flow filtering parameters through the main control program, such as the target IP from the Modem device being xxx.xxx.xx.xxx, TCP data flow. traffic_acc saves the matching parameters to the hash table and enables the Modem TX direction filtering function of the hw_nat device.

[0221] Step 2, IP data matching, includes: (1) The user initiates a game connection to xxx.xxx.xx.xxx through CPE; (2) The hw_nat driver detects Modem TX filtering enabled and calls the hook function set by traffic_acc; (3) traffic_acc parses the IP data packet and determines whether it is the first packet established by TCP or a packet released by TCP connection. If so, it searches the hash table to determine whether the IP data packet is the IP that needs to be accelerated; (4) If the matching fails, it returns and the IP data continues to be processed according to the original process. If the matching is successful, the IP 5-tuple and matching information are reported to the main control program, and then it returns and the IP data continues to be processed according to the original process.

[0222] Step 3, bearer establishment, includes: (1) defining a dedicated bearer using the AT+CGDCONT command; (2) activating the dedicated bearer using the AT+CGACT command; (3) setting QoS using the AT+C5GQOS command; (4) requesting a dedicated bearer from the base station using the AT+CGTFT command, carrying the IP 5-tuple; (5) after the base station allocates a dedicated channel, the CPE forces the game traffic to be routed to that channel.

[0223] Step 4, dynamic maintenance, including: (1) sampling latency data every 5 seconds, triggering bearer reconstruction if the threshold is exceeded (e.g., 30ms); (2) sending AT+CGACT=0 to release resources when the game ends.

[0224] For those skilled in the art, various other corresponding changes and modifications can be made based on the technical solutions and concepts described above, and all such changes and modifications should fall within the protection scope of the claims of this invention.

[0225] It should be noted that the embodiments described above are only some embodiments of the present invention, and not all embodiments. The singular forms "a," "described," and "itself" used in the embodiments and the appended claims are also intended to include the plural forms, unless the context clearly indicates otherwise.

Claims

1. A method for dynamic game acceleration based on CPE, characterized in that, The method includes the following steps: S1, inject a hook function into the CPE device. The hook function includes an interface monitoring hook, which is used to intercept the data stream transmitted between the user terminal and the external network through the network interface of the CPE device. S2, the data stream is filtered according to the matching rules pre-configured on the CPE device. The matching rules can be based on one or more parameters of interface type, data flow direction, IP 5-tuple, and support subnet-level matching through mask; S3, when a game data stream that matches the matching rules is detected, the CPE device initiates a dedicated channel establishment request to the mobile network base station based on the identification information of the data stream, carrying the IP five-tuple parameters of the successful match; S4. After the mobile network base station confirms that the dedicated channel has been successfully established, the CPE device will route the matching game data stream to the dedicated channel. S5, the CPE device continuously monitors the transmission quality of the game data stream. When an abnormal transmission quality is detected, it triggers the reconstruction of the dedicated channel. When the game data stream is detected to be ending, it sends a channel release command to the base station to reclaim the dedicated transmission resources. The hook function also includes: Protocol parsing hooks are used to extract IP 5-tuple features from data streams and identify TCP connection states; The business identification hook is used to match the extracted features with preset rules in real time and identify connection establishment and release messages; Routing decision hooks are used to dynamically adjust the transmission path of matching data streams based on the channel establishment results; The hook functions work together in sequence according to the data processing flow: after the interface monitoring hook captures the raw data stream, it is passed to the protocol parsing hook. After the protocol parsing hook extracts the features, it is handed over to the business identification hook for rule matching. After the business identification hook confirms the matching result, it triggers the routing decision hook to perform path adjustment, forming a complete data stream processing chain. The matching rules support dynamic updates, and the method further includes the following steps: S6, the CPE device works with the service identification hook through a kernel-level communication mechanism to maintain real-time synchronization of the game server address database and obtain address change information of the game server cluster; S7, the CPE device automatically updates the matching rules based on the address change information, and keeps the data flow identification mechanism synchronized with the game server deployment strategy by dynamically adjusting the IP address range and its mask parameters.

2. The CPE-based dynamic game acceleration method according to claim 1, wherein, Step S2 specifically includes the following steps: S21, A hierarchical matching mechanism is pre-configured in the CPE device. The hierarchical matching mechanism matches the data flow based on the interface type, data flow direction and IP five-tuple in sequence, and supports subnet-level matching through the mask. S22, perform game feature identification on the data stream that has been matched hierarchically to confirm that it is a game data stream; S23 records the IP 5-tuple information of the confirmed game data stream for use in subsequent dedicated channel establishment requests.

3. The CPE-based dynamic game acceleration method of claim 1, wherein, Step S3 specifically includes the following steps: S31, When a game data stream that matches the matching rules is detected, the CPE device identifies the application characteristics of the game data stream and determines the corresponding transmission priority; S32, the CPE device constructs an enhanced channel establishment request based on the identification result, the request including the successfully matched IP 5-tuple, transmission quality requirement parameters and application priority identifier; S33, the CPE device dynamically evaluates the load status and link quality of nearby mobile network base stations and selects the optimal base station that meets the transmission quality requirements; S34, Send a request to the selected base station to establish a dedicated channel with encrypted verification; S35, if no confirmation response is received from the base station, the CPE device initiates channel establishment requests to other base stations in sequence according to the base station candidate sequence.

4. The CPE-based dynamic game acceleration method of claim 1, wherein, Step S4 specifically includes the following steps: S41, after receiving the dedicated channel confirmation information returned by the mobile network base station, the CPE device creates a stream mapping table to record the correspondence between the matched game data streams and the dedicated channels; S42, perform lossless conversion on each data group in the matched game data stream, and add a dedicated channel identifier while maintaining the integrity of the original content; S43, Dynamic priority scheduling is implemented for data groups in the data stream based on the characteristics of game interaction to ensure that highly sensitive interactive data is transmitted through a dedicated channel first; S44 establishes a session continuity guarantee mechanism to maintain stable transmission of game sessions during network switching or connection fluctuations. S45, monitor the performance difference between the dedicated channel and the regular transmission path in real time, and dynamically adjust the routing strategy of the matched game data stream to optimize transmission quality.

5. The dynamic game acceleration method based on CPE according to claim 1, characterized in that, Step S5 specifically includes the following steps: S51, the CPE device performs real-time quality assessment of the game data stream based on multi-dimensional transmission indicators, including latency jitter, packet loss mode and throughput stability. S52, when a continuous degradation in transmission quality is detected, the CPE device starts a predictive channel reconstruction mechanism, and then performs traffic switching after the new channel is established to ensure the continuity of the gaming experience; S53, CPE devices distinguish between temporary sleep states and actual session termination by analyzing the interaction characteristics of data streams and traffic sleep modes; S54, upon confirming the complete termination of the game session, the CPE device sends a hierarchical resource release command to the base station to achieve orderly reclamation of dedicated transmission resources and network load balancing.

6. The dynamic game acceleration method based on CPE according to claim 5, characterized in that, Step S5 further includes a dynamic service quality adjustment step, specifically: S55, the CPE device dynamically adjusts the service quality parameters of the dedicated channel based on real-time transmission quality indicators, including bandwidth allocation priority and traffic scheduling strategy; S56: When transmission quality is abnormal, service quality parameters are renegotiation is triggered, and the transmission stability of the dedicated channel is optimized through base station collaboration.

7. A dynamic game acceleration system based on CPE, characterized in that, The system is configured in a CPE device, and the system includes: The hook function injection unit, configured in the CPE device, is used to intercept the data stream transmitted between the user terminal and the external network through the CPE device's Modem, WiFi, or Ethernet network interface; A data stream filtering unit is used to filter the data stream according to pre-configured matching rules. The matching rules are based on one or more parameters of interface type, data flow direction, and IP 5-tuple, and support subnet-level matching through mask. The dedicated channel establishment unit is used to initiate a dedicated channel establishment request to the mobile network base station based on the identification information of the game data stream when a game data stream that matches the matching rules is detected, carrying the IP five-tuple parameters of the successfully matched stream. The data stream routing unit is used to route the matching game data stream to the dedicated channel after the mobile network base station confirms that the dedicated channel has been successfully established; The transmission quality monitoring unit is used to continuously monitor the transmission quality of the game data stream. When an abnormal transmission quality is detected, it triggers the reconstruction of the dedicated channel. When the game data stream is detected to be over, it sends a channel release command to the base station to reclaim the dedicated transmission resources. The dynamic rule update unit is used to maintain real-time synchronization of the game server address database with the business identification module through a kernel-level communication mechanism, obtain address change information of the game server cluster, and automatically adjust the matching rules based on the address change information. By updating the IP address range and its mask parameters, the data flow identification mechanism and the game server deployment strategy are kept dynamically synchronized. The hook function injection unit includes: The interface monitoring module is used to capture the raw data stream; The protocol parsing module is used to extract the IP 5-tuple characteristics of the data stream and identify the TCP connection state; The service identification module is used to match the extracted features with preset rules in real time and identify connection establishment and release messages; The routing decision module is used to dynamically adjust the transmission path of the matching data stream based on the channel establishment results; The interface monitoring module, protocol parsing module, service identification module, and routing decision module work together in sequence according to the data processing flow to form a complete data flow processing chain. The matching rules support dynamic updates, and the system also includes: The communication synchronization and change detection unit is used by the CPE device to maintain real-time synchronization of the game server address database through kernel-level communication mechanisms and in cooperation with the service identification hook, and to obtain address change information of the game server cluster. The rule adjustment unit is used by the CPE device to automatically update the matching rules based on the address change information, and to keep the data flow identification mechanism synchronized with the game server deployment strategy by dynamically adjusting the IP address range and its mask parameters.

Images