Multi-link collaborative wireless communication resource scheduling method and system

Abstract

The invention relates to the technical field of wireless communication, and discloses a multi-link collaborative wireless communication resource scheduling method and system. The invention aims to solve the problems of low service matching degree and insufficient resource utilization rate caused by single scheduling strategy in the prior art. According to the method, the demand type and the data scale characteristic of the data stream are identified by acquiring the multi-link physical parameters; differentiated weighted scoring is carried out on each link based on the demand type, and a link adaptive value is generated by combining real-time load correction; formulating a transmission strategy including preferential mapping, load sharing or a concurrent redundancy mode according to the characteristics; and triggering dynamic rescheduling based on network state monitoring. According to the invention, refined resource scheduling based on service awareness is realized, and the transmission efficiency and stability of a heterogeneous network are effectively improved.

Description

Technical Field

[0001] This invention relates to the field of wireless communication technology, and more specifically, to a method and system for scheduling wireless communication resources in a multi-link collaborative manner. Background Technology

[0002] With the rapid development of mobile communication technology, smart terminal devices (such as smartphones and tablets) generally possess multi-mode communication capabilities, simultaneously supporting multiple wireless access methods such as cellular mobile communication networks (4G / 5G) and wireless local area networks. This hardware architecture, which allows for the coexistence of heterogeneous networks, aims to provide users with a wider coverage and higher bandwidth communication experience by leveraging the complementary advantages of different network standards. The rational management of multiple wireless communication link resources to ensure the continuity and efficiency of data transmission is a fundamental issue in the field of wireless communication.

[0003] Currently, most mainstream scheduling solutions in the industry adopt a single link switching mechanism based on physical signal strength. This means that the system prioritizes connecting to the wireless LAN by default, only triggering a hard switch to the cellular network when the wireless LAN signal strength drops below a specific threshold or the connection is completely interrupted. However, this coarse-grained scheduling strategy, which relies heavily on physical signal strength, is inadequate when facing increasingly complex and diverse upper-layer application requirements. Different types of data streams (such as real-time battle games, high-definition video downloads, and background message pushes) have vastly different sensitivities to network performance indicators (such as low latency, high bandwidth, and low power consumption). Furthermore, channel interference and congestion in heterogeneous network environments are often not entirely positively correlated with signal strength (e.g., a full wireless LAN signal but extremely congested channel). The existing single scheduling model easily leads to a mismatch between network resource allocation and actual service needs, resulting in problems such as network connectivity but lag or idle high-speed network resources, making it difficult to achieve efficient utilization of wireless communication resources and optimal user experience. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a multi-link collaborative wireless communication resource scheduling method to solve the problems of low service matching and insufficient resource utilization caused by the single scheduling strategy in existing methods.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] This invention provides a multi-link cooperative wireless communication resource scheduling method, which includes the following steps:

[0008] S1: Available Link Set Generation: Obtain the physical layer parameters of each wireless communication link of the computer device and generate an available link set;

[0009] S2: Data Stream Feature Analysis: Obtain the data stream to be transmitted, determine the demand type of the data stream based on the header features and traffic statistics features of the data stream, and obtain the data scale features of the data stream;

[0010] S3: Link Adaptation Quantification: Based on the required type, the physical layer parameters of each link in the available link set are weighted and calculated to generate a set of link adaptation values;

[0011] S4: Determine the transmission strategy: Based on the data scale characteristics, the demand type, and the link adaptation value set, determine and implement the transmission strategy for the data stream;

[0012] S5: Status Monitoring and Rescheduling: Monitor network resource status change events, update the link adaptation value set of each link, and optimize the transmission strategy of active data streams.

[0013] As a preferred embodiment of the multi-link collaborative wireless communication resource scheduling method of the present invention, the process of generating an available link set specifically includes: scanning the current communication environment through the underlying driver interface of the computer device, obtaining the reference signal received power, signal-to-noise ratio, channel busy ratio and physical negotiation rate of each wireless communication link, and associating the above physical layer parameters with the corresponding link identifiers to construct an available link set.

[0014] As a preferred embodiment of the multi-link collaborative wireless communication resource scheduling method of the present invention, the process of determining the demand type of the data stream specifically includes: extracting addressing and protocol identification information from the header features; extracting data packet size and transmission frequency features from the traffic statistics features; making a preliminary judgment based on the addressing and protocol identification information, and then optimizing based on the data packet size and transmission frequency features; when the data stream exhibits high-frequency transmission and low load characteristics, it is judged to be a latency-sensitive type; when the data stream exhibits continuous high load and high throughput characteristics, it is judged to be a bandwidth-consuming type; the rest are judged to be ordinary transmission types.

[0015] As a preferred embodiment of the multi-link collaborative wireless communication resource scheduling method of the present invention, the process of obtaining the data scale characteristics specifically includes: by monitoring the sending queue status of the operating system kernel and parsing the application layer protocol header information, extracting the average data packet size, the total amount of data to be transmitted, and the amount of data queue backlog of the data stream as data scale characteristics.

[0016] As a preferred embodiment of the multi-link collaborative wireless communication resource scheduling method of the present invention, the specific method of weighted calculation includes: converting the physical layer parameters of each link into signal quality indicators reflecting the link signal strength and purity, transmission capacity indicators reflecting the link transmission bandwidth and rate, and channel idle indicators reflecting the channel availability and idleness; configuring an appropriate weight strategy according to the determined demand type; and substituting the above indicators into the weight strategy for weighted calculation to obtain a set of link adaptation values.

[0017] As a preferred embodiment of the multi-link collaborative wireless communication resource scheduling method of the present invention, the weighting strategy includes: if the demand type is a delay-sensitive type, assigning the channel idle index the highest weight value, assigning the signal quality index the second highest weight value, and assigning the transmission capacity index the lowest weight value; if the demand type is a bandwidth consumption type, assigning the transmission capacity index the highest weight value, assigning the signal quality index the second highest weight value, and assigning the channel idle index the lowest weight value; if the demand type is a normal transmission type, assigning the channel idle index, the signal quality index, and the transmission capacity index weight values ​​that are similar to each other.

[0018] As a preferred embodiment of the multi-link collaborative wireless communication resource scheduling method of the present invention, the process of determining the transmission strategy specifically includes: combining the data scale characteristics, the demand type and the link adaptation value set, selecting physical links from the available link set to construct a link combination, and matching the transmission mode for the link combination according to the demand type.

[0019] As a preferred embodiment of the multi-link collaborative wireless communication resource scheduling method of the present invention, the transmission modes include: a preferred mapping mode in which the data stream is completely mapped to a single main link in the link combination for transmission; a load-sharing mode in which the transmission load of the data stream is distributed to each link in the link combination and bandwidth aggregation is achieved through multi-path concurrent transmission; and a concurrent redundancy mode in which the same data packet is copied and simultaneously injected into the transmission queues of each link in the link combination.

[0020] As a preferred embodiment of the multi-link collaborative wireless communication resource scheduling method of the present invention, the process of state monitoring and rescheduling specifically includes: real-time monitoring of network resource state change events; after detecting an event, first recalculating the link adaptation value set of active data streams, and then adjusting the transmission strategy of the active data streams according to the latest link adaptation value set.

[0021] This invention also provides a multi-link collaborative wireless communication resource scheduling system for executing the above method, specifically including the following modules:

[0022] Link acquisition module: Receives communication status information reported by the underlying hardware of the computer device, obtains the physical layer parameters of each wireless communication link, and generates a set of available links accordingly;

[0023] Feature analysis module: Receives data stream requests initiated by the application, parses the header features and traffic statistics features of the data stream, and outputs the determined demand type and data scale features;

[0024] Adaptation quantification module: Receives the physical layer parameters output by the link acquisition module and the demand type output by the feature analysis module, performs weighted calculations on them, and generates a set of link adaptation values ​​for the data stream;

[0025] Strategy determination module: Receives the data scale characteristics and demand type output by the feature analysis module, and the link adaptation value set output by the adaptation quantification module, determines the target transmission strategy for carrying the data stream, and implements it;

[0026] Status monitoring module: monitors network resource status change events in real time, and generates a rescheduling trigger signal when the event is detected;

[0027] Rescheduling module: Receives the rescheduling trigger signal sent by the status monitoring module, coordinates with the adaptation quantification module to update the link adaptation value set, and optimizes the transmission strategy of active data streams based on the update result.

[0028] The beneficial effects of this invention are as follows: it acquires multi-link physical parameters, identifies the demand type and data scale characteristics of data flows; it performs differentiated weighted scoring on each link based on the demand type, and generates link adaptation values ​​by combining real-time load correction; it formulates transmission strategies including optimal mapping, load sharing, or concurrent redundancy modes based on the above characteristics; and it triggers dynamic rescheduling based on network status monitoring. This invention achieves fine-grained resource scheduling based on service awareness, effectively improving the transmission efficiency and stability of heterogeneous networks. Attached Figure Description

[0029] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0030] Figure 1 This is a flowchart of a multi-link collaborative wireless communication resource scheduling method.

[0031] Figure 2 Define the flowchart for the data flow requirement type.

[0032] Figure 3 Flowchart for generating the link adaptation value set.

[0033] Figure 4 Flowchart for determining the transmission strategy.

[0034] Figure 5 This is a flowchart of a multi-link collaborative wireless communication resource scheduling system. Detailed Implementation

[0035] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0036] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0037] Secondly, the term "one embodiment" or "example" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the invention. The appearance of an embodiment in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that mutually excludes other embodiments.

[0038] Example 1

[0039] Reference Figures 1-4 This is one embodiment of the present invention, which provides a multi-link cooperative wireless communication resource scheduling method, including the following steps:

[0040] S1. Generation of Available Link Set: Obtain the physical layer parameters of each wireless communication link of the computer device and generate a set of available links.

[0041] The process of generating a set of available links specifically includes scanning the current communication environment through the low-level driver interface of the computer device, obtaining the reference signal received power, signal-to-noise ratio, channel busy ratio and physical negotiation rate of each wireless communication link, and associating the above physical layer parameters with the corresponding link identifiers to construct a set of available links.

[0042] The communication radio frequency (RF) components of computer equipment perform periodic channel scanning tasks to obtain physical layer status data for each link. These RF components encompass the network interface controller for wireless LANs, the baseband processor for cellular mobile networks, and the Bluetooth controller for Bluetooth near-field communication (NFC). In current smart terminal architectures, wireless LANs and cellular mobile networks constitute the two core pillars of data transmission. The former relies on widely deployed hotspots to provide high-bandwidth, low-cost local access, while the latter relies on operator base stations to provide wide-coverage, highly reliable wide-area access. Their widespread application makes them the preferred choice for multi-link collaborative scheduling. Furthermore, to address the special needs of ultra-low power consumption or NFC, Bluetooth links are also included as backup special link resources. To ensure the uniformity of heterogeneous network parameters, the acquired data follows standard network device interface specifications. All physical links in active or standby states are registered and mapped at the virtual network layer, forming a set of available links consisting of multiple link objects with independent transmission capabilities. This forms the basis for subsequent multi-link collaborative scheduling. Due to the significant heterogeneity in the hardware implementation of the underlying wireless LAN links, cellular mobile network links, and Bluetooth near-field communication links, directly scheduling these raw hardware interfaces would lead to extremely complex and difficult-to-extend upper-layer algorithms. By introducing a virtual network layer as an intermediate abstraction layer, different physical network cards are mapped to a unified virtual link object.

[0043] Differentiated scanning analysis is performed on links of different standards based on the underlying driver interface. For wireless LAN links, the underlying driver periodically executes an idle channel assessment mechanism to calculate the proportion of time the channel is occupied by interference signals within a unit time window, i.e., the channel busy ratio, and simultaneously reads the signal-to-noise ratio and negotiation rate information from the physical layer status information. For cellular mobile network links, the cellular mobile network communication radio frequency components measure the reference signal received power and signal-to-noise ratio of the currently connected cellular network nodes and estimate the current physical negotiation rate based on the scheduling information issued by the base station. For Bluetooth links, the connection interval and received signal strength are mainly monitored to assess its ability to carry low-speed heartbeat packets. These characteristics are organized into a standardized parameter structure, which includes not only first-order indicators describing signal coverage (such as RSRP and RSSI) but also second-order indicators describing channel interference levels and transmission potential (such as channel busy ratio and MCS index).

[0044] The raw physical layer parameters collected from each network interface are standardized and encapsulated, and bound to the corresponding link identifier. Heterogeneous data from different driver interfaces are uniformly mapped to a predefined data structure, thereby masking the differences in underlying hardware. All active physical links are fully included to ensure the completeness of the scheduling decision space. For example, even if a link has a weak signal or a congested channel, its current physical state value is still recorded and retained in the set of available links. All identified links and their corresponding real-time physical parameter vectors together constitute a structured dataset, namely the set of available links, providing the most basic data support for subsequent score-based fine-grained scheduling.

[0045] S2. Data Flow Feature Analysis: Based on the header features and traffic statistics of the data flow, determine the demand type of the data flow.

[0046] The process of determining the data stream demand type specifically includes: extracting addressing and protocol identification information from header features; extracting packet size and transmission frequency features from traffic statistics features; making a preliminary judgment based on addressing and protocol identification information, and then optimizing based on packet size and transmission frequency features; when the data stream exhibits high-frequency transmission and low load characteristics, it is judged as a latency-sensitive type; when the data stream exhibits continuous high load and high throughput characteristics, it is judged as a bandwidth-consuming type; the rest are judged as ordinary transmission types.

[0047] The process of obtaining data scale characteristics specifically includes: by monitoring the sending queue status of the operating system kernel and parsing the application layer protocol header information, extracting the average data packet size, the total amount of data to be transmitted, and the amount of data queue backlog in the data stream as data scale characteristics.

[0048] Specifically, hook functions are implanted between the network layer and the transport layer to perform lossless interception and metadata extraction of all inbound and outbound data packets. To achieve accurate determination of the demand type, a two-layer recognition architecture is constructed, which includes static identity recognition and dynamic behavior verification.

[0049] In the initial stages (such as when the Transmission Control Protocol (TCP) three-way handshake packets arrive, or when the first packet of a User Datagram Protocol (UDP) data stream arrives), addressing and protocol identification information (such as source IP address, destination IP address, source port number, destination port number, and transport protocol) is extracted by parsing the packet header. This static information provides a preliminary basis for determining the type of requirement. For example, the transport protocol field distinguishes between connection-oriented reliable transport (TCP) and connectionless real-time transport (UDP); the destination port number may correspond to a specific well-known service (port 1935 corresponds to the Real-Time Streaming Protocol, indicating live or video services); and the destination IP address may match the Content Delivery Network (CDN) address database.

[0050] Relying solely on header features carries the risk of false positives due to spoofing or reuse (e.g., web games reuse HTTP ports). Therefore, a dynamic verification mechanism based on traffic statistics is needed. This mechanism maintains a sliding time window (e.g., window length T = 500ms) for each active data stream and continuously tracks two core statistical metrics within the time window:

[0051] Average packet payload is used to accumulate the effective payload length of all packets within a window and calculate the average. For bandwidth-intensive services (such as video streaming and file downloads), in order to improve transmission efficiency, the protocol stack usually fills the maximum transmission unit as much as possible, so its average payload will continue to approach thousands of bytes; while for latency-sensitive services (such as online games), in order to reduce latency, the application layer usually generates a large number of tiny instruction packets of tens to hundreds of bytes.

[0052] The packet arrival interval variation coefficient is used to record the time difference between the arrival of consecutive data packets. Delay-sensitive services typically exhibit a very high frequency and uniform transmission rhythm, characterized by low intervals and low jitter; while ordinary transmission services (such as web browsing) exhibit obvious bursts and intermittentness.

[0053] Based on the aforementioned statistical characteristics, a real-time state machine is used to verify or correct the initial labels. Specifically, if a user datagram protocol (UDP) data stream initially identified as latency-sensitive shows a significant increase in its average packet load and approaches the maximum data packet length (MTU) over multiple consecutive time windows, the stream will be reclassified as actually performing large file transfers (such as QUIC downloads) based on UDP, thus real-time correcting its demand type to bandwidth-intensive. This dynamic identification mechanism, based on the entire lifecycle, ensures that subsequent scheduling strategies always match the actual current business needs.

[0054] The system periodically queries the network protocol stack's running status through the kernel communication interface provided by the operating system, extracting the average data packet size, send queue backlog, and total amount of data to be transmitted. First, using the kernel's traffic statistics function, it calculates the ratio of the total number of bytes transmitted to the total number of packets transmitted within a preset time window, thus obtaining the average data packet size. Next, it reads the send buffer status of the network socket to obtain the real-time number of bytes currently backed up in memory waiting to be sent, representing the send queue backlog and indicating the current transmission congestion. For application layer protocols (such as HTTP), it uses deep packet inspection or application layer callback interfaces to parse the content length field in the packet header, obtaining the total amount of data to be transmitted. Finally, it combines all this information to obtain the data scale characteristics.

[0055] S3. Link Adaptation Quantification: Based on the required type, the physical layer parameters of each link in the available link set are weighted and calculated to generate a set of link adaptation values.

[0056] The specific method of weighted calculation includes converting the physical layer parameters of each link into signal quality indicators representing the link signal strength and purity, transmission capacity indicators representing the link transmission bandwidth and rate, and channel idleness indicators representing the channel availability and idleness; configuring an appropriate weighting strategy according to the demand type; and then substituting the above indicators into the corresponding weighting strategy for weighted calculation to obtain a set of link adaptation values.

[0057] The weighting strategies include: for demand types that are latency-sensitive, the channel idle index is assigned the highest weight value, the signal quality index the second highest weight value, and the transmission capacity index the lowest weight value; for demand types that are bandwidth-consuming, the transmission capacity index is assigned the highest weight value, the signal quality index the second highest weight value, and the channel idle index the lowest weight value; for demand types that are normal transmission, the channel idle index, signal quality index, and transmission capacity index are assigned similar weight values.

[0058] In order to solve the technical problem that the parameters of heterogeneous networks are not directly calculated due to the huge differences in physical definition, value range and unit of measurement, a full-dimensional parameter normalization mapping mechanism was established.

[0059] Because different types of communication links are constrained by their respective underlying protocol standards, their performance bottlenecks manifest in different ways, making simple numerical comparisons of raw parameters mathematically unsound. Therefore, this mechanism constructs an abstract evaluation system independent of specific hardware implementations. This system includes three common core performance indicators: signal quality, representing link signal strength and purity; transmission capacity, representing link transmission bandwidth and rate, and the link's maximum data carrying potential under ideal conditions; and channel idleness, representing the availability and idleness of the channel. The core function of this mechanism is to uniformly project the diverse and attributed underlying raw data onto these three standardized evaluation dimensions through specific mathematical transformation logic, thereby eliminating logical barriers between heterogeneous parameters and laying the data foundation for unified scheduling of multiple links.

[0060] Specifically, for wireless LAN links, signal quality primarily depends on two core parameters: Received Signal Strength Indicator (RSSI) and Signal-to-Noise Ratio (SNR). RSSI reflects the power of the signal received by the terminal, while SNR reflects the clarity of the effective signal amidst background noise. To calculate the final signal quality index, effective physical intervals for both parameters need to be defined, and their relative scores within each interval are calculated. Then, based on weighted proportions, the processed RSSI and SNR scores are weighted and summed to synthesize a comprehensive signal quality index that reflects the robustness of the link's physical connection. In terms of capacity, the current physical layer negotiation rate is directly read and compared to the theoretical maximum rate of currently available links to obtain the transmission capacity index. In terms of interference, the idle channel assessment busy ratio (ISBN) from the underlying driver statistics is directly read. Since this value is a negative indicator and a percentage, the channel idleness index can be calculated by subtracting the ISBN from 1, representing the channel's idleness level. For cellular mobile communication links, the physical layer parameters reflect the characteristics of base station scheduling. Reference Received Power (RSRP) and Reference Received Quality (RSRQ) are extracted and mapped to signal quality indices. In terms of capacity, the carrier bandwidth configured at the base station and the maximum modulation and coding scheme supported by the current signal-to-noise ratio are obtained. The theoretical peak rate of the link under ideal conditions is calculated and compared with the theoretical maximum rate of currently available links to obtain the transmission capacity index. In terms of interference, since cellular networks do not have a contention backoff mechanism, their congestion is mainly manifested as an increase in interference level. Therefore, the normalized value of the signal-to-noise ratio is used to indirectly represent the purity of the environment, generating a channel idle index.

[0061] After completing the index mapping, the data flow demand types determined in S2 are obtained, and corresponding weight schemes are set according to the demand types to define the relative importance of signal quality indicators, transmission capacity indicators, and channel idle indicators in the scoring system. The aim is to ensure that the final scoring results accurately reflect the carrying capacity of each link for the current specific demand type through differentiated weight allocation. Finally, the processed index values ​​of each link and the determined weight coefficients are substituted into the weighting formula for calculation.

[0062] For latency-sensitive types (such as online games), the core requirements are extremely low network jitter and extremely high transmission stability, while bandwidth size is not a major concern. A low channel idle rate indicates severe congestion in the current communication environment, making data packets highly susceptible to queuing and blocking. A low signal quality rate indicates unstable physical connections, also leading to packet loss and retransmission. Both of these directly result in severe service interruptions. Therefore, the channel idle rate, representing environmental purity, is assigned the first priority weight, and the signal quality rate, representing connection stability, is assigned the second priority weight. Since latency-sensitive data packets are typically small, the transmission capacity rate is assigned the third priority weight.

[0063] For bandwidth-intensive types (such as large file downloads), the core requirement is the maximum data throughput per unit time to shorten the total transmission time. A low transmission capacity indicator means the physical channel of the link is narrow, directly limiting the speed ceiling and becoming a decisive factor affecting user experience. A low signal quality indicator, while leading to connection congestion and window adjustments, can be managed through retransmissions. As for the channel idleness indicator, due to the filling effect of high-volume transmission, even in a slightly congested environment, considerable throughput can still be achieved as long as the channel is wide enough. Therefore, the transmission capacity indicator, representing the link's throughput potential, is assigned the first priority weight; the signal quality indicator, representing connection stability, is assigned the second priority weight; and the channel idleness indicator, which has a relatively smaller impact on the total throughput, is assigned the third priority weight.

[0064] For common transmission types (such as instant messaging text), there is no extreme one-way preference for latency, bandwidth, and stability; the goal is a balance of performance across all aspects. If any single indicator is too low, it may affect the user's basic experience at different levels. Therefore, a balanced strategy is adopted, which allocates weights evenly to channel idleness indicators, signal quality indicators, and transmission capacity indicators to ensure that the scheduling results are not hampered by a deviation in any one dimension.

[0065] Finally, for the specific data flow, the adaptation score of each available link is calculated using a linear weighted model, and the calculation formula is as follows:

[0066] in, Indicates the first The link adaptation value for this data stream; This indicates the channel idle index of the link; This indicates the signal quality metric of the link; This indicates the transmission capacity of the link; These are the corresponding weight coefficients determined based on the data flow requirement type. By performing the above calculations across all available links, a list containing all... An ordered list, or link adaptation value set, is generated. This set quantifies the differences in the service capabilities of each physical link for the current data flow, providing data support for determining subsequent transmission strategies. Through this mechanism, heterogeneous parameters at the physical layer are successfully transformed into business adaptation scores that are understandable and comparable to those of upper-layer applications.

[0067] S4. Determine the transmission strategy: Based on the data scale characteristics of the data stream, the demand type, and the link adaptation value set, determine and implement the transmission strategy of the data stream.

[0068] The process of determining the transmission strategy specifically includes: combining data scale characteristics, demand type and link adaptation value set, selecting physical links from the available link set to construct link combinations, and matching transmission modes for the link combinations according to demand type.

[0069] The transmission modes include: a preferred mapping mode that maps the entire data stream to a single main link in the link combination for transmission; a load-sharing mode that distributes the transmission load of the data stream to each link in the link combination and achieves bandwidth aggregation through multi-path concurrent transmission; and a concurrent redundancy mode that copies the same data packet and injects it into the sending queues of each link in the link combination.

[0070] Among them, the multi-link aggregation mechanism has a startup cost. If the total amount of data to be transmitted is too low (e.g., 1MB), the additional overhead caused by forcibly opening multiple links will exceed the benefits of bandwidth aggregation, resulting in a decrease in overall transmission efficiency. Therefore, if the total amount of data to be transmitted is too low, a single-link optimal mapping transmission strategy is directly adopted.

[0071] Next, by combining the characteristics of data scale, demand type and link adaptation value set, physical links are selected from the available link set to construct link combinations and match the corresponding transmission modes.

[0072] For ordinary transmission types, based on the set of link adaptation values, the single physical link with the highest score is selected to construct a single link combination, and the optimal mapping mode is directly matched to guide the data flow to the optimal link for transmission in order to minimize overhead.

[0073] For data types exhibiting high bandwidth consumption, the primary link with the highest adaptation value is selected, and the second-best link is incorporated as an auxiliary link to construct a multi-link combination including primary and auxiliary links. Once constructed, the load-sharing mode is matched, and the primary and auxiliary links are controlled to distribute different data packets or fragments of the data stream according to their respective bandwidth capabilities, thereby achieving effective aggregation of bandwidth resources.

[0074] For latency-sensitive types, without considering the data size, the two physical links with the highest and second highest scores are directly selected based on the link adaptation value set. This forces the construction of a multi-link combination and matches a concurrent redundancy mode. All links in the combination are controlled to send the same data packet copy of the data stream simultaneously, using the redundancy of physical resources to achieve the ultimate transmission reliability.

[0075] S5. Status Monitoring and Rescheduling: Monitor network resource status change events, update the link adaptation value set of each link, and optimize the transmission strategy of active data streams.

[0076] The process of state monitoring and rescheduling specifically includes: real-time monitoring of network resource state change events; after an event is detected, first recalculating the link adaptation value set of active data streams, and then adjusting the transmission strategy of the active data streams based on the latest link adaptation value set.

[0077] A closed-loop feedback mechanism covering the entire lifecycle has been constructed to ensure that transmission strategies can dynamically adapt to real-time changes in the network environment and service status. A lightweight state daemon runs, continuously receiving asynchronous event notifications from the physical link layer and application layer, focusing on capturing three core events: link quality degradation, discovery of better links, and service lifecycle changes. When the adaptation value of the carrying link is continuously lower than the preset service quality threshold, or when a new link's adaptation value is found to be significantly better than the current link, or when a data stream transmission task is detected to have ended, it is determined that the network resource status has changed, thus triggering rescheduling. In addition, there are link access point change events (such as base station replacement, wireless LAN replacement, etc.). Since the basic attributes of the physical link (such as frequency band, bandwidth, IP address, etc.) may have been completely changed, a full process reset will be performed.

[0078] During rescheduling, differentiated processing is performed based on the nature of the triggering event. For events involving link quality degradation or the discovery of a better link, or changes in the link environment, the system first establishes a new link connection for the optimized data stream and confirms the data path's availability before dismantling the old link connection. This minimizes packet loss risks during the handover process and ensures service continuity. For events involving changes in the service lifecycle, the system performs resource release and global re-optimization. For example, when a large file download task is detected as complete, the link bandwidth resources occupied by that task are released, significantly reducing the real-time load on that link and thus improving its adaptability. Subsequently, all currently active data streams are traversed, and the latest link status after resource release is reassessed. If it is found that for a running data stream (such as a video stream using a cellular network), the adaptability of the released WLAN link has surpassed that of the current link, the system will perform a handover operation, migrating the data stream to a better WLAN link, thereby maximizing global resource utilization.

[0079] Example 2

[0080] Reference Figure 5 These are two embodiments of the present invention. This embodiment provides a multi-link collaborative wireless communication resource scheduling system, which is typically deployed on a high-performance computing server to execute the steps described in Embodiment 1.

[0081] The system's software architecture consists of a set of highly collaborative functional modules, specifically including:

[0082] Link acquisition module: Receives communication status information reported by the underlying hardware of the computer device, obtains the physical layer parameters of each wireless communication link, and generates a set of available links accordingly;

[0083] Feature analysis module: Receives data stream requests initiated by the application, parses the header features and traffic statistics features of the data stream, and outputs the determined demand type and data scale features;

[0084] Adaptation quantification module: Receives the physical layer parameters output by the link acquisition module and the demand type output by the feature analysis module, performs weighted calculations on them, and generates a set of link adaptation values ​​for the data stream;

[0085] Strategy determination module: Receives the data scale characteristics and demand type output by the feature analysis module, and the link adaptation value set output by the adaptation quantification module, determines the target transmission strategy for carrying the data stream, and implements it;

[0086] Status monitoring module: monitors network resource status change events in real time, and generates a rescheduling trigger signal when the event is detected;

[0087] Rescheduling module: Receives the rescheduling trigger signal sent by the status monitoring module, coordinates with the adaptation quantification module to update the link adaptation value set, and optimizes the transmission strategy of active data streams based on the update result.

[0088] In the initial stage of data flow establishment, the link acquisition module continuously reads real-time status data reported by the underlying communication hardware (such as Wi-Fi chips and baseband processors) through the bus interface and aggregates it into a set of available links. At the same time, when the upper-layer application initiates a network connection request, the feature analysis module immediately intercepts the request, extracts its header features, and starts a traffic statistics window to quickly identify whether the data flow is latency-sensitive, bandwidth-consuming, or normal transmission, and calculates its data scale characteristics.

[0089] The processed business feature information is passed to the adaptation and quantification module. This module, combined with the physical layer parameters provided by the link acquisition module, calls the corresponding weighting algorithm based on the identified demand type to calculate the adaptation score for each available link in the data stream, generating a set of link adaptation values. Next, the strategy determination module, acting as the decision-making center, simultaneously receives business demand data from the feature analysis module and link score data from the adaptation and quantification module. After comprehensive judgment, it outputs the final transmission strategy and instructs the underlying driver to execute data distribution.

[0090] During the data stream transmission phase, the status monitoring module remains continuously active, monitoring network environment fluctuations and changes in the service lifecycle in real time. Once a change event triggering rescheduling is detected, it immediately sends a signal to the rescheduling module. Upon receiving the signal, the rescheduling module reactivates the adaptation metric module to recalculate the current link score and performs seamless policy adjustments to the active data streams based on the latest score, thus forming a closed-loop adaptive control system.

[0091] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A multi-link collaborative wireless communication resource scheduling method, characterized in that, Performed by a computer device, including the following steps: S1: Available Link Set Generation: Obtain the physical layer parameters of each wireless communication link of the computer device and generate an available link set; S2: Data Stream Feature Analysis: Obtain the data stream initiated by the application, determine the demand type of the data stream based on the header features and traffic statistics features of the data stream, and obtain the data scale features of the data stream; S3: Link Adaptation Quantification: Based on the required type, the physical layer parameters of each link in the available link set are weighted and calculated to generate a set of link adaptation values; S4: Determine the transmission strategy: Based on the data scale characteristics, the demand type, and the link adaptation value set, determine and implement the transmission strategy for the data stream; S5: Status Monitoring and Rescheduling: Monitor network resource status change events, update the link adaptation value set of each link, and optimize the transmission strategy of active data streams.

2. The multi-link collaborative wireless communication resource scheduling method according to claim 1, characterized in that, The process of generating a set of available links specifically includes: scanning the current communication environment through the underlying driver interface of the computer device, obtaining the reference signal received power, signal-to-noise ratio, channel busy ratio and physical negotiation rate of each wireless communication link, and associating the above physical layer parameters with the corresponding link identifiers to construct a set of available links.

3. The multi-link collaborative wireless communication resource scheduling method according to claim 1, characterized in that, The process of determining the demand type of the data stream specifically includes: extracting addressing and protocol identification information from the header features; extracting data packet size and transmission frequency features from the traffic statistics features; making a preliminary judgment based on the addressing and protocol identification information, and then optimizing based on the data packet size and transmission frequency features; when the data stream exhibits high-frequency transmission and low-load characteristics, it is judged as a latency-sensitive type; when the data stream exhibits continuous high load and high throughput characteristics, it is judged as a bandwidth-consuming type; the rest are judged as ordinary transmission types.

4. The multi-link collaborative wireless communication resource scheduling method according to claim 1, characterized in that, The process of obtaining the data scale characteristics specifically includes: by monitoring the sending queue status of the operating system kernel and parsing the application layer protocol header information, extracting the average data packet size, the total amount of data to be transmitted, and the amount of data queue backlog of the data stream as data scale characteristics.

5. The multi-link collaborative wireless communication resource scheduling method according to claim 1, characterized in that, The specific method of weighted calculation includes: converting the physical layer parameters of each link into signal quality indicators reflecting the link signal strength and purity, transmission capacity indicators reflecting the link transmission bandwidth and rate, and channel idleness indicators reflecting the availability and idleness of the channel; configuring an appropriate weighting strategy according to the determined demand type; and substituting the above indicators into the weighting strategy for weighted calculation to obtain a set of link adaptation values.

6. The multi-link cooperative wireless communication resource scheduling method according to claim 4, characterized in that, The weighting strategy includes: if the demand type is latency-sensitive, assign the highest weight value to the channel idle index, the second highest weight value to the signal quality index, and the lowest weight value to the transmission capacity index; if the demand type is bandwidth consumption, assign the highest weight value to the transmission capacity index, the second highest weight value to the signal quality index, and the lowest weight value to the channel idle index; if the demand type is normal transmission, assign weight values ​​to the channel idle index, the signal quality index, and the transmission capacity index that are similar to each other.

7. The multi-link cooperative wireless communication resource scheduling method according to claim 1, characterized in that, The process of determining the transmission strategy specifically includes: combining the data scale characteristics, the demand type, and the link adaptation value set, selecting physical links from the available link set to construct a link combination, and matching the transmission mode for the link combination according to the demand type.

8. The multi-link cooperative wireless communication resource scheduling method according to claim 7, characterized in that, The transmission modes include: a preferred mapping mode that maps the data stream completely to a single main link in the link combination for transmission; a load-sharing mode that distributes the transmission load of the data stream to each link in the link combination and achieves bandwidth aggregation through multi-path concurrent transmission; and a concurrent redundancy mode that copies the same data packet and injects it into the sending queue of each link in the link combination.

9. The multi-link cooperative wireless communication resource scheduling method according to claim 1, characterized in that, The state monitoring and rescheduling process specifically includes: real-time monitoring of network resource state change events; Upon detecting an event, the link adaptation value set of the active data stream is recalculated, and then the transmission strategy of the active data stream is adjusted based on the latest link adaptation value set.

10. A multi-link collaborative wireless communication resource scheduling system, characterized in that, The system is used to perform the method according to any one of claims 1 to 9, specifically including: Link acquisition module: Receives communication status information reported by the underlying hardware of the computer device, obtains the physical layer parameters of each wireless communication link, and generates a set of available links accordingly; Feature analysis module: Receives data stream requests initiated by the application, parses the header features and traffic statistics features of the data stream, and outputs the determined demand type and data scale features; Adaptation quantification module: Receives the physical layer parameters output by the link acquisition module and the demand type output by the feature analysis module, performs weighted calculations on them, and generates a set of link adaptation values ​​for the data stream; Strategy determination module: Receives the data scale characteristics and demand type output by the feature analysis module, and the link adaptation value set output by the adaptation quantification module, determines the target transmission strategy for carrying the data stream, and implements it; Status monitoring module: monitors network resource status change events in real time, and generates a rescheduling trigger signal when the event is detected; Rescheduling module: Receives the rescheduling trigger signal sent by the status monitoring module, coordinates with the adaptation quantification module to update the link adaptation value set, and optimizes the transmission strategy of active data streams based on the update result.

Images