A stream media resource scheduling method and system in a high-concurrency live broadcast scene

Abstract

The application provides a stream media resource scheduling method and system in a high-concurrency live broadcast scene, relates to the technical field of edge computing, collects live broadcast request data, stream media transmission states and edge node resource states, and then constructs resource competition relations; the resource competition relations are associated and partitioned, a resource competition topology is constructed and resource competition groups are determined, resource association matching is performed on the resource competition groups based on the edge node resource states, resource coordination capabilities are obtained; the evolution trend of live broadcast popularity is determined through historical access records, the resource coordination capabilities are constrained and associated through the evolution trend, and resource scheduling constraints are obtained; collaborative optimization is performed according to the resource competition topology and the resource scheduling constraints, and a scheduling strategy for edge node resources and stream media transmission paths is obtained; the application can realize collaborative constraint scheduling of resource competition relations and live broadcast popularity evolution, so as to enhance the dynamic adaptation capability of stream media resource allocation in a high-concurrency live broadcast scene.

Description

Technical Field

[0001] This application relates to the field of edge computing technology, and more specifically, to a method and system for scheduling streaming media resources in high-concurrency live streaming scenarios. Background Technology

[0002] With the rapid development of internet video services, live streaming platforms have gradually become an important carrier for online information dissemination and real-time interaction. In high-concurrency live streaming scenarios such as large-scale sports events, e-commerce live streaming, online education, and online entertainment, a large number of concurrent user access requests will be generated in a short period of time, causing the live streaming server, edge nodes, and network transmission links to be under great resource scheduling pressure.

[0003] To reduce the load on central servers and shorten user access latency, edge computing technology is increasingly being applied to streaming media transmission. By deploying computing, caching, and network resources at edge nodes closer to the user, streaming media content can be cached locally and distributed quickly, thereby improving streaming media transmission efficiency. However, current streaming media resource scheduling in high-concurrency live streaming scenarios mostly focuses on a single dimension of edge node resource status or changes in live stream popularity, lacking a unified characterization of resource competition relationships between live stream requests. Furthermore, existing technologies do not adequately consider the synergistic relationship between changes in live stream popularity and resource competition status. Under conditions of rapid fluctuations in high-concurrency access, problems such as unbalanced edge node resource allocation, frequent switching of streaming media transmission paths, and delayed resource scheduling responses can easily occur. Therefore, how to achieve coordinated constraint scheduling based on resource competition relationships and the evolution of live stream popularity to enhance the dynamic adaptability of streaming media resource allocation in high-concurrency live streaming scenarios is a challenge facing the industry. Summary of the Invention

[0004] This application provides a streaming media resource scheduling method and system for high-concurrency live streaming scenarios, which can realize coordinated scheduling based on resource competition and the evolution of live streaming popularity, thereby enhancing the dynamic adaptability of streaming media resource allocation in high-concurrency live streaming scenarios.

[0005] Firstly, this application provides a method for scheduling streaming media resources in high-concurrency live streaming scenarios, the scheduling method comprising the following steps: Collect live streaming request data, streaming media transmission status, and edge node resource status from the live streaming server, and construct resource competition relationships based on the live streaming request data and the streaming media transmission status; The resource competition relationship is partitioned into related regions, a resource competition topology is constructed and a resource competition group is determined, and resource association matching is performed on the resource competition group based on the resource status of the edge nodes to obtain resource coordination capabilities; Obtain historical access records of the live streaming server, determine the evolution trend of live streaming popularity through the historical access records, and constrain and correlate the resource coordination capability through the evolution trend to obtain resource scheduling constraints; Based on the resource contention topology and the resource scheduling constraints, a scheduling strategy for edge node resources and streaming media transmission paths is obtained through collaborative optimization.

[0006] In this embodiment, collecting live streaming request data, streaming media transmission status, and edge node resource status from the live streaming server specifically includes: Collect the live streaming request information received by the live streaming server to obtain live streaming request data; Based on the live stream request data, obtain the transmission information of the corresponding live stream to obtain the streaming media transmission status; The edge nodes of the corresponding live stream are determined by the live stream request data, and the resource status of the edge nodes is collected to obtain the resource status of the edge nodes.

[0007] In this embodiment, constructing a resource contention relationship based on the live stream request data and the streaming media transmission status specifically includes: Based on the live stream request data, the degree of resource access overlap between live stream requests is statistically analyzed to obtain a request competition association set; Based on the analysis of the streaming media transmission status, the degree of link sharing and cache reuse between live streams is determined to obtain a resource sharing association set; The request competition association set and the resource sharing association set are fused together to obtain the resource competition relationship.

[0008] In this embodiment, the process of associating and partitioning the resource contention relationships, constructing a resource contention topology, and determining resource contention groups specifically includes: Based on the resource competition relationship, live streaming resources with similar levels of competitive correlation are aggregated to obtain multiple resource competition regions; Based on the resource competition relationships between the resource competition regions, resource competition connections are established, thereby constructing a resource competition topology; The degree of competition between resource competition regions is calculated using the resource competition topology to obtain a topological influence sequence; The resource competition topology is hierarchically divided by the topological influence sequence to obtain resource competition groups.

[0009] In this embodiment, the resource association matching of the resource competition groups based on the resource status of the edge nodes to obtain resource coordination capabilities specifically includes: Resource capability analysis is performed on the resource status of the edge nodes to obtain the resource capability parameters of the edge nodes; Based on the resource capability parameters, the resource competition groups are associated and mapped to obtain resource matching results; The resource matching results are analyzed for collaboration to obtain the degree of resource collaboration in each resource competition group, thereby determining the resource collaboration capability.

[0010] In this embodiment, obtaining the historical access records of the live streaming server specifically includes: Obtain live streaming request data from the live streaming server over multiple consecutive scheduling periods, and aggregate the accesses according to the live streaming identifier to obtain the historical access sequence. The historical access sequence is sorted by time to obtain the historical access records.

[0011] In this embodiment, determining the evolution trend of live stream popularity through the historical access records specifically includes: The historical access records are segmented by time to obtain popularity change data; Based on the aforementioned popularity change data, a continuous correlation analysis was performed to obtain the trajectory of live stream popularity changes; By performing trend fitting on the trajectory of the live stream popularity change, the evolution trend of the live stream popularity can be obtained.

[0012] In this embodiment, the resource coordination capability is constrained and correlated through the evolutionary trend to obtain resource scheduling constraints, specifically including: The range of popularity variation for live streaming resources is determined based on the aforementioned live streaming popularity evolution trend; The resource coordination capability is categorized and constrained based on the heat variation range to obtain resource constraint results; The resource constraint results are correlated and corrected to obtain resource scheduling constraints.

[0013] In this embodiment, the scheduling strategy for edge node resources and streaming media transmission paths is obtained through collaborative optimization based on the resource contention topology and the resource scheduling constraints, specifically including: Determine the resource competition paths between edge nodes based on the resource competition topology; The resource scheduling path is obtained by coordinating the resource contention path through the resource scheduling constraints. Based on the resource scheduling path, the edge node resources and streaming media transmission path are jointly optimized to obtain a scheduling strategy for edge node resources and streaming media transmission path.

[0014] Secondly, this application provides a streaming media resource scheduling system for high-concurrency live streaming scenarios, used to execute a streaming media resource scheduling method for high-concurrency live streaming scenarios, the scheduling system comprising: The contention awareness module is used to collect live streaming request data, streaming media transmission status and edge node resource status from the live streaming server, and to construct resource contention relationships based on the live streaming request data and the streaming media transmission status. The collaborative analysis module is used to partition the resource competition relationship, construct the resource competition topology and determine the resource competition group, and perform resource association matching on the resource competition group based on the resource status of the edge node to obtain resource collaboration capability. The trend constraint module is used to obtain historical access records of the live streaming server, determine the evolution trend of the live streaming popularity through the historical access records, and constrain the resource coordination capability through the evolution trend to obtain resource scheduling constraints. The collaborative scheduling module is used to perform collaborative optimization based on the resource contention topology and the resource scheduling constraints to obtain a scheduling strategy for edge node resources and streaming media transmission paths.

[0015] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects: The technical solution provided in this application enables coordinated scheduling of resource competition relationships and live streaming popularity evolution. Firstly, by collecting live streaming request data, streaming media transmission status, and edge node resource status from the live streaming server, and constructing resource competition relationships based on the live streaming request data and streaming media transmission status, it comprehensively reflects the resource occupancy association and transmission competition status between different live streaming resources in high-concurrency live streaming scenarios. This provides a data foundation for accurately characterizing resource competition characteristics and avoids the problem of traditional methods that rely solely on node load for resource scheduling, which fails to comprehensively reflect resource competition relationships. Secondly, by partitioning the resource competition relationships, constructing a resource competition topology, and determining resource competition groups, and then combining the edge node resource status to perform resource association matching on the resource competition groups, resource coordination capabilities are obtained. This establishes a coordinated mapping mechanism between resource competition relationships and edge node resource status, enabling differentiated coordinated analysis of resource competition areas and improving resource scheduling efficiency. The process involves several steps: First, determining the rationality of resource matching and the efficiency of resource utilization during the live streaming process. Second, by acquiring historical access records of the live streaming server, the evolution trend of live streaming popularity is determined. Based on this evolution trend, the resource coordination capability is constrained and correlated to obtain resource scheduling constraints. This allows the dynamic changes in live streaming popularity to be introduced into the resource coordination analysis process, achieving joint constraints between resource competition status and the evolution of live streaming popularity, thus enhancing the dynamic response capability of resource scheduling to business changes. Finally, based on the resource competition topology and the resource scheduling constraints, a coordinated optimization is performed to obtain scheduling strategies for edge node resources and streaming media transmission paths. This achieves coordinated constraint scheduling between resource competition relationships and the evolution of live streaming popularity, enabling dynamic coordinated optimization of edge node resource allocation and streaming media transmission path adjustment as resource competition status and live streaming popularity change. This enhances the dynamic adaptability of streaming media resource allocation in high-concurrency live streaming scenarios and improves the resource scheduling efficiency and transmission stability of live streaming services.

[0016] In summary, the technical solution adopted in this application can achieve coordinated scheduling of resource competition and live streaming popularity evolution, thereby enhancing the dynamic adaptability of streaming media resource allocation in high-concurrency live streaming scenarios. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of a streaming media resource scheduling method for high-concurrency live streaming scenarios provided in this application; Figure 2This is an exemplary flowchart for determining resource collaboration capabilities provided in this application; Figure 3 This is an exemplary flowchart for determining the evolution trend of live stream popularity based on the information provided in this application; Figure 4 This is a schematic diagram of a high-concurrency live streaming scenario provided in this application; Figure 5 This is a module structure diagram of a streaming media resource scheduling system for high-concurrency live streaming scenarios provided in this application. Detailed Implementation

[0019] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0020] Example 1: To better understand the technical solution of this application, the above technical solution will be described in detail below with reference to the accompanying drawings and specific embodiments. Figure 1 As shown in the figure, this is a flowchart of a streaming media resource scheduling method in a high-concurrency live streaming scenario according to this embodiment of the present application. The scheduling method includes the following steps: In step S1, the live streaming request data, streaming media transmission status, and edge node resource status of the live streaming server are collected, and a resource competition relationship is constructed based on the live streaming request data and the streaming media transmission status.

[0021] In this embodiment, the collection of live streaming request data, streaming media transmission status, and edge node resource status from the live streaming server can be carried out in the following manner: Collect the live streaming request information received by the live streaming server to obtain live streaming request data; Based on the live stream request data, obtain the transmission information of the corresponding live stream to obtain the streaming media transmission status; The edge nodes of the corresponding live stream are determined by the live stream request data, and the resource status of the edge nodes is collected to obtain the resource status of the edge nodes.

[0022] In specific implementation, firstly, the request receiving interface deployed in the live streaming server access layer continuously receives live streaming access requests initiated by user terminals, and extracts basic access information such as live stream identifier, user identifier, request time, request source address, and request resource identifier for each live streaming access request. Then, the collected basic access information is categorized according to the live stream identifier and sorted by time according to the order of request arrival, forming a request data set that reflects the access process of the same live streaming service. This request data set is then combined into live streaming request data. Next, based on the live stream identifier and request resource identifier in the live streaming request data, the corresponding streaming media distribution record of the live streaming server is queried to obtain the data transmission information of the current live stream during transmission. Specifically, the transmission link information, cache node information, data transmission latency information, and data transmission bandwidth usage information corresponding to the live stream can be further obtained and associated and organized according to the same live stream, forming a data set that reflects the actual transmission status of the live stream. This data set is then combined into streaming media transmission status. Finally, based on the live stream identifier and request source information in the live stream request data, the edge node responsible for the current live stream service is determined, and a resource status collection instruction is sent to the corresponding edge node. After receiving the resource status collection instruction, the edge node collects its current processor resource usage, storage resource usage, cache resource usage, network bandwidth usage, and current service load in real time, and returns the collected resource status information to the live stream server. The returned resource status information is then uniformly organized to form the resource status data of the corresponding edge node, and the resource status data is used as the resource status of the edge node.

[0023] It should be noted that the live streaming request data in this application refers to the set of request information generated by users accessing the live streaming service received by the live streaming server. This data characterizes the access status of the live streaming service at different times, by different users, and on different resource objects, providing a foundational data source for subsequent analysis of the competitive relationships between live streaming resources. Streaming media transmission status refers to the actual operational status information of the live streaming stream during network transmission, including data transmission path, cache node distribution, network transmission latency, network bandwidth usage, and data transmission stability. This data characterizes the resource usage and transmission status of the live streaming stream during transmission. Edge node resource status refers to the current resource usage of edge nodes available for live streaming services, including computing resource usage, storage resource usage, remaining cache resources, network bandwidth utilization, and node service load. This data characterizes the resource carrying capacity of edge nodes for the live streaming service, providing a basis for subsequent resource competition grouping and resource association matching.

[0024] In this embodiment, the resource contention relationship can be constructed based on the live streaming request data and the streaming media transmission status in the following ways: Based on the live stream request data, the degree of resource access overlap between live stream requests is statistically analyzed to obtain a request competition association set; Based on the analysis of the streaming media transmission status, the degree of link sharing and cache reuse between live streams is determined to obtain a resource sharing association set; The request competition association set and the resource sharing association set are fused together to obtain the resource competition relationship.

[0025] In practice, firstly, the live stream identifier, edge node identifier, and request time are extracted from the live stream request data, and any two live stream requests are compared item by item. If two live stream requests access the same live stream resource or are provided by the same edge node, a corresponding resource access association record is generated. The number of times any two live stream requests jointly access the live stream resource and the duration of jointly occupying the same edge node are then counted, and the degree of resource access overlap is determined based on the number of joint accesses and the duration of joint occupation. After all live stream requests are counted, a data set consisting of the correspondence between live stream requests and the degree of resource access overlap is formed, and this data set is then used as a request competition association set. Then, based on the streaming media transmission status, the data transmission path, cache node information, and cache content identifier corresponding to each live stream are read, and the data transmission path is split into multiple data transmission nodes and data transmission links. The data transmission nodes and data transmission links of any two live streams are compared item by item, and the number of identical data transmission nodes and identical data transmission links are counted. The link sharing degree is determined based on the statistical results. At the same time, it is determined whether the two live streams access the same cache node and read the same cache content. The cache reuse count is counted, and the cache reuse degree is determined based on the ratio of the cache reuse count to the total cache access count. The link sharing degree and cache reuse degree are then organized according to the live stream identifier to form a data set composed of live stream identifier, link sharing degree, and cache reuse degree. The data set is then combined into a resource sharing association set. Finally, an association relationship is established between the request competition association set and the resource sharing association set according to the same live stream resource identifier, and the request competition information and resource sharing information corresponding to the same live stream resource are uniformly organized; a correspondence relationship is established between the request competition information and resource sharing information corresponding to the same live stream resource; if a live stream resource has both a request competition relationship and a resource sharing relationship, a resource competition connection relationship is established between the corresponding live stream resources; if only one type of association relationship exists, a resource competition connection relationship is established according to the existing association relationship; after the association relationship between all live stream resources is established, the resulting live stream resource association structure is taken as the resource competition relationship.

[0026] It should be noted that, in this application, the degree of resource access overlap refers to the extent to which different live streaming requests jointly access the same live streaming resource or jointly occupy the same edge node resource, used to characterize the strength of resource competition among different live streaming requests. The request competition association set refers to a data set composed of resource access associations among multiple live streaming requests, used to reflect the distribution of resource competition among live streaming requests. The link sharing degree refers to the extent to which different live streaming streams share the same data transmission path or the same data transmission node, used to characterize the data transmission resource sharing among live streaming streams. The cache reuse degree refers to the extent to which different live streaming streams jointly call the same cache resource, used to characterize the situation where cache resources are reused by multiple live streaming streams. The resource sharing association set refers to a data set composed of resource sharing relationships among multiple live streaming streams, used to reflect the sharing association among live streaming resources. The resource competition relationship refers to the resource association relationship formed by combining the resource competition among live streaming requests and the resource sharing among live streaming streams, used to characterize the degree of resource competition among different live streaming resources, providing a basis for subsequent resource competition association partitioning, resource competition topology construction, and resource competition grouping.

[0027] In step S2, the resource competition relationship is partitioned into related regions, a resource competition topology is constructed and a resource competition group is determined, and resource association matching is performed on the resource competition group based on the resource status of the edge nodes to obtain resource coordination capabilities.

[0028] In this embodiment, the resource contention relationships are associated and partitioned to construct a resource contention topology and determine resource contention groups, which can be done in the following way: Based on the resource competition relationship, live streaming resources with similar levels of competitive correlation are aggregated to obtain multiple resource competition regions; Based on the resource competition relationships between the resource competition regions, resource competition connections are established, thereby constructing a resource competition topology; The degree of competition between resource competition regions is calculated using the resource competition topology to obtain a topological influence sequence; The resource competition topology is hierarchically divided by the topological influence sequence to obtain resource competition groups.

[0029] In specific implementation, firstly, resource competition association information between each live streaming resource is extracted based on the resource competition relationship, and the number and strength of resource competition associations between each live streaming resource and other live streaming resources are counted. Then, comparisons are made according to the number of resource competition associations, the objects of resource competition associations, and the strength of resource competition associations. Live streaming resources with the same object of resource competition association or similar strength of resource competition associations are grouped into the same set, and the grouped live streaming resources are merged to form several live streaming resource sets. Next, a consistency check is performed on the resource competition associations in each live streaming resource set to ensure that the live streaming resources in the same set maintain similar resource competition association characteristics, thus treating each live streaming resource set as a resource competition region. Secondly, a correspondence relationship between regions is established based on the resource competition associations existing between the resource competition regions. Specifically, the live streaming resources in each resource competition region are traversed, and it is determined whether there are resource competition associations between live streaming resources in different resource competition regions. If resource competition associations exist, a connection relationship is established between the corresponding resource competition regions. Then, the connection relationships between all resource competition regions are uniformly organized to form an overall association structure between the resource competition regions, and this overall association structure is used as the resource competition topology. Then, based on the resource competition topology, the number and strength of connections between each resource competition region and other resource competition regions are counted one by one. Combining the connection positions of each resource competition region in the resource competition topology, the number of resource competition connections established between each resource competition region and other resource competition regions is counted. The resource competition association of each resource competition connection with corresponding live resources is then counted. Based on the number and strength of resource competition connections, the magnitude of the resource competition impact of each resource competition region on other resource competition regions is determined, and the magnitude of the impact is taken as the degree of competition influence. The analyzed competition influence of each resource competition region is then arranged in descending order of influence degree, forming a ranking result for the resource competition regions. This ranking result is then used as the topological influence sequence. Finally, each resource competition region is traversed sequentially according to the topological influence sequence, and resource competition regions with similar degrees of competition influence are grouped into the same level. The resource competition regions in each level are then uniformly aggregated to form multiple resource sets with different levels of competition influence, and the resource sets corresponding to each level are used as resource competition groups.

[0030] It should be noted that, in this application, a resource competition region refers to a set of resources formed by multiple live-streaming resources with similar resource competition relationships, used to characterize the distribution range of live-streaming resources with similar resource competition characteristics. Resource competition connection relationship refers to the correspondence established between different resource competition regions due to the resource competition relationship between live-streaming resources, used to represent the mutual influence between resource competition regions. Resource competition topology refers to the overall association structure formed based on resource competition connection relationships, used to characterize the resource competition organization relationship and competition propagation relationship between various resource competition regions. Competition influence degree refers to the strength of the resource competition influence of one resource competition region on other resource competition regions, used to reflect the influence capacity of a resource competition region in the overall resource competition. Topological influence sequence refers to the ranking result formed based on the competition influence degree, used to characterize the order of influence of each resource competition region in the resource competition topology. Resource competition grouping refers to a set of multiple resource competition regions formed based on the topological influence sequence, used to group resource competition regions with similar competition influence degrees into the same group, providing a resource organization basis for subsequent resource association matching and resource synergy capability analysis.

[0031] Preferred, Reference Figure 2 As shown, this diagram is an exemplary flowchart for determining resource collaboration capabilities according to the present application. In this embodiment, the resource collaboration capabilities are obtained by performing resource association matching on the resource competition groups based on the resource status of the edge nodes, which can be specifically implemented using the following steps: In step S21, the resource capability analysis of the edge node resource status is performed to obtain the resource capability parameters of the edge node; In step S22, the resource competition groups are associated and mapped according to the resource capability parameters to obtain resource matching results; In step S23, a collaborative correlation analysis is performed on the resource matching results to obtain the degree of resource collaboration of each resource competition group, thereby determining the resource collaboration capability.

[0032] In specific implementation, firstly, the processor resource usage, storage resource usage, cache resource usage, network bandwidth usage, and node service load in the edge node resource status are statistically analyzed separately, and the resource capacity that each resource can continue to provide services is calculated. Then, the resource capacity of each resource is summarized according to the edge node to form the resource status data corresponding to each edge node. The remaining processor resources, remaining storage resources, remaining cache resources, and remaining network bandwidth resources are statistically analyzed separately, summarized according to the edge node, and the number of live streams that each remaining resource can support is used as the resource capability parameter. Then, based on the resource competition groups, the composition of live streaming resources and the distribution of resource competition for each resource competition group are extracted one by one. The resource demand is determined according to the number of live streaming resources and the resource competition situation in each resource competition group. The resource demand of each resource competition group is then compared with the resource capability parameters of each edge node to determine whether each edge node meets the resource demand of the corresponding resource competition group. If the resource demand is met, an association relationship is established between the resource competition group and the corresponding edge node. All established association relationships are then uniformly organized to form the resource allocation relationship for each resource competition group, and the resource allocation relationship is used as the resource matching result. Finally, based on the resource matching results, the number of edge nodes corresponding to each resource competition group and the resource distribution of each edge node are counted one by one. It is determined whether multiple edge nodes can jointly undertake the tasks of caching, computing and data forwarding of the live broadcast resources corresponding to the same resource competition group. Then, the sum of the remaining amount of resources after each edge node undertakes the tasks and the amount of resources evenly distributed is taken as the resource coordination degree. Then, the resource coordination degree corresponding to all resource competition groups is uniformly sorted out to form the resource coordination analysis results of each resource competition group, and the resource coordination analysis results are taken as the resource coordination capability.

[0033] It should be noted that the resource capability parameter in this application refers to the resource carrying capacity index that the edge node can currently provide live streaming services, used to characterize the ability of the edge node to undertake live streaming resource scheduling tasks. The resource matching result refers to the resource allocation result formed after establishing a correspondence between resource competition groups and edge nodes, used to represent the resource configuration of edge nodes corresponding to each resource competition group. The resource coordination degree refers to the degree of cooperation between resources when multiple edge nodes jointly undertake the service task of the same resource competition group, used to characterize the strength of the ability of multiple edge nodes to collaboratively provide live streaming services. The resource coordination capability refers to a comprehensive capability index formed based on the resource coordination degree corresponding to each resource competition group, used to characterize the collaborative support capability of edge node resources for different resource competition groups, providing a resource coordination basis for subsequently establishing resource scheduling constraints based on the evolution trend of live streaming popularity.

[0034] In step S3, historical access records of the live streaming server are obtained, the evolution trend of live streaming popularity is determined through the historical access records, and the resource coordination capability is constrained and correlated through the evolution trend to obtain resource scheduling constraints.

[0035] In this embodiment, the historical access records of the live streaming server can be obtained in the following way: Obtain live streaming request data from the live streaming server over multiple consecutive scheduling periods, and aggregate the accesses according to the live streaming identifier to obtain the historical access sequence. The historical access sequence is sorted by time to obtain the historical access records.

[0036] In specific implementation, firstly, live request data stored over multiple consecutive scheduling cycles is read from the historical logs of the live streaming server, and the live stream identifier, request time, request source information, and request resource information corresponding to each live stream request are extracted. Then, live stream requests belonging to the same live streaming service are categorized and grouped according to the live stream identifier, ensuring that live stream requests generated by the same service in different scheduling cycles are centrally stored. Next, each type of live stream request is arranged sequentially according to the scheduling cycle, forming a data sequence that continuously reflects changes in live streaming service access, and this data sequence is used as the historical access sequence. Then, the corresponding request time information is extracted from each live stream request in the historical access sequence, and the requests are rearranged according to their chronological order, ensuring that the live stream requests are arranged continuously in the actual order of occurrence. Finally, the live stream requests after time arrangement are uniformly organized according to the live stream identifier, forming time-series data that completely reflects the historical access process of the live streaming service, and this time-series data is used as the historical access record.

[0037] It should be noted that the scheduling period in this application refers to the time interval during which the live streaming server performs one resource scheduling and records the corresponding operating status, ensuring that live streaming request data generated in different time periods has a unified statistical range. The historical access sequence refers to a data sequence formed by aggregating live streaming requests generated by the same live streaming service within multiple consecutive scheduling periods according to the live streaming identifier, used to characterize the continuous process of changes in live streaming service access. Historical access records refer to time-series data formed by sorting according to the actual occurrence time of live streaming requests, used to fully reflect the changes in access to the live streaming service at different time stages, providing a historical data foundation for subsequent analysis of live streaming popularity patterns and determination of live streaming popularity evolution trends.

[0038] Preferred, Reference Figure 3 As shown, this diagram is an exemplary flowchart for determining the evolution trend of live stream popularity according to the present application. In this embodiment, determining the evolution trend of live stream popularity through the historical access records can be achieved through the following steps: In step S31, the historical access records are segmented by time to obtain popularity change data; In step S32, continuous correlation analysis is performed based on the popularity change data to obtain the live broadcast popularity change trajectory; In step S33, trend fitting is performed on the trajectory of the live stream popularity change to obtain the evolution trend of the live stream popularity.

[0039] In specific implementation, firstly, the historical access records are divided into time periods according to scheduling cycles, and the number of live streaming requests, the number of users accessing the service, and the number of resource accesses are counted for each scheduling cycle. Then, the access statistics obtained within the same scheduling cycle are uniformly organized to form the access statistics results for that scheduling cycle. The number of live streaming requests, the number of users accessing the service, and the number of resource accesses are counted separately for each scheduling cycle, and the increase or decrease in the statistical results for adjacent scheduling cycles is compared to form the access change results for each scheduling cycle. These access change results are used as popularity change data. Next, the popularity change data is read sequentially according to time, and the popularity change data for adjacent scheduling cycles are compared item by item to analyze the direction and magnitude of changes in live streaming access activity between adjacent scheduling cycles. Finally, the change results between multiple consecutive scheduling cycles are connected in chronological order to form a change sequence that continuously reflects the process of live streaming popularity change, and this change sequence is used as the live streaming popularity change trajectory. Finally, the time nodes in the trajectory of the live stream popularity change are continuously fitted to form a continuous change curve between the time nodes; then, the overall change direction and change law of the live stream popularity over time are analyzed based on the continuous change curve, and the overall change direction and change law are taken as the evolution trend of the live stream popularity.

[0040] It should be noted that the popularity change data in this application refers to the statistical results of the live streaming service's access activity within each scheduling period, used to characterize the degree of attention received by the live streaming service within the corresponding scheduling period. The live streaming popularity change trajectory refers to the continuous change process formed by connecting the popularity change data corresponding to multiple scheduling periods in chronological order, used to characterize the development process of live streaming popularity over time. Trend fitting refers to the process of establishing a continuous change relationship based on the live streaming popularity change trajectory to reflect the overall change pattern of live streaming popularity, used to eliminate the impact of local fluctuations on the analysis of the overall change pattern. The evolution trend of live streaming popularity refers to the overall direction and pattern of change exhibited by live streaming popularity within multiple consecutive scheduling periods, used to characterize the development trend of live streaming service popularity changes over a future period, providing a trend basis for subsequently establishing resource scheduling constraints.

[0041] In this embodiment, the resource coordination capability is constrained and correlated through the evolutionary trend to obtain resource scheduling constraints, which can be specifically achieved in the following manner: The range of popularity variation for live streaming resources is determined based on the aforementioned live streaming popularity evolution trend; The resource coordination capability is categorized and constrained based on the heat variation range to obtain resource constraint results; The resource constraint results are correlated and corrected to obtain resource scheduling constraints.

[0042] In specific implementation, firstly, the changes in live streaming popularity corresponding to each live streaming resource are extracted based on the evolution trend of the live streaming popularity, and each live streaming resource is classified and organized according to the direction and magnitude of the changes in live streaming popularity; then, live streaming resources with similar changes in live streaming popularity are grouped into the same range of changes, forming multiple different ranges of changes in live streaming popularity; then, each range of changes in live streaming popularity is uniformly arranged according to the order of the changes in live streaming popularity, forming the range of changes in popularity corresponding to the live streaming resource, and the range of changes in popularity corresponding to the live streaming resource is taken as the range of changes in popularity. Then, the popularity change range corresponding to each live stream resource is read one by one, and the resource coordination status of the corresponding live stream resource in the resource coordination capability is obtained; then, the popularity change range corresponding to each live stream resource is associated with the corresponding resource coordination capability, and the constraint degree of the resource coordination capability is determined according to the popularity change range; for live stream resources with large popularity changes, the resource guarantee degree corresponding to the resource coordination capability is increased; for live stream resources with small popularity changes, the resource guarantee degree corresponding to the resource coordination capability is decreased; then, the resource coordination capabilities after constraint processing are uniformly organized to form the resource constraint results corresponding to each live stream resource, and the resource constraint results are used as the resource constraint results. Finally, the resource constraint results are correlated one by one to determine whether there is a resource competition relationship between different live stream resources and the impact of resource competition; if there is a resource competition relationship, the resource constraint results of the corresponding live stream resources are coordinated and adjusted in combination with the resource competition relationship to ensure that the live stream resources with resource competition relationships maintain a consistent resource constraint relationship; then, the adjusted constraint results of each live stream resource are uniformly organized to form the resource constraint set corresponding to the live stream resource, and the resource constraint set is used as the resource scheduling constraint.

[0043] It should be noted that the "popularity change range" in this application refers to different ranges of live stream popularity change formed based on the evolution trend of live stream popularity, used to characterize the popularity changes of live stream resources at different stages. "Hierarchical constraint" refers to the processing method of applying different levels of resource guarantee to resource coordination capabilities based on different popularity change ranges, used to maintain a correspondence between resource coordination capabilities and live stream popularity changes. "Resource constraint result" refers to the resource allocation result formed after hierarchical constraint processing of resource coordination capabilities, used to characterize the resource guarantee requirements corresponding to different live stream resources. "Association correction" refers to coordinating and adjusting the resource constraint result based on the resource competition relationship between live stream resources, ensuring that live stream resources with resource competition relationships maintain a consistent resource allocation relationship, used to reduce resource allocation conflicts. "Resource scheduling constraint" refers to the resource allocation restrictions formed based on the resource constraint result, used to constrain the scheduling process of subsequent edge node resources and streaming media transmission paths, ensuring that the resource scheduling process simultaneously conforms to resource competition relationships and the patterns of live stream popularity changes.

[0044] In step S4, collaborative optimization is performed based on the resource contention topology and the resource scheduling constraints to obtain a scheduling strategy for edge node resources and streaming media transmission paths.

[0045] In this embodiment, the scheduling strategy for edge node resources and streaming media transmission paths is obtained through collaborative optimization based on the resource contention topology and the resource scheduling constraints, which can be specifically implemented in the following manner: Determine the resource competition paths between edge nodes based on the resource competition topology; The resource scheduling path is obtained by coordinating the resource contention path through the resource scheduling constraints. Based on the resource scheduling path, the edge node resources and streaming media transmission path are jointly optimized to obtain a scheduling strategy for edge node resources and streaming media transmission path.

[0046] In practice, firstly, the resource competition groups in the resource competition topology can be traversed to read the resource competition connection relationships between the resource competition groups, and the corresponding edge nodes can be connected in sequence according to the resource competition connection relationships to form resource access association links between edge nodes. Then, the resource competition order and resource flow direction between the corresponding edge nodes can be counted for each resource access association link. If the live streaming resource corresponding to the first edge node and the live streaming resource corresponding to the second edge node have a resource competition relationship, and the live streaming resource corresponding to the second edge node and the live streaming resource corresponding to the third edge node also have a resource competition relationship, then the first edge node, the second edge node, and the third edge node can be connected in sequence to form a continuous resource competition path. Then, resource constraint information corresponding to each resource contention group can be read sequentially according to resource scheduling constraints, and the resource constraint information can be mapped one by one to each edge node in the resource contention path. For edge nodes with high resource contention and limited by resource scheduling constraints, the resource access order in the corresponding resource contention path is adjusted, and the resource access tasks are redistributed to edge nodes with lower resource contention. For resource contention paths that meet the resource scheduling constraints, the original resource access order remains unchanged. After adjusting all resource contention paths one by one, the adjusted resource access path is used as the resource scheduling path. Finally, the target edge node corresponding to each live stream can be determined according to the resource scheduling path, and the data transmission path that can reach the target edge node can be re-queried based on the redistributed edge node. The data transmission path with fewer nodes and lower current network resource consumption is selected as the new live stream transmission path. Then, the edge node resource configuration results are mapped to the corresponding streaming media transmission path to form the resource allocation scheme and data transmission scheme corresponding to the live resources, and the two are used together as the scheduling strategy.

[0047] It should be noted that, in this embodiment, the resource contention path refers to the connection path between edge nodes established based on the resource contention topology, used to represent the propagation direction of the resource contention relationship between different edge nodes. The resource scheduling path refers to the resource access path that meets resource scheduling constraints, used to represent the adjustment order and resource allocation relationship of edge node resources. The scheduling strategy refers to the execution plan formed by the edge node resource allocation method and the streaming media transmission path configuration method, used to guide the resource scheduling of the live stream between different edge nodes.

[0048] Furthermore, the high-concurrency live streaming scenario in this application refers to a business scenario where a large number of users simultaneously access the live streaming service, and the live streaming content is distributed jointly by the live streaming server, edge nodes, and network transmission resources. This scenario characterizes the resource scheduling environment of the live streaming service under conditions of rapid growth in user requests. Preferably, [the following is a reference to a specific scenario]. Figure 4As shown in the figure, this is a schematic diagram of a high-concurrency live streaming scenario provided in this application. The live streaming server receives live streaming requests initiated by user terminals and completes the generation, management, and distribution control of the live stream through edge nodes. Edge nodes refer to computing and caching devices deployed at the network edge, used to carry out live stream caching, resource allocation, and live data forwarding, thereby reducing the computing load on the live streaming server and shortening user access latency. User terminals include mobile terminals, tablet terminals, computer terminals, and smart TV terminals, which are terminal devices with live streaming viewing capabilities, used to send live streaming requests to the live streaming server and receive corresponding live stream data. The streaming media transmission link represents the data transmission path from the edge nodes to the user terminals, and can represent the transmission direction and transmission relationship of live data in the network.

[0049] In summary, the technical solution adopted in this application can achieve coordinated scheduling of resource competition and live streaming popularity evolution, thereby enhancing the dynamic adaptability of streaming media resource allocation in high-concurrency live streaming scenarios.

[0050] Example 2: This application provides a streaming media resource scheduling system for high-concurrency live streaming scenarios, referring to... Figure 5 As shown, this figure is a module structure diagram of a streaming media resource scheduling system for high-concurrency live streaming scenarios provided in this application. The scheduling system includes: The contention perception module 100 is used to collect live streaming request data, streaming media transmission status and edge node resource status of the live streaming server, and to construct resource contention relationships based on the live streaming request data and the streaming media transmission status. The collaborative analysis module 200 is used to partition the resource competition relationship, construct the resource competition topology and determine the resource competition group, and perform resource association matching on the resource competition group based on the resource status of the edge node to obtain resource collaboration capability. The trend constraint module 300 is used to obtain the historical access records of the live streaming server, determine the evolution trend of the live streaming popularity through the historical access records, and constrain the resource coordination capability through the evolution trend to obtain resource scheduling constraints. The collaborative scheduling module 400 is used to perform collaborative optimization based on the resource contention topology and the resource scheduling constraints to obtain a scheduling strategy for edge node resources and streaming media transmission paths.

[0051] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0052] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, including read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electrically-Erasable Programmable Read-Only Memory (EEPROM), compactdisc read-only memory (CD-ROM) or other optical disc storage, disk storage, magnetic tape storage, or any other computer-readable medium capable of carrying or storing data.

[0053] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

Claims

1. A method for scheduling streaming media resources in high-concurrency live streaming scenarios, characterized in that, The scheduling method includes the following steps: Collect live streaming request data, streaming media transmission status, and edge node resource status from the live streaming server, and construct resource competition relationships based on the live streaming request data and the streaming media transmission status; The resource competition relationship is partitioned into related regions, a resource competition topology is constructed and a resource competition group is determined, and resource association matching is performed on the resource competition group based on the resource status of the edge nodes to obtain resource coordination capabilities; Obtain historical access records of the live streaming server, determine the evolution trend of live streaming popularity through the historical access records, and constrain and correlate the resource coordination capability through the evolution trend to obtain resource scheduling constraints; Based on the resource contention topology and the resource scheduling constraints, a scheduling strategy for edge node resources and streaming media transmission paths is obtained through collaborative optimization.

2. The streaming media resource scheduling method in a high-concurrency live streaming scenario as described in claim 1, characterized in that, The collection of live streaming request data, streaming media transmission status, and edge node resource status from the live streaming server specifically includes: Collect the live streaming request information received by the live streaming server to obtain live streaming request data; Based on the live stream request data, obtain the transmission information of the corresponding live stream to obtain the streaming media transmission status; The edge nodes of the corresponding live stream are determined by the live stream request data, and the resource status of the edge nodes is collected to obtain the resource status of the edge nodes.

3. The streaming media resource scheduling method in a high-concurrency live streaming scenario as described in claim 1, characterized in that, Constructing resource contention relationships based on the live stream request data and the streaming media transmission status specifically includes: Based on the live stream request data, the degree of resource access overlap between live stream requests is statistically analyzed to obtain a request competition association set; Based on the analysis of the streaming media transmission status, the degree of link sharing and cache reuse between live streams is determined to obtain a resource sharing association set; The request competition association set and the resource sharing association set are fused together to obtain the resource competition relationship.

4. The streaming media resource scheduling method in a high-concurrency live streaming scenario as described in claim 1, characterized in that, The process of associating and partitioning the resource competition relationships, constructing a resource competition topology, and determining resource competition groups specifically includes: Based on the resource competition relationship, live streaming resources with similar levels of competitive correlation are aggregated to obtain multiple resource competition regions; Based on the resource competition relationships between the resource competition regions, resource competition connections are established, thereby constructing a resource competition topology; The degree of competition between resource competition regions is calculated using the resource competition topology to obtain a topological influence sequence; The resource competition topology is hierarchically divided by the topological influence sequence to obtain resource competition groups.

5. A streaming media resource scheduling method for high-concurrency live streaming scenarios as described in claim 1, characterized in that, Based on the resource status of the edge nodes, resource association matching is performed on the resource competition groups to obtain resource collaboration capabilities, specifically including: Resource capability analysis is performed on the resource status of the edge nodes to obtain the resource capability parameters of the edge nodes; Based on the resource capability parameters, the resource competition groups are associated and mapped to obtain resource matching results; The resource matching results are analyzed for collaboration to obtain the degree of resource collaboration in each resource competition group, thereby determining the resource collaboration capability.

6. The streaming media resource scheduling method in a high-concurrency live streaming scenario as described in claim 1, characterized in that, Obtaining the historical access records of the live streaming server specifically includes: Obtain live streaming request data from the live streaming server over multiple consecutive scheduling periods, and aggregate the accesses according to the live streaming identifier to obtain the historical access sequence. The historical access sequence is sorted by time to obtain the historical access records.

7. A streaming media resource scheduling method for high-concurrency live streaming scenarios as described in claim 1, characterized in that, Determining the evolution trend of live stream popularity through the aforementioned historical access records specifically includes: The historical access records are segmented by time to obtain popularity change data; Based on the aforementioned popularity change data, a continuous correlation analysis was performed to obtain the trajectory of live stream popularity changes; By performing trend fitting on the trajectory of the live stream popularity change, the evolution trend of the live stream popularity can be obtained.

8. A streaming media resource scheduling method for high-concurrency live streaming scenarios as described in claim 1, characterized in that, By constraining and associating the resource coordination capability through the evolutionary trend, the resource scheduling constraints specifically include: The range of popularity variation for live streaming resources is determined based on the aforementioned live streaming popularity evolution trend; The resource coordination capability is categorized and constrained based on the heat variation range to obtain resource constraint results; The resource constraint results are correlated and corrected to obtain resource scheduling constraints.

9. A streaming media resource scheduling method in a high-concurrency live streaming scenario as described in claim 1, characterized in that, Based on the resource contention topology and the resource scheduling constraints, a collaborative optimization is performed to obtain a scheduling strategy for edge node resources and streaming media transmission paths, specifically including: Determine the resource competition paths between edge nodes based on the resource competition topology; The resource scheduling path is obtained by coordinating the resource contention path through the resource scheduling constraints. Based on the resource scheduling path, the edge node resources and streaming media transmission path are jointly optimized to obtain a scheduling strategy for edge node resources and streaming media transmission path.

10. A streaming media resource scheduling system for high-concurrency live streaming scenarios, characterized in that, The scheduling system is used to execute a streaming media resource scheduling method in a high-concurrency live streaming scenario as described in any one of claims 1 to 9, the scheduling system comprising: The contention awareness module is used to collect live streaming request data, streaming media transmission status and edge node resource status from the live streaming server, and to construct resource contention relationships based on the live streaming request data and the streaming media transmission status. The collaborative analysis module is used to partition the resource competition relationship, construct the resource competition topology and determine the resource competition group, and perform resource association matching on the resource competition group based on the resource status of the edge node to obtain resource collaboration capability. The trend constraint module is used to obtain historical access records of the live streaming server, determine the evolution trend of the live streaming popularity through the historical access records, and constrain the resource coordination capability through the evolution trend to obtain resource scheduling constraints. The collaborative scheduling module is used to perform collaborative optimization based on the resource contention topology and the resource scheduling constraints to obtain a scheduling strategy for edge node resources and streaming media transmission paths.

Images