A service area weight coefficient updating method and device

Abstract

The embodiment of the present application provides a weight coefficient updating method and device for a service area. The technical scheme provided by the embodiment of the present application determines a candidate coefficient according to a weight coefficient for distributing services in a current area, determines a test coefficient based on set coefficient adjustment information, performs running test evaluation based on the test coefficient and the candidate coefficient to obtain a first evaluation result, updates the candidate coefficient and the coefficient adjustment information according to the first evaluation result, updates the test coefficient based on the updated coefficient adjustment information, and updates the weight coefficient for distributing services in the current area based on the updated candidate coefficient when a test end condition is met. According to the scheme, the candidate coefficient is determined according to the currently used weight coefficient, and the candidate coefficient is continuously updated according to the running test evaluation result of the test coefficient and the candidate coefficient. The updating of the weight coefficient does not need to rely on a large number of manual tests, and the weight coefficient updating efficiency of the service area is effectively improved.

Description

Technical Field

[0001] This application relates to the field of communication technology, and in particular to a method and apparatus for updating the weight coefficients of a service area. Background Technology

[0002] With the rapid development of multimedia services and the widespread adoption of audio and video communication software, real-time audio and video calls have been greatly promoted on the Internet.

[0003] To ensure call quality, multiple audio and video service zones are deployed globally. Different service zones support different network connections to suit specific user groups. The selection of the audio and video service zone for audio and video calls significantly impacts user experience. To determine the optimal service zone for each user, administrators pre-set weighting coefficients for service zone allocation (e.g., packet loss weighting, latency weighting, and jitter weighting). Network connectivity data (e.g., packet loss, latency, and jitter) between different service providers and service zones is obtained through network probing. Each zone is then scored based on the network connectivity data and the set weighting coefficients, thereby selecting the best service zone for users during audio and video calls.

[0004] The initial configuration of the weighting coefficients is selected by observing the quality of actual calls. Adjusting different weighting coefficients can affect the user's service area allocation, thereby affecting the user's access quality. Currently, weighting coefficients are generally updated manually based on network conditions. Each adjustment requires a test run to verify the effectiveness of the current weighting coefficients. Moreover, these tests take a long time to obtain sufficient and stable experimental data for a small user group, resulting in low efficiency in updating weighting coefficients. Summary of the Invention

[0005] This application provides a method and apparatus for updating the weight coefficient of a service area, which solves the technical problem in related technologies that the updating of weight coefficients relies on a large amount of manual testing and the updating efficiency of the weight coefficient of the service area is low. It effectively reduces the manual testing of weight coefficients and improves the updating efficiency of the weight coefficient of the service area.

[0006] In a first aspect, embodiments of this application provide a method for updating the weight coefficients of a service area, comprising:

[0007] Candidate coefficients are determined based on the weight coefficients of the current regional allocation service, and test coefficients are determined based on the set coefficient adjustment information. The weight coefficients include one or more combinations of packet loss weight coefficients, latency weight coefficients, and jitter weight coefficients.

[0008] Based on the test coefficients and the candidate coefficients, an operational test evaluation is performed to obtain a first evaluation result;

[0009] The candidate coefficients and the coefficient adjustment information are updated based on the first evaluation result, and the test coefficients are updated based on the updated coefficient adjustment information.

[0010] If the test termination conditions are met, the weight coefficients of the current region allocation service are updated based on the updated candidate coefficients.

[0011] In a second aspect, embodiments of this application provide a service area weight coefficient update device, including a coefficient setting module, an operation evaluation module, a coefficient adjustment module, and a coefficient update module, wherein:

[0012] The coefficient setting module is configured to determine candidate coefficients based on the weight coefficients of the current region allocation service, and to determine test coefficients based on the set coefficient adjustment information. The weight coefficients include one or more combinations of packet loss weight coefficients, latency weight coefficients, and jitter weight coefficients.

[0013] The operation evaluation module is configured to perform operation test evaluation based on the test coefficients and the candidate coefficients to obtain a first evaluation result;

[0014] The coefficient adjustment module is configured to update the candidate coefficients and the coefficient adjustment information according to the first evaluation result, and to update the test coefficients based on the updated coefficient adjustment information;

[0015] The coefficient update module is configured to update the weight coefficient of the current region allocation service based on the updated candidate coefficients when the test termination condition is met.

[0016] In a third aspect, embodiments of this application provide a service area weight coefficient update device, including: a memory and one or more processors;

[0017] The memory is used to store one or more programs;

[0018] When the one or more programs are executed by the one or more processors, the one or more processors implement the service area weight coefficient update method as described in the first aspect.

[0019] In a fourth aspect, embodiments of this application provide a non-volatile storage medium for storing computer-executable instructions, which, when executed by a computer processor, are used to perform the service area weight coefficient update method as described in the first aspect.

[0020] In a fifth aspect, embodiments of this application provide a computer program product comprising a computer program stored in a computer-readable storage medium, wherein at least one processor of the device reads from the computer-readable storage medium and executes the computer program, causing the device to perform the service area weight coefficient update method as described in the first aspect.

[0021] This application embodiment determines candidate coefficients based on the weight coefficients of the current regional service allocation, and determines test coefficients based on set coefficient adjustment information. A first evaluation result is obtained by performing an operational test based on the test coefficients and candidate coefficients. The candidate coefficients and coefficient adjustment information are updated based on the first evaluation result, and the test coefficients are updated based on the updated coefficient adjustment information. When the test termination condition is met, the weight coefficients of the current regional service allocation are updated based on the updated candidate coefficients. This solution determines candidate coefficients based on the currently used weight coefficients and continuously updates the candidate coefficients based on the operational test evaluation results of the test coefficients and candidate coefficients. The update of the weight coefficients does not rely on a large amount of manual testing, effectively improving the efficiency of weight coefficient updates for service areas. Attached Figure Description

[0022] Figure 1 This is a flowchart of a service area weight coefficient update method provided in an embodiment of this application;

[0023] Figure 2 This is a flowchart of another service area weight coefficient update method provided in an embodiment of this application;

[0024] Figure 3 This is a schematic diagram of a first evaluation result determination process provided in an embodiment of this application;

[0025] Figure 4 This is a schematic diagram illustrating the update process of candidate coefficients and coefficient adjustment information provided in an embodiment of this application;

[0026] Figure 5 This is a schematic diagram of the structure of a service area weight coefficient updating device provided in an embodiment of this application;

[0027] Figure 6 This is a schematic diagram of the structure of a service area weight coefficient update device provided in an embodiment of this application. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of this application clearer, specific embodiments of this application will be described in further detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely for explaining this application and not for limiting it. It should also be noted that, for ease of description, only the parts relevant to this application are shown in the drawings, not all of them. Before discussing exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although the flowcharts describe operations (or steps) as sequential processes, many of these operations can be performed in parallel, concurrently, or simultaneously. Furthermore, the order of the operations can be rearranged. The process can be terminated when its operation is completed, but additional steps not included in the drawings may also be present. The above processes can correspond to methods, functions, procedures, subroutines, subroutines, etc.

[0029] The service area weight coefficient update method provided in this application can be applied to service area (AV) allocation in audio and video calls (AV calls) to update the weight coefficients used for service area allocation. It aims to continuously update the candidate coefficients based on the evaluation results of operational tests using candidate and test coefficients in the region allocation service, thereby updating the weight coefficients efficiently. Traditional weight coefficient update methods typically require staff to manually adjust the weight coefficients based on actual user call data collected over a period of time and their experience. The updated weight coefficients are then tested in the region allocation service, such as performing A / B testing (AB testing, which compares the application of a new feature (updated weight coefficient) to a portion of AV audio and video call data (Group B) and the application of an old feature (original weight coefficient) to another portion of AV audio and video call data (Group A), comparing call metrics between the two groups to verify whether the new feature has a positive or negative impact on the call), to determine the effectiveness of the weight coefficient update. However, each weight coefficient update requires manual testing, resulting in low efficiency for service area weight coefficient updates. Based on this, an embodiment of this application provides a method for updating the weight coefficients of a service area to solve the technical problem of low efficiency in existing weight coefficient updates.

[0030] Figure 1 A flowchart of a service area weight coefficient update method provided in this application embodiment is given. The service area weight coefficient update method provided in this application embodiment can be executed by a service area weight coefficient update device, which can be implemented by hardware and / or software and integrated into the service area weight coefficient update equipment.

[0031] The following description uses the service area weight coefficient update device as an example to illustrate the method for updating the weight coefficients of a service area. (Reference) Figure 1 The method for updating the weight coefficients of this service area includes:

[0032] S101: Determine candidate coefficients based on the weight coefficients of the current regional allocation service, and determine test coefficients based on the set coefficient adjustment information. The weight coefficients include one or more combinations of packet loss weight coefficients, latency weight coefficients, and jitter weight coefficients.

[0033] The area allocation service is used to assign service areas to users who need to make audio / video calls (AV calls). Users conduct AV calls through the AV servers provided in the service areas. To ensure call quality, multiple service areas are deployed (e.g., globally), each supporting different network connections to accommodate different user groups. The service area allocation (i.e., service area routing) for AV calls significantly impacts the user's call experience. To determine the optimal service area for each user, the area allocation service typically uses network probing to obtain call performance metrics of the network connections between specific operators, network service providers (ISPs), countries, and various service areas. The process of selecting the optimal service area for a user in an AV call involves scoring each service area based on the call performance metrics obtained from network probing of different service areas (based on the user's operator, ISP, country, etc.) and selecting the service area with the highest score. Network probing can be done using PTM (Packet Train Map) probing, a method similar to ping, which consists of multiple probe packets sent from one access point to another; the probe is a series of data packets.

[0034] In one embodiment, call performance metrics may include one or more combinations of packet loss information, latency information, and jitter information. Different weighting coefficients can be configured for different call performance metrics, such as a packet loss weighting coefficient, a latency weighting coefficient, and a jitter weighting coefficient corresponding to one or more combinations of packet loss information, latency information, and jitter information. In one embodiment, different weighting coefficients can be configured for video calls and audio calls respectively. Optionally, a user's rating to a service area can be measured based on the following formula:

[0035] score = (α * packet_loss) 2 +(β*latency) 2 +(γ*jitter) 2

[0036] Here, packet_loss, latency, and jitter represent packet loss, latency, and jitter information, respectively, while α, β, and γ are the weighting coefficients for packet loss, latency, and jitter, respectively. It can be understood that adjusting different weighting coefficients can affect the service area allocated to a user under the same network conditions. For example, increasing the jitter weighting coefficient makes the impact of jitter on service area allocation take precedence over packet loss and latency, and the area allocation service will attempt to select the service area with the least jitter. Decreasing the jitter weighting coefficient will cause the area allocation service to prioritize the other two call performance metrics.

[0037] In one embodiment, service area allocation may also consider the price cost of the service area. After calculating a score for each possible service area for a call, these service areas are sorted according to their scores (lower scores are ranked higher), and only service areas with scores not exceeding a set cost threshold are retained. The remaining service areas are then sorted according to price cost, and the service area with the lowest price cost is selected as the target area for allocation to the user. That is, the service area with the highest call quality is selected first, and then the service area with the lowest price cost is chosen, achieving a balance between quality and cost in the final service area allocation result.

[0038] For example, the weighting coefficients currently used by the region allocation service when allocating service regions are determined, and initial candidate coefficients are determined based on these weighting coefficients. Optionally, the weighting coefficients of the current region allocation service can be used as candidate coefficients, that is, the initially configured candidate coefficients are consistent with the weighting coefficients configured by the region allocation service in the current production environment.

[0039] Furthermore, based on the set coefficient adjustment information, initial test coefficients are determined from the initial candidate coefficients. The coefficient adjustment information is used to adjust the initial candidate coefficients to obtain the initial test coefficients, or to update the test coefficients to obtain new test coefficients. Determining the test coefficients based on the set coefficient adjustment information can involve adjusting different weight coefficients, such as adjusting one or more combinations of weight coefficients including packet loss weight coefficients, latency weight coefficients, and jitter weight coefficients. Adjusting the weight coefficients can involve increasing or decreasing the weight coefficients.

[0040] S102: Perform runtime testing and evaluation based on test coefficients and candidate coefficients to obtain the first evaluation result.

[0041] For example, test coefficients and candidate coefficients are applied to the regional allocation service. The service area service will allocate service areas to users within a defined range based on the test coefficients and candidate coefficients, and record the call performance metrics corresponding to these users. This is used to conduct a test evaluation based on the test coefficients and candidate coefficients, resulting in a first evaluation result. This first evaluation result can be used to indicate whether the service area allocation effect of applying the test coefficients to the regional allocation service is better than that of applying the candidate coefficients. Optionally, if the test coefficients are better than (or significantly better than) the candidate coefficients, the first evaluation result can be considered to indicate that the attempt of this test coefficient was successful; if the candidate coefficients are better than (or significantly better than) the test coefficients, the first evaluation result can be considered to indicate that the attempt of this test coefficient was unsuccessful; or if the test coefficients are not significantly better than the candidate coefficients, the first evaluation result can be considered to indicate that the attempt of this test coefficient was neutral.

[0042] In one embodiment, the test evaluation can be based on A / B testing, which involves comparing a portion of the AV audio and video call data (Group B) to which the test coefficients are applied to the regional allocation service and another portion of the AV audio and video call data (Group A) to which the candidate coefficients are applied. The call metrics (call performance metrics) between Group A and Group B are compared to verify whether the impact of the test coefficients on the call is positive or negative, that is, to determine whether the regional allocation effect of the test coefficients applied to the regional allocation service is better than that of the candidate coefficients.

[0043] Optionally, the setting range provided by this solution can be a set range of countries or regions, which is smaller than the total service range served by the regional allocation service. That is, the operation test and evaluation can be carried out within a small part of the total service range served by the regional allocation service.

[0044] S103: Update the candidate coefficients and coefficient adjustment information based on the first evaluation results, and update the test coefficients based on the updated coefficient adjustment information.

[0045] For example, the candidate coefficients and coefficient adjustment information are updated based on the first evaluation result. For instance, if the first evaluation result indicates that the attempt to test the coefficient was successful, the adjustment direction indicated by the current coefficient adjustment information is considered appropriate, and the candidate coefficient can be replaced using the current test coefficient while maintaining the adjustment direction indicated by the coefficient adjustment information; if the first evaluation result indicates that the attempt to test the coefficient failed, the adjustment direction indicated by the current coefficient adjustment information is considered inappropriate, and the current candidate coefficient can be maintained while changing the adjustment direction indicated by the coefficient adjustment information; if the first evaluation result indicates that the attempt to test the coefficient was neutral, the current candidate coefficient and the adjustment direction indicated by the coefficient adjustment information can be maintained.

[0046] For example, you can set the adjustment direction indicated by the priority adjustment coefficient information. After trying all directions of a weight coefficient, you can try all directions of the next weight coefficient (for example, try them in the order of delay weight coefficient, packet loss weight coefficient, and jitter weight coefficient). After trying all directions of adjustment of various weight coefficients, you can further reduce the adjustment step size indicated by the coefficient adjustment information and try all directions of adjustment of various weight coefficients with smaller adjustment step sizes to obtain better candidate coefficients.

[0047] Furthermore, after updating the coefficient adjustment information, the test coefficients are updated according to the latest coefficient adjustment information. For example, if the adjustment direction and step size indicated by the coefficient adjustment information have not been adjusted, the current test coefficients are adjusted according to the previously determined adjustment direction and step size, thereby updating the test coefficients. However, if the adjustment direction and / or step size indicated by the coefficient adjustment information have been adjusted, the current test coefficients are adjusted according to the new adjustment direction and step size, thereby updating the test coefficients. The adjustment direction indicated by the coefficient adjustment information can be either increasing or decreasing the weight coefficient. Optionally, when the adjustment direction indicated by the coefficient adjustment information changes, for the weight coefficient type indicated by the coefficient adjustment information, adjustments can be made based on the weight coefficient type corresponding to the initially determined candidate coefficients (the weight coefficients for the current regional allocation service) to obtain new test coefficients. For weight coefficient types not indicated by the coefficient adjustment information, the currently determined values ​​for the corresponding weight coefficient types can be maintained.

[0048] After updating the candidate coefficients, coefficient adjustment information, and / or test coefficients, continue to run test evaluation based on the latest test coefficients and candidate coefficients to obtain the first evaluation result. Further update the candidate coefficients and coefficient adjustment information based on the first evaluation result, and update the test coefficients based on the updated coefficient adjustment information, until the test termination conditions set by this scheme are met.

[0049] S104: If the test termination condition is met, update the weight coefficients of the current region allocation service based on the updated candidate coefficients.

[0050] For example, when the test termination conditions are met, the weight coefficients of the current regional allocation service can be updated using candidate coefficients, essentially replacing the current regional allocation service's weight coefficients with candidate coefficients. It's understood that during test evaluation, if the candidate coefficients are found to be more effective than the current weight coefficients in regional allocation service, then manual adjustment of the weight coefficients and multiple test evaluations are unnecessary, effectively improving the efficiency and accuracy of weight coefficient updates. After updating the weight coefficients of the current regional allocation service, the updated weight coefficients are configured in the regional allocation service, which will then allocate applicable service areas to users requiring audio / video calls based on the updated weight coefficients.

[0051] The above-described method determines candidate coefficients based on the weight coefficients of the current regional service allocation, and determines test coefficients based on the set coefficient adjustment information. A first evaluation result is obtained by performing an operational test based on the test coefficients and candidate coefficients. The candidate coefficients and coefficient adjustment information are then updated based on the first evaluation result, and the test coefficients are updated based on the updated coefficient adjustment information. When the test termination condition is met, the weight coefficients of the current regional service allocation are updated based on the updated candidate coefficients. This solution determines candidate coefficients based on the currently used weight coefficients and continuously updates the candidate coefficients based on the operational test evaluation results of the test coefficients and candidate coefficients. The update of the weight coefficients does not rely on extensive manual testing, effectively improving the efficiency of weight coefficient updates for service areas.

[0052] Based on the above embodiments, Figure 2 A flowchart of another service area weight coefficient update method provided in this application embodiment is given, which is a specific implementation of the above-described service area weight coefficient update method. (Reference) Figure 2 The method for updating the weight coefficients of this service area includes:

[0053] S201: Based on the weight coefficients of the current regional allocation service, determine the candidate coefficients, and adjust the candidate coefficients according to the adjustment direction and adjustment step size corresponding to the set coefficient adjustment information to obtain the test coefficients.

[0054] The weighting coefficients provided in this solution include one or more combinations of packet loss weighting coefficients, latency weighting coefficients, and jitter weighting coefficients. The coefficient adjustment information includes the adjustment direction and adjustment step size. For example, an adjustment direction list can be initialized based on the type of weighting coefficient to be adjusted. This list records different adjustment directions for different weighting coefficients. When an adjustment direction in the coefficient adjustment information needs to be adjusted, the next adjustment direction can be determined according to the recorded order of the adjustment directions in the adjustment direction list.

[0055] In one embodiment, the adjustment directions recorded in the adjustment direction list are, in order: increase the delay weight coefficient, decrease the delay weight coefficient, increase the packet loss weight coefficient, decrease the packet loss weight coefficient, increase the jitter weight coefficient, and decrease the jitter weight coefficient. For example, if the current adjustment direction is to increase the delay weight coefficient, when it is necessary to change the adjustment direction, the updated adjustment direction will be to decrease the delay weight coefficient.

[0056] For example, the weight coefficient of the current region allocation service (hereinafter referred to as the current weight coefficient) is obtained and used as the initial candidate coefficient. An adjustment direction list is loaded, and the first adjustment direction in the list is used as the adjustment direction in the initial coefficient adjustment information. Optionally, the adjustment step size in the initial coefficient adjustment information can be set to a predetermined value, or it can be a value corresponding to a predetermined percentage of the weight coefficient of the current region allocation service (the initial candidate coefficient). For example, the adjustment step size in the initial coefficient adjustment information can be set to 20% of the weight coefficient of the current region allocation service (the initial candidate coefficient). At this time, the step size reduction count can be reset, meaning the step size reduction count is now 0.

[0057] Furthermore, according to the adjustment direction and step size corresponding to the aforementioned coefficient adjustment information, the candidate coefficients are adjusted to obtain the test coefficients. For example, initially, according to the adjustment direction of increasing the latency weight coefficient recorded in the coefficient adjustment information and the adjustment step size of 20% of the current weight coefficient, the latency weight coefficient in the aforementioned determined candidate coefficients is increased by 20%, while the packet loss weight coefficient and jitter weight coefficient remain unchanged, thus obtaining the test coefficients. This scheme can continuously update the test coefficients according to different adjustment directions and step sizes, searching for test coefficients with better service area allocation effects to update the candidate coefficients. It eliminates the need for staff to manually adjust the weight coefficients based on the timing and area allocation effects of the current weight coefficients, resulting in better weight coefficient update effects and higher update efficiency.

[0058] S202: Obtain the first test data of the test coefficients applied to the regional allocation service and the second test data of the candidate coefficients applied to the regional allocation service.

[0059] For example, each time a new test coefficient is obtained, it is applied to the region allocation service. Within a defined range, the service area service allocates service areas to users based on the test coefficient, records the call performance metrics for these users, and uses these call performance metrics as the first test data for applying the test coefficient to the region allocation service. Simultaneously, within the defined range, the service area service allocates service areas to users based on candidate coefficients, records the call performance metrics for these users, and uses these call performance metrics as the second test data for applying the candidate coefficients to the region allocation service.

[0060] In one embodiment, the call performance metrics corresponding to the test data (including the first test data and the second test data) provided by this solution include performance metric data of one or more performance types and cost metric data of one or more cost types. Optionally, the performance metric data may include one or more combinations of call duration, percentage of calls that failed (zero length), percentage of video calls that didn't show any video, number of times the server received at least one packet from the client, and number of times the client received at least one packet from the server. The cost metric data may include cost per minute and / or payment to cloud providers per minute. The higher the call duration, the number of times the server received at least one packet from the client, and the number of times the client received at least one packet from the server, the better the corresponding service area allocation effect. The lower the percentage of calls that failed, the percentage of video calls that didn't show any video, the cost per minute, and the payment to cloud providers per minute, the better the corresponding service area allocation effect.

[0061] S203: Based on the first test data and the second test data, perform a running test evaluation to obtain the first evaluation result.

[0062] For example, based on the first test data and the second test data, the effectiveness of applying the test coefficients and candidate coefficients to the regional allocation service is evaluated through operational testing, resulting in a first evaluation result. This solution collects the first test data and the second test data respectively, using the test coefficients and candidate coefficients to apply them to the regional allocation service, and performs operational testing to obtain a first evaluation result that accurately reflects the allocation of the regions corresponding to the test coefficients and candidate coefficients, thus improving the accuracy of weight coefficient updates.

[0063] In one possible embodiment, such as Figure 3As shown in the schematic diagram of the first evaluation result determination process, the weight coefficient update method provided in this solution, when performing a test evaluation based on the first test data and the second test data to obtain the first evaluation result, includes:

[0064] S2031: The first and second test data are denoised to obtain the third test data, wherein the third test data is routed to different service areas based on different weight coefficients.

[0065] S2032: Based on the third test data, conduct a running test evaluation to obtain the first evaluation result.

[0066] For example, the collected first and second test data are denoised to obtain the third test data. Optionally, the denoising process for the first and second test data can be to remove test data that are routed to the same service area under different weighting coefficients from the first and second test data, that is, to retain the test data that are routed to different service areas based on different weighting coefficients from the first and second test data, and use the retained test data as the third test data.

[0067] Furthermore, based on the third test data obtained after noise reduction, the test coefficients and candidate coefficients are evaluated through operational testing to obtain the first evaluation result. This scheme obtains the third test data by denoising the first and second test data, removing test data whose coefficient changes have little impact on the service area, and then conducts operational testing evaluation based on the third test data. This yields a first evaluation result that better reflects the differences in the service area allocation effect of the test coefficients and candidate coefficients when applied to the regional allocation service, thereby improving the accuracy of weight coefficient updates.

[0068] In one possible embodiment, the weight coefficient update method provided by this solution, when performing a test evaluation based on third test data and obtaining a first evaluation result, includes:

[0069] S20321: If the amount of data in the third test data is less than the first set threshold, and the total amount of data corresponding to the first test data and the second test data reaches the second set threshold, determine the first evaluation result that indicates neutrality.

[0070] S20322: Determine whether the third test data meets the negative significance condition based on the set first performance threshold and the set first cost threshold, and if the negative significance condition is met, determine the first evaluation result indicating failure.

[0071] S20323: If the negative significance condition is not met, determine whether the third test data meets the positive significance condition based on the first performance threshold and the first cost threshold, and if the positive significance condition is met, determine the first evaluation result indicating success.

[0072] S20324: If the positive significance condition is not met and the amount of data of the third test data reaches the third set threshold, determine whether the third test data meets the negative significance condition or whether the third test data meets the positive significance condition based on the set second performance threshold and the set second cost threshold. The second performance threshold is less than the first performance threshold and the second cost threshold is greater than the first cost threshold.

[0073] For example, after denoising the first test data and the second test data to obtain the third test data, it is first determined whether the amount of data in the third test data reaches a first set threshold, such as whether the amount of data for each call performance indicator in the third test data reaches the first set threshold.

[0074] If the amount of data in the third test data is less than the first set threshold (i.e., the number of noise-free data points is less than the first set threshold), it is further determined whether the total amount of data corresponding to the first test data and the second test data reaches the second set threshold. For example, it is determined whether the amount of data for each call performance indicator of the full data corresponding to the first test data and the second test data reaches the second set threshold. The second set threshold (e.g., 100,000) is greater than the first set threshold (e.g., 5,000).

[0075] If the amount of data in the third test data is less than the first set threshold, and the total amount of data corresponding to the first and second test data reaches the second set threshold, the call volume of the corresponding service area needs to be changed, and a neutral first evaluation result is determined. If the amount of data in the third test data is less than the first set threshold, but the total amount of data corresponding to the first and second test data does not reach the second set threshold, it is considered that more test data is needed for the run test evaluation, and test data continues to be collected without generating a first evaluation result. This solution, when the amount of data in the third test data is less than the first set threshold, determines whether a neutral first evaluation result is generated or more test data is needed based on whether the total amount of test data reaches the second set threshold. This accurately judges the run test evaluation result of the test coefficients, improves the accuracy of weight coefficient updates, and ensures that users receive a more suitable service area allocation quality.

[0076] In one embodiment, when the amount of third test data reaches a first set threshold, it is determined whether the third test data meets the negative significance condition based on a set first performance threshold and a set first cost threshold. If the third test data meets the negative significance condition, a first evaluation result indicating failure is determined. The negative significance condition can be understood as a criterion for judging whether the test coefficient is significant relative to the candidate coefficient in a negative direction. When the test coefficient is significant relative to the candidate coefficient in a negative direction, the negative significance condition is considered met. For example, if the negative standard deviation of a certain performance indicator data of the third test data (e.g., call failure rate, video call rate without video, call cost per minute, fee paid to the cloud provider per minute) is less than the corresponding threshold, the third test data is not significant relative to the candidate coefficient in a negative direction for that performance indicator data. However, if the negative standard deviation of a certain performance indicator data of the third test data reaches the corresponding threshold, the third test data is significant relative to the candidate coefficient in a negative direction for that performance indicator data. This solution accurately determines whether to generate a first evaluation result indicating failure based on whether the third test data meets the negative significance condition, accurately judges the running test evaluation result of the test coefficient, improves the update accuracy of the weight coefficients, and ensures that users receive a more suitable service area allocation quality.

[0077] Furthermore, if the third test data does not meet the negative significance condition, the system further determines whether the third test data meets the positive significance condition based on the set first performance threshold and the set first cost threshold. If the third test data meets the positive significance condition, the first evaluation result indicating success is determined. The positive significance condition can be understood as the criterion for judging whether the test coefficient is significant relative to the candidate coefficient in a positive direction. When the test coefficient is significant relative to the candidate coefficient in a positive direction, the positive significance condition is considered met. For example, if the positive standard deviation difference of a certain performance indicator data of the third test data (e.g., call duration, number of times the server receives at least one data packet from the client, number of times the client receives at least one data packet from the server) is less than the corresponding threshold, the third test data is not significant relative to the candidate coefficient in a positive direction for that performance indicator data. However, if the positive standard deviation difference of a certain performance indicator data of the third test data reaches the corresponding threshold, the third test data is significant relative to the candidate coefficient in a positive direction for that performance indicator data. This solution accurately determines whether to generate the first evaluation result indicating success based on whether the third test data meets the positive significance condition, accurately judges the running test evaluation result of the test coefficient, improves the update accuracy of the weight coefficients, and ensures that users receive a more suitable service area allocation quality.

[0078] If the third test data does not meet the positive significance condition, and the amount of the third test data reaches a third set threshold (e.g., 5000), then it is determined whether the third test data meets the negative significance condition based on the set second performance threshold and the set second cost threshold; or, if the third test data does not meet the negative significance condition, it is further determined whether the third test data meets the positive significance condition. Here, the second performance threshold provided by this scheme is lower than the first performance threshold, and the second cost threshold is higher than the first cost threshold.

[0079] If the third test data does not meet the positive significance condition, and the amount of the third test data is less than the third set threshold, then it is considered that more test data is needed for run-test evaluation, and test data collection continues, without generating a first evaluation result. If it cannot be determined whether the third test data meets the negative or positive significance condition based on the set second performance threshold and the set second cost threshold, the first evaluation result with neutral indications in each row is reviewed.

[0080] When the third test data does not meet the positive significance condition but the data volume reaches the third set threshold, this solution uses a smaller second performance threshold and a larger second cost threshold to determine whether the third test data meets the negative or positive significance condition. Under the condition of sufficient noise-free data volume, it accurately judges the running test evaluation results of the test coefficients, improves the update accuracy of the weight coefficients, and ensures that users receive a more suitable service area allocation quality.

[0081] The call performance metrics corresponding to the first and second test data provided in this solution include one or more performance metric data types and one or more cost metric data types. The performance metric data includes one or more combinations of call duration, call failure rate, video call rate without video, number of times the server receives at least one data packet from the client, and number of times the client receives at least one data packet from the server. The cost metric data includes call cost per minute and / or fees paid to the cloud provider per minute.

[0082] In one possible embodiment, the weight coefficient update method provided by this solution, when determining whether the third test data meets the negative significance condition based on a set first performance threshold and a set first cost threshold, includes: determining that the negative significance condition is met when there are performance index data in the third test data whose negative standard deviation difference reaches the set first performance threshold, or cost index data whose negative standard deviation difference reaches the set first cost threshold.

[0083] In one possible embodiment, the weight coefficient update method provided by this solution determines whether the third test data meets the positive significance condition based on the first performance threshold and the first cost threshold, including: if there are performance index data in the third test data with a positive standard deviation difference reaching the first performance threshold, or cost index data with a positive standard deviation difference reaching the first cost threshold, then the positive significance condition is met.

[0084] For example, the standard deviation difference corresponding to each performance index data in the third test data is calculated. For instance, for one performance index data, the difference between the average of the performance index data corresponding to the test coefficient and the average of the performance index data corresponding to the candidate coefficient in the third test data, as well as the standard deviation (sigma value) of the performance index data corresponding to the test coefficient and the candidate coefficient, is calculated. The ratio of the average difference to the standard deviation (sigma difference) is used as the standard deviation difference of the third test data on the corresponding performance index data. The standard deviation difference includes positive standard deviation difference (e.g., a positive standard deviation difference) and negative standard deviation difference (e.g., a negative standard deviation difference). It should be explained that performance index data with an absolute standard deviation difference below a corresponding set threshold are considered insignificant, while performance index data with an absolute standard deviation difference reaching the corresponding set threshold are considered significant. The success or failure of the test coefficient attempt can be determined based on whether the positive or negative standard deviation difference is significant.

[0085] In one embodiment, the negative standard deviation of the performance index data calculated in the third test data is compared with a set first performance threshold (e.g., the first performance threshold is set to 3.4), and the negative standard deviation of the cost index data calculated in the third test data is compared with a first cost threshold (e.g., the first cost threshold is set to 15.0). If there are performance index data in the third test data whose negative standard deviation reaches the set first performance threshold, or cost index data whose negative standard deviation reaches the set first cost threshold, then the negative significance condition is determined to be met; otherwise, the negative significance condition is considered not met.

[0086] If the negative significance condition is not met, the positive standard deviation differences of the performance index data calculated in the third test data are compared with the set first performance threshold, and the positive standard deviation differences of the cost index data calculated in the third test data are compared with the first cost threshold. If there are performance index data in the third test data whose positive standard deviation differences reach the first performance threshold, or cost index data whose positive standard deviation differences reach the first cost threshold, then the positive significance condition is determined to be met; otherwise, the positive significance condition is considered not met.

[0087] In one embodiment, when determining whether the third test data meets the negative significance condition or the positive significance condition based on the set second performance threshold and the set second cost threshold, it can be as follows: if there are performance index data in the third test data whose negative standard deviation difference reaches the set second performance threshold, or cost index data whose negative standard deviation difference reaches the set second cost threshold, the negative significance condition is determined to be met; if there are performance index data in the third test data whose positive standard deviation difference reaches the second performance threshold, or cost index data whose positive standard deviation difference reaches the second cost threshold, the positive significance condition is determined to be met.

[0088] This solution determines whether negative or positive significance conditions are met by comparing the standard deviation differences in the third test data with the set performance and cost thresholds. This accurately judges the running test evaluation results of the test coefficients, improves the update accuracy of the weight coefficients, and ensures that users receive a more suitable service area allocation quality.

[0089] S204: Update the candidate coefficients and coefficient adjustment information based on the first evaluation results, and update the test coefficients based on the updated coefficient adjustment information.

[0090] In one possible embodiment, such as Figure 4 The provided schematic diagram illustrates an update process for candidate coefficients and coefficient adjustment information. The weight coefficient update method provided in this solution, when updating candidate coefficients and coefficient adjustment information based on the first evaluation result, includes:

[0091] S2041: If the first evaluation result indicates neutrality, maintain the candidate coefficients and the adjustment direction and adjustment step size in the coefficient adjustment information.

[0092] S2042: If the first evaluation result indicates success, update the candidate coefficients using the test coefficients and maintain the adjustment direction and adjustment step size in the coefficient adjustment information.

[0093] S2043: If the first evaluation result indicates failure, maintain the candidate coefficients and update the coefficient adjustment information based on the set adjustment direction list. The adjustment directions recorded in the set adjustment direction list include the increase direction and decrease direction corresponding to different types of weight coefficients.

[0094] For example, when the first evaluation result of the instruction is neutral is obtained, if it is believed that the effect of applying the test coefficient to the service area allocation service is not significantly better than that of applying the candidate coefficient to the service area allocation service, then the current candidate coefficient and the adjustment direction and adjustment step size in the coefficient adjustment information are maintained. That is, when the test coefficient is updated next, the adjustment direction and adjustment step size used in the last update of the test coefficient will still be followed.

[0095] Upon receiving the first successful evaluation result, if it is determined that the test coefficients applied to the regional allocation service have a better effect on the regional allocation service than the candidate coefficients applied to the regional allocation service, then the candidate coefficients will be updated to the current test coefficients, and the adjustment direction and adjustment step size in the current coefficient adjustment information will be maintained.

[0096] Upon receiving the first evaluation result indicating failure, if it is determined that the application of the candidate coefficient to the regional allocation service is more effective than the application of the test coefficient to the regional allocation service, then the current candidate coefficient is maintained, and the coefficient adjustment information is updated based on the set adjustment direction list. That is, the adjustment direction and / or adjustment step size in the coefficient adjustment information are updated according to the adjustment direction recorded in the set adjustment direction list (including the increase direction and decrease direction corresponding to different types of weight coefficients).

[0097] This solution determines the update method for candidate coefficients and coefficient adjustment information based on the initial evaluation results. It can use a single, small adjustment step size to find better weight coefficients after multiple iterations while minimizing the impact on audio and video calls. This ensures that the impact on user quality is controllable during the adjustment of weight coefficients. It also tracks a set of core performance and cost metrics, uses statistical significance to verify A / B test results, runs small-scale A / B tests and tracks the results of each step size adjustment. This effectively optimizes call quality while reducing testing costs, and limits the number of attempts for test coefficients within an adjustment step size, effectively reducing the situation of being stuck for too long at a specific adjustment step size.

[0098] In one possible embodiment, the weight coefficient update method provided by this solution, when updating coefficient adjustment information based on a set adjustment direction list, includes: determining the next adjustment direction from the adjustment directions recorded in the set adjustment direction list; if the next adjustment direction is the first adjustment direction recorded in the set adjustment direction list, reducing the current adjustment step size according to a set step size reduction magnitude, until the number of step size reductions reaches a set number threshold.

[0099] For example, when it is necessary to update the coefficient adjustment information according to the adjustment direction list, the next adjustment direction corresponding to the current adjustment direction is determined from the multiple adjustment directions in the adjustment direction list, and the next adjustment direction is updated to the coefficient adjustment information.

[0100] If the next adjustment direction (i.e., the adjustment direction updated to the coefficient adjustment information) is the first adjustment direction recorded in the set adjustment direction list, it can be determined that all adjustment directions in the set adjustment direction list have been traversed under the current adjustment step size. Then, the current adjustment step size is reduced according to the set step size reduction margin (e.g., halving the adjustment step size each time), and the reduced adjustment step size is updated in the coefficient adjustment information. Optionally, when it is necessary to update the adjustment step size, it can be determined whether the number of step size reductions has reached a set threshold (e.g., 3 times). If the number of step size reductions reaches the set threshold, the attempt to test the coefficients is stopped, and the currently determined candidate coefficient is taken as the current optimal candidate coefficient for weight coefficient update. This scheme limits the possibility of excessive computation due to attempts with excessively small adjustment step sizes, thus affecting the efficiency of weight coefficient updates, by setting a threshold.

[0101] In one possible embodiment, after performing a test evaluation based on the test coefficients and candidate coefficients to obtain a first evaluation result, the weight coefficient update method provided by this solution further includes: recording the test coefficients in a historical test list. Correspondingly, after updating the test coefficients based on the updated coefficient adjustment information, the weight coefficient update method provided by this solution further includes: if the updated test coefficients are recorded in the historical test list, updating the coefficient adjustment information based on a set adjustment direction list, and updating the test coefficients again based on the updated coefficient adjustment information, until test coefficients not recorded in the historical test list are determined.

[0102] The historical test list records the test coefficients recorded in previous service area weight coefficient update processes. For example, when it's necessary to update the weight coefficients of a service area, the historical test list recorded in previous service area weight coefficient update processes can be loaded. After updating the test coefficients based on coefficient adjustment information, or after running test evaluations based on the test coefficients and candidate coefficients, the test coefficients obtained from this update or the run test evaluation are recorded in the historical test list. Furthermore, after updating the test coefficients based on coefficient adjustment information, it is determined whether the updated test coefficients are recorded in the historical test list. If the updated test coefficients are not recorded in the historical test list, run test evaluations can continue based on the test coefficients and candidate coefficients. If the updated test coefficients are recorded in the historical test list, the next coefficient adjustment information is updated based on the set adjustment direction list (i.e., the updated test coefficients are updated according to the current adjustment direction and adjustment step size to obtain the next updated test coefficients), and it continues to determine whether the test coefficients corresponding to the next coefficient adjustment information are recorded in the historical test list, until a test coefficient not recorded in the historical test list is determined. This solution records the tested coefficients in a historical test list, reducing redundant test evaluations, minimizing redundancy in weight coefficient update data processing, and improving the efficiency of weight coefficient updates.

[0103] S205: If the test termination condition is met, update the weight coefficients of the current region allocation service based on the updated candidate coefficients.

[0104] In one possible embodiment, the weight coefficient update method provided by this solution, when updating the weight coefficient of the current region allocation service based on the updated candidate coefficients under the condition that the test ends, includes: updating the weight coefficient of the current region allocation service based on the updated candidate coefficients when the number of step reductions of the adjustment step size corresponding to the coefficient adjustment information reaches a set number threshold.

[0105] For example, this solution determines whether the test termination condition is met based on the number of step reductions in the adjustment step size. Specifically, when the number of step reductions in the adjustment step size corresponding to the coefficient adjustment information reaches a set threshold (e.g., 3 times), the test termination condition is considered met, and the weight coefficients for the current region allocation service are updated based on the updated candidate coefficients. This solution sets the test termination condition by setting a threshold for the number of step reductions. When the adjustment step size becomes sufficiently small, it can be considered that the test coefficient side has approached the optimal effect. While ensuring the accuracy of the adjustment step size in different adjustment directions of the test coefficients, it reduces the impact of excessively small adjustment step sizes on the efficiency of weight coefficient updates, thus ensuring the efficiency of weight coefficient updates.

[0106] In one possible embodiment, the weight coefficient update method provided by this solution, when updating the weight coefficient of the current region allocation service based on the updated candidate coefficients, includes: performing a running test evaluation based on the updated candidate coefficients and the weight coefficient of the current region allocation service to obtain a second evaluation result, and updating the weight coefficient of the current region allocation service based on the second evaluation result.

[0107] For example, upon determining that the test termination condition is met, a second evaluation result can be obtained by running a test based on the updated candidate coefficients and the current weight coefficients of the regional allocation service. If a second evaluation result indicating failure or neutrality is obtained, the current weight coefficients of the regional allocation service are maintained. Conversely, if a second evaluation result indicating success is obtained, the current weight coefficients of the regional allocation service are replaced with the updated candidate coefficients to update the weight coefficients of the current regional allocation service and achieve a better service region allocation effect. This solution, after determining the optimal candidate coefficients, further conducts a run-test evaluation between the last updated candidate coefficients and the current weight coefficients of the regional allocation service to determine whether the candidate coefficients are superior to the current weight coefficients, ensuring the service region allocation effect after the weight coefficients are updated. Optionally, the run-test evaluation based on the updated candidate coefficients and the current weight coefficients of the regional allocation service can be performed automatically or manually by staff. By manually conducting A / B testing on the last updated candidate coefficients and the current weight coefficients for the regional allocation service, it can be effectively ensured that the service region allocation effect of the candidate coefficients is better than that of the weight coefficients used in the current production environment. Staff only need to verify the best candidate coefficients (e.g., manually verify them once at a set time interval), eliminating the need for staff to conduct frequent A / B testing, reducing manual operation and labor costs, improving the efficiency of weight coefficient updates, and improving the service region allocation effect.

[0108] The above-described method determines candidate coefficients based on the current regional service allocation weight coefficients, and determines test coefficients based on predefined coefficient adjustment information. A first evaluation result is obtained through runtime testing using the test coefficients and candidate coefficients. The candidate coefficients and coefficient adjustment information are then updated based on the first evaluation result, and the test coefficients are updated again based on the updated coefficient adjustment information. When the test termination condition is met, the weight coefficients for the current regional service allocation are updated based on the updated candidate coefficients. This solution determines candidate coefficients based on the currently used weight coefficients and continuously updates them based on the runtime testing evaluation results of the test coefficients and candidate coefficients. The update of weight coefficients does not rely on extensive manual testing, effectively improving the efficiency of service area weight coefficient updates. Simultaneously, noise reduction processing is applied to the first and second test data to obtain third test data, removing test data with minimal impact on service areas due to coefficient changes, thus improving the accuracy of weight coefficient updates. Furthermore, the runtime testing evaluation results of the test coefficients are judged based on the standard deviation differences in the third test data and the comparison with predefined performance and cost thresholds, further improving the accuracy of weight coefficient updates and ensuring users receive more suitable service area allocation quality.

[0109] Figure 5 This is a schematic diagram of a service area weight coefficient updating device provided in an embodiment of this application. (Reference) Figure 5The weight coefficient update device for the service area includes a coefficient setting module 51, an operation evaluation module 52, a coefficient adjustment module 53, and a coefficient update module 54.

[0110] The coefficient setting module 51 is configured to determine candidate coefficients based on the weight coefficients of the current region allocation service, and to determine test coefficients based on the set coefficient adjustment information. The weight coefficients include one or more combinations of packet loss weight coefficients, latency weight coefficients, and jitter weight coefficients. The operation evaluation module 52 is configured to perform operation test evaluation based on the test coefficients and candidate coefficients to obtain a first evaluation result. The coefficient adjustment module 53 is configured to update the candidate coefficients and coefficient adjustment information according to the first evaluation result, and to update the test coefficients based on the updated coefficient adjustment information. The coefficient update module 54 is configured to update the weight coefficients of the current region allocation service based on the updated candidate coefficients when the test termination conditions are met.

[0111] The above-described method determines candidate coefficients based on the weight coefficients of the current regional service allocation, and determines test coefficients based on the set coefficient adjustment information. A first evaluation result is obtained by performing an operational test based on the test coefficients and candidate coefficients. The candidate coefficients and coefficient adjustment information are then updated based on the first evaluation result, and the test coefficients are updated based on the updated coefficient adjustment information. When the test termination condition is met, the weight coefficients of the current regional service allocation are updated based on the updated candidate coefficients. This solution determines candidate coefficients based on the currently used weight coefficients and continuously updates the candidate coefficients based on the operational test evaluation results of the test coefficients and candidate coefficients. The update of the weight coefficients does not rely on extensive manual testing, effectively improving the efficiency of weight coefficient updates for service areas.

[0112] In one possible embodiment, when determining the test coefficient based on the set coefficient adjustment information, the coefficient setting module 51 is configured to adjust the candidate coefficients according to the adjustment direction and adjustment step size corresponding to the set coefficient adjustment information to obtain the test coefficients.

[0113] In one possible embodiment, the evaluation module 52 is configured as follows:

[0114] Obtain the first test data of the test coefficients applied to the regional allocation service and the second test data of the candidate coefficients applied to the regional allocation service;

[0115] Based on the first and second test data, the running test evaluation was conducted to obtain the first evaluation result.

[0116] In one possible embodiment, when the runtime evaluation module 52 performs a runtime test evaluation based on the first test data and the second test data to obtain a first evaluation result, it is configured as follows:

[0117] The first and second test data are denoised to obtain the third test data, in which the third test data is routed to different service areas based on different weight coefficients;

[0118] Based on the third test data, the running test evaluation was conducted to obtain the first evaluation result.

[0119] In one possible embodiment, when the runtime evaluation module 52 performs a runtime test evaluation based on the third test data and obtains a first evaluation result, it is configured as follows:

[0120] If the amount of data in the third test data is less than the first set threshold, and the total amount of data corresponding to the first test data and the second test data reaches the second set threshold, then a neutral first evaluation result is determined.

[0121] Based on the set first performance threshold and the set first cost threshold, determine whether the third test data meets the negative significance condition, and if the negative significance condition is met, determine the first evaluation result indicating failure;

[0122] If the negative significance condition is not met, the third test data is determined to be positively significant based on the first performance threshold and the first cost threshold. If the positive significance condition is met, the first evaluation result indicating success is determined.

[0123] If the positive significance condition is not met, and the amount of data in the third test data reaches the third set threshold, the system determines whether the third test data meets the negative significance condition or whether the third test data meets the positive significance condition based on the set second performance threshold and the set second cost threshold. The second performance threshold is less than the first performance threshold, and the second cost threshold is greater than the first cost threshold.

[0124] In one possible embodiment, the first test data and the second test data include performance indicator data of one or more performance types and cost indicator data of one or more cost types. The performance indicator data includes one or more of the following combinations: call duration, call failure rate, video call rate without video, number of times the server receives at least one data packet from the client, and number of times the client receives at least one data packet from the server. The cost indicator data includes call cost per minute and / or payment to cloud provider per minute.

[0125] When the evaluation module 52 determines whether the third test data meets the negative significance condition based on the set first performance threshold and the set first cost threshold, it is configured as follows:

[0126] If there are performance index data in the third test data whose negative standard deviation difference reaches the set first performance threshold, or cost index data whose negative standard deviation difference reaches the set first cost threshold, then the negative significance condition is determined to be met.

[0127] In one possible embodiment, the first test data and the second test data include performance index data of one or more performance types and cost index data of one or more cost types;

[0128] When the evaluation module 52 determines whether the third test data meets the positive significance condition based on the first performance threshold and the first cost threshold, it is configured as follows:

[0129] If there are performance index data in the third test data that have a positive standard deviation difference reaching the first performance threshold, or cost index data that have a positive standard deviation difference reaching the first cost threshold, then the positive significance condition is determined to be met.

[0130] In one possible embodiment, when the coefficient adjustment module 53 updates the candidate coefficients and coefficient adjustment information based on the first evaluation result, it is configured as follows:

[0131] If the first evaluation result indicates neutrality, the candidate coefficients and the adjustment direction and adjustment step size in the coefficient adjustment information shall be maintained.

[0132] If the first evaluation result indicates success, the candidate coefficients are updated using the test coefficients, while maintaining the adjustment direction and adjustment step size in the coefficient adjustment information;

[0133] If the first evaluation result indicates failure, the candidate coefficients are maintained, and the coefficient adjustment information is updated based on the set adjustment direction list. The adjustment directions recorded in the set adjustment direction list include the increase direction and decrease direction corresponding to different types of weight coefficients.

[0134] In one possible embodiment, when the coefficient adjustment module 53 updates the coefficient adjustment information based on the set adjustment direction list, it is configured as follows:

[0135] Select the next adjustment direction from the adjustment directions recorded in the set adjustment direction list;

[0136] If the next adjustment direction is the first adjustment direction recorded in the set adjustment direction list, reduce the current adjustment step size according to the set step size reduction range until the number of step size reductions reaches the set number threshold.

[0137] In one possible embodiment, the weight coefficient update device further includes a history record module, which is used to record the test coefficients in a history test list after running test evaluation based on test coefficients and candidate coefficients and obtaining a first evaluation result.

[0138] After updating the test coefficients based on the updated coefficient adjustment information, the coefficient adjustment module 53 is further configured to: update the coefficient adjustment information based on the set adjustment direction list if the updated test coefficients are recorded in the historical test list, and update the test coefficients again based on the updated coefficient adjustment information until test coefficients not recorded in the historical test list are determined.

[0139] In one possible embodiment, the coefficient update module 54 is configured to update the weight coefficient of the current region allocation service based on the updated candidate coefficients when the number of step reductions of the adjustment step corresponding to the coefficient adjustment information reaches a set threshold.

[0140] In one possible embodiment, when the coefficient update module 54 updates the weight coefficient of the current region allocation service based on the updated candidate coefficients, it is configured to: perform a running test evaluation based on the updated candidate coefficients and the weight coefficient of the current region allocation service to obtain a second evaluation result, and update the weight coefficient of the current region allocation service based on the second evaluation result.

[0141] It is worth noting that in the embodiments of the weight coefficient update device for the service area described above, the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of each functional unit are only for easy differentiation and are not used to limit the protection scope of the embodiments of the present invention.

[0142] This application also provides a service area weight coefficient update device, which can integrate the service area weight coefficient update apparatus provided in this application. Figure 6 This is a schematic diagram of the structure of a service area weight coefficient updating device provided in an embodiment of this application. (Reference) Figure 6 The service area weight coefficient update device includes: an input device 63, an output device 64, a memory 62, and one or more processors 61; the memory 62 is used to store one or more programs; when one or more programs are executed by one or more processors 61, the one or more processors 61 implement the service area weight coefficient update method as provided in the above embodiments. The service area weight coefficient update device, equipment, and computer provided above can be used to execute the service area weight coefficient update method provided in any of the above embodiments, and have corresponding functions and beneficial effects.

[0143] This application also provides a non-volatile storage medium storing computer-executable instructions, which, when executed by a computer processor, are used to perform the service area weight coefficient update method provided in the above embodiments. Of course, the computer-executable instructions provided in this application are not limited to the service area weight coefficient update method provided above; they can also perform related operations in the service area weight coefficient update method provided in any embodiment of this application. The service area weight coefficient update apparatus, device, and storage medium provided in the above embodiments can execute the service area weight coefficient update method provided in any embodiment of this application. Technical details not described in detail in the above embodiments can be found in the service area weight coefficient update method provided in any embodiment of this application.

[0144] Based on the above embodiments, this application also provides a computer program product. The technical solution of this application, in essence or in other words, the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer program product is stored in a storage medium and includes several instructions to cause a computer device, mobile terminal, or processor therein to execute all or part of the steps of the service area weight coefficient update method provided in the various embodiments of this application.

Claims

1. A method of updating weight coefficients of a service area, characterized by, The method comprises: determining a candidate coefficient based on a weight coefficient of a current regional distribution service, and determining a test coefficient based on set coefficient adjustment information, wherein the weight coefficient comprises a combination of one or more of a packet loss weight coefficient, a delay weight coefficient, and a jitter weight coefficient; obtaining first test data in which the test coefficient is applied to the regional distribution service, and second test data in which the candidate coefficient is applied to the regional distribution service; performing a running test evaluation based on the first test data and the second test data to obtain a first evaluation result; updating the candidate coefficient and the coefficient adjustment information according to the first evaluation result, and updating the test coefficient based on the updated coefficient adjustment information; in a case where a test end condition is met, updating the weight coefficient of the current regional distribution service based on the updated candidate coefficient.

2. The service area weight coefficient update method of claim 1, wherein, The method comprises: adjusting the candidate coefficient according to an adjustment direction and an adjustment step corresponding to the set coefficient adjustment information to obtain the test coefficient.

3. The service area weight coefficient update method of claim 1, wherein, The method comprises: performing denoising processing on the first test data and the second test data to obtain third test data, wherein the third test data is routed to different service regions based on different weight coefficients; performing a running test evaluation based on the third test data to obtain a first evaluation result.

4. The service area weight coefficient update method of claim 3, wherein, The method comprises: in a case where a data amount of the third test data is less than a first set threshold, and a total data amount corresponding to the first test data and the second test data reaches a second set threshold, determining a neutral first evaluation result; determining whether the third test data meets a negative significant condition based on a first performance threshold and a first cost threshold, and in a case where the negative significant condition is met, determining a failed first evaluation result; in a case where the negative significant condition is not met, determining whether the third test data meets a positive significant condition based on the first performance threshold and the first cost threshold, and in a case where the positive significant condition is met, determining a successful first evaluation result; in a case where the positive significant condition is not met, and the data amount of the third test data reaches a third set threshold, determining whether the third test data meets a negative significant condition or determining whether the third test data meets a positive significant condition based on a second performance threshold and a second cost threshold, wherein the second performance threshold is less than the first performance threshold, and the second cost threshold is greater than the first cost threshold.

5. The service area weight coefficient update method of claim 4, wherein, The first test data and the second test data include performance indicator data of one or more performance types and cost indicator data of one or more cost types, the performance indicator data includes one or more combinations of call duration, call failure ratio, video call ratio without video, number of times that the server receives at least one data packet from the client, and number of times that the client receives at least one data packet from the server, and the cost indicator data includes per-minute call cost and / or per-minute payment cost to the cloud provider; The determination of whether the third test data meets the negative significant condition based on the set first performance threshold and the set first cost threshold includes: In a case where there is performance indicator data with a negative standard deviation difference reaching the set first performance threshold or cost indicator data with a negative standard deviation difference reaching the set first cost threshold in the third test data, it is determined that the negative significant condition is met.

6. The service area weight coefficient update method of claim 4, wherein, The first test data and the second test data include performance indicator data of one or more performance types and cost indicator data of one or more cost types; The determination of whether the third test data meets the positive significant condition based on the first performance threshold and the first cost threshold includes: In a case where there is performance indicator data with a positive standard deviation difference reaching the first performance threshold or cost indicator data with a positive standard deviation difference reaching the first cost threshold in the third test data, it is determined that the positive significant condition is met.

7. The method of claim 1, wherein, The updating of the candidate coefficient and the coefficient adjustment information according to the first evaluation result includes: In a case where the first evaluation result indicates neutrality, the adjustment direction and the adjustment step in the candidate coefficient and the coefficient adjustment information are maintained; In a case where the first evaluation result indicates success, the candidate coefficient is updated by using the test coefficient, and the adjustment direction and the adjustment step in the coefficient adjustment information are maintained; In a case where the first evaluation result indicates failure, the candidate coefficient is maintained, and the coefficient adjustment information is updated based on a set adjustment direction list, the adjustment directions recorded in the set adjustment direction list include increasing directions and decreasing directions corresponding to different types of weight coefficients.

8. The service area weight coefficient update method of claim 7, wherein, The updating of the coefficient adjustment information based on the set adjustment direction list includes: A next adjustment direction is determined from the adjustment directions recorded in the set adjustment direction list; In a case where the next adjustment direction is the first adjustment direction recorded in the set adjustment direction list, a current adjustment step is reduced by a set step reduction amplitude until a step reduction number reaches a set number threshold.

9. The service area weight coefficient update method of claim 7, wherein, After the first evaluation result is obtained, the method further includes: The test coefficient is recorded in a historical test list; After the test coefficient is updated based on the updated coefficient adjustment information, the method further includes: In a case where the updated test coefficient is recorded in the historical test list, the coefficient adjustment information is updated based on a set adjustment direction list, and the test coefficient is again updated based on the updated coefficient adjustment information until a test coefficient not recorded in the historical test list is determined.

10. The method of claim 1, wherein, The weight coefficient of the current region distribution service is updated based on the updated candidate coefficient in a case where a test end condition is met. In a case where the step reduction number of the adjustment step corresponding to the coefficient adjustment information reaches a set number threshold, the weight coefficient of the current region distribution service is updated based on the updated candidate coefficient.

11. The method of claim 1, wherein, The weight coefficient of the current region distribution service is updated based on the updated candidate coefficient, including: A second evaluation result is obtained by performing running test evaluation based on the updated candidate coefficient and the weight coefficient of the current region distribution service, and the weight coefficient of the current region distribution service is updated based on the second evaluation result.

12. A service area weight coefficient updating apparatus characterized by comprising: The coefficient setting module is configured to determine a candidate coefficient based on the weight coefficient of the current region distribution service, and determine a test coefficient based on the set coefficient adjustment information, the weight coefficient including one or more combinations of a packet loss weight coefficient, a delay weight coefficient, and a jitter weight coefficient. The running evaluation module is configured to obtain first test data in which the test coefficient is applied to the region distribution service and second test data in which the candidate coefficient is applied to the region distribution service, perform running test evaluation based on the first test data and the second test data, and obtain a first evaluation result. The coefficient adjustment module is configured to update the candidate coefficient and the coefficient adjustment information according to the first evaluation result, and update the test coefficient based on the updated coefficient adjustment information. The coefficient updating module is configured to update the weight coefficient of the current region distribution service based on the updated candidate coefficient in a case where a test end condition is met. The computer program is executed by the processor to implement the weight coefficient updating method of the service region according to any one of claims 1-11.

13. A service area weight coefficient updating apparatus characterized by comprising: The computer program is executed by the processor to implement the weight coefficient updating method of the service region according to any one of claims 1-11. The computer program is executed by the processor to implement the weight coefficient updating method of the service region according to any one of claims 1-11. ​ ​ 14. A non-volatile storage medium for storing computer-executable instructions, characterized in that, ​ 15. A computer program product comprising a computer program, characterized in that, ​

Images