Call center capacity evaluation method and device and related equipment

Abstract

The embodiment of the invention provides a call center capacity evaluation method, a call center capacity evaluation device and related equipment. In the embodiment of the invention, the maximum audio concurrency support capability of a single server and the resource consumption ratio of video to audio can be obtained, and the video traffic proportion of a call center and the call duration ratio of the video to the audio can be obtained; determining the maximum concurrency support capability of a single server under the mixed telephone traffic through a first calculation model, and predicting the total concurrency demand of the call center under the mixed telephone traffic through a second calculation model according to the number of calls per second, the audio call duration, the video telephone traffic proportion and the call duration proportion of the call center; and based on the total concurrency demand and the maximum concurrency support capability of the single server under the mixed telephone traffic, determining a preliminary server demand quantity of the call center, and based on the performance difference correction factor, correcting the preliminary server demand quantity to obtain the corrected server demand quantity of the call center.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to a method, apparatus and related equipment for assessing call center capacity. Background Technology

[0002] With the popularization and commercial application of 5G technology, the business model of call centers has evolved from traditional voice calls to multimedia services such as video calls and real-time audio and video interaction. Currently, call center capacity assessment mainly relies on stress testing methods, which simulate high-concurrency call scenarios to test the server's processing capacity. Common stress testing tools include SIPP and LoadRunner, among which SIPP is widely used for stress testing of call center media due to its open-source, flexible, and high-performance characteristics.

[0003] Taking the widely used SIPP tool as an example, its implementation is relatively simple and direct in pure audio call scenarios: load testing is used to obtain the maximum concurrent audio calls a single server can support under a specific configuration, and then the required server scale is estimated proportionally based on business needs. For example, if a single server can support 800 concurrent audio calls, then only about 13 servers are needed to support 10,000 calls. Furthermore, to handle more complex business scenarios, some dynamic server adjustment solutions have emerged in existing technologies. For example, comparative document 1 (CN116938724A) discloses a method for scaling up and down servers in audio and video conferencing, which dynamically adjusts the number of server connections or cloud storage capacity by monitoring the server's actual memory usage and the number of concurrent conference requests. Another existing technology, such as comparative document 2 (CN114040192A), discloses a load testing method for audio and video conferencing. This method simulates multi-conference concurrent scenarios by calling audio and video interfaces and obtains interface load test data and server performance data to generate a load test report.

[0004] However, when dealing with mixed audio and video call scenarios, the relevant technologies generally suffer from problems such as inaccurate capacity assessment and unreasonable resource planning due to the static and singular evaluation model and the deviation between stress test data and the production environment. Summary of the Invention

[0005] This application provides a call center capacity assessment method, apparatus, and equipment to address the common problems in related technologies when dealing with mixed audio and video call scenarios, such as inaccurate capacity assessment and unreasonable resource planning caused by static and singular assessment models and deviations between stress test data and the production environment.

[0006] This application provides a call center capacity assessment method, including: Based on the stress test data of the call center, the maximum audio concurrency support capacity of a single server and the resource consumption ratio of video and audio are obtained. Based on the business data of the call center, the video call volume ratio and the video and audio call duration ratio of the call center are obtained. Based on the maximum audio concurrency support capacity, the video call volume ratio, and the resource consumption ratio, the maximum concurrency support capacity of a single server under mixed call volume is determined by a first calculation model. Based on the number of calls per second of the call center, the audio call duration, the video call volume ratio, and the call duration ratio, the total concurrency demand of the call center under mixed call volume is predicted by a second calculation model. Based on the total concurrent demand and the maximum concurrent support capacity of a single server under mixed call traffic, the initial server demand of the call center is determined, and the initial server demand is corrected based on a performance difference correction factor to obtain the corrected server demand of the call center; wherein, the performance difference correction factor is determined by historical data analysis or linear regression methods based on the hardware performance difference between the production environment and the stress test environment.

[0007] This application embodiment also provides a call center capacity assessment device, including: The acquisition module is used to acquire the maximum audio concurrency support capacity of a single server and the resource consumption ratio of video and audio based on the stress test data of the call center, and to acquire the video call traffic ratio and the video and audio call duration ratio of the call center based on the business data of the call center. The prediction module is used to determine the maximum concurrent support capacity of a single server under mixed traffic based on the maximum audio concurrent support capacity, the video call volume ratio, and the resource consumption ratio through a first calculation model, and to predict the total concurrent demand of the call center under mixed traffic based on the number of calls per second of the call center, the audio call duration, the video call volume ratio, and the call duration ratio through a second calculation model. The evaluation module is used to determine the initial server requirements of the call center based on the total concurrent demand and the maximum concurrent support capacity of a single server under mixed call traffic, and to correct the initial server requirements based on a performance difference correction factor to obtain the corrected server requirements of the call center; wherein, the performance difference correction factor is determined by historical data analysis or linear regression methods based on the hardware performance differences between the production environment and the stress test environment.

[0008] This application also provides an electronic device, including: a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. The processor communicates with the memory via the bus. When the machine-readable instructions are executed by the processor, they perform the steps in the call center capacity assessment method provided in this application.

[0009] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to implement the steps in the call center capacity assessment method provided in this application.

[0010] This application also provides a computer program product that stores instructions that, when executed by a computer, cause the computer to perform the steps in the call center capacity assessment method provided in this application.

[0011] The call center capacity assessment method provided in this application first conducts targeted audio, video, and mixed call load tests on the target call center, and analyzes its actual business data to obtain the baseline performance parameters of a single server, namely the maximum audio concurrency support capacity and audio / video resource consumption ratio, as well as core business characteristic parameters, namely the video call ratio and audio / video call duration ratio. Based on these parameters, on the one hand, a first calculation model is used to correlate and calculate the baseline single-machine capacity, video call ratio, and resource consumption differences to accurately determine the actual carrying capacity limit of a single server under mixed call loads; on the other hand, a second calculation model is used to accurately predict the total concurrent demand of the system under mixed call loads by combining business traffic, call duration, and video call characteristics. On this basis, the total demand is combined with the single-machine capacity to calculate the theoretical server demand, and finally, a performance difference correction factor based on the hardware differences between production and load testing environments, determined through historical data or linear regression, is introduced to calibrate the theoretical server demand, thereby outputting a server demand that closely reflects the actual needs of the production environment. Attached Figure Description

[0012] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 A flowchart illustrating a call center capacity assessment method provided for an exemplary embodiment of this application; Figure 2 A schematic diagram of an actual process for applying the call center capacity assessment method provided in an exemplary embodiment of this application; Figure 3A schematic diagram of the structure of a call center capacity assessment device provided as an exemplary embodiment of this application; Figure 4 This is a schematic diagram of the structure of an electronic device provided as an exemplary embodiment of this application. Detailed Implementation

[0013] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0014] The following is a description of the terms used in this application: A call center, as described in this application, refers to a centralized service system that utilizes computer telephony integration technology to handle a large volume of incoming and outgoing calls. Its core functions include automatic call distribution, interactive voice response, agent management, and multimedia communication support. In the context of this invention, a call center specifically refers to a modern communication service platform that supports mixed audio and video call services, and is the object and target system for implementing the dynamic capacity assessment method of this invention.

[0015] Stress testing is a testing method that evaluates the performance, stability, and reliability of a system or component by simulating extreme loads beyond normal operating conditions. In this invention, it specifically refers to the process of using tools such as SIPp to simulate high-concurrency audio, video, and mixed audio-video call scenarios to perform load testing on the call center media server, in order to obtain its key performance indicators (such as maximum concurrency, CPU utilization, etc.) under extreme or saturated conditions.

[0016] SIPp (SIP Protocol Tester) is an open-source, high-performance SIP protocol testing tool used to simulate SIP User Agents (UAC / UAS) and generate customizable SIP signaling traffic. In this invention, SIPp is the core tool for performing stress tests, using its command-line parameters (such as -r to set the number of calls per second, -d to set the call duration, -m to set the number of calls, etc.) to achieve accurate load simulation and performance data collection for call center media services.

[0017] Concurrent Calls. Concurrent Calls refer to the number of call sessions that the system is processing simultaneously at any given moment. It is a key capacity metric for measuring the processing ability of a call center. In this invention, the formula for calculating Concurrent Calls is: Concurrent Calls = Calls Attempts Per Second (CAPS) × Average Call Duration. The evaluation objective is to accurately predict the maximum concurrent calls that the system can stably support under different traffic models.

[0018] Calls Attempts Per Second (CAPS). Calls Attempts Per Second is a metric for measuring the frequency of business requests in a call center, representing the number of new call establishment attempts received by the system per second. It is one of the basic input parameters for calculating concurrent calls and evaluating the throughput capacity of the system.

[0019] Traffic Model. The Traffic Model is an abstract representation that mathematically describes the traffic characteristics of a call center's business. In this invention, it specifically refers to the mixed audio - video traffic model, whose key parameters include: Video Traffic Ratio (x), Resource Consumption Ratio of Video to Audio (b), and Call Duration Ratio of Video to Audio (a). Establishing an accurate traffic model is a prerequisite for achieving dynamic capacity evaluation.

[0020] Video Traffic Ratio. The Video Traffic Ratio refers to the percentage of video call requests in the total call traffic volume (denoted as x, 0 < x < 1). This parameter reflects the business composition, is the core variable in the mixed traffic model, and directly affects the total resource consumption and the calculation of concurrent calls.

[0021] Resource Consumption Ratio of Video to Audio. The Resource Consumption Ratio of Video to Audio (denoted as b, b > 1) refers to the ratio of the average occupancy of key server resources (such as CPU) by a single - path video call to that by a single - path audio call. This parameter is obtained through stress testing, quantifies the resource - intensive degree of video services relative to audio services, and is a key coefficient for calculating the equivalent load under mixed traffic.

[0022] Call Duration Ratio of Video to Audio. The Call Duration Ratio of Video to Audio (denoted as a, a > 1) refers to the ratio of the average duration of video calls to that of audio calls. This parameter is obtained based on historical business data analysis, reflects the business characteristic that video calls generally take longer, and is an important factor for predicting changes in the total concurrent calls.

[0023] The Traffic Correlation Model, the first mathematical model proposed in this invention, is used to calculate the maximum concurrency (m) that a single server can support under mixed traffic conditions, given a video call percentage (x) and an audio / video resource consumption ratio (b). Its core logic is to convert video calls into equivalent audio calls, as shown in the formula: ,in This represents the maximum concurrent audio requests per server.

[0024] The Traffic Concurrency Statistical Model, the second mathematical model proposed in this invention, is used to predict the change in total system concurrency demand when some audio traffic is converted to video traffic. This model is based on the assumption of a constant total traffic volume and calculates concurrency by combining the proportion of video traffic (x) and the audio-to-video call duration ratio (a), using the formula: Total Concurrency = Audio Concurrency × [x*(a-1)+1]. Where audio concurrency = Caps·s represents the system concurrency assuming all calls are audio. This model, by introducing the video proportion x and the duration ratio a, calculates the increase in concurrency caused by some calls being converted to longer video calls, assuming a constant total traffic volume. Here, Caps represents the number of calls per second, indicating the number of new call establishment attempts processed by the system per second, and s represents the average call duration, indicating the average duration of a single call session, in seconds. For example, if a system has 40 calls per second (CAPS) and an average audio call duration of 20 seconds, then the system needs to support approximately 40 × 20 = 800 concurrent audio calls.

[0025] The Performance Discrepancy Correction Factor (k) is a coefficient used to calibrate the differences between theoretical assessment results and actual production environments. Because the stress testing environment and the production environment inherently differ in hardware configuration, software load, network conditions, etc., directly using stress testing data for planning may lead to biases. This factor is determined through historical data comparative analysis or regression analysis and is ultimately used to adjust the theoretical server requirements. The calculation formula is: Actual Requirements = k × Theoretical Requirements.

[0026] As described in the background section, with video calls and hybrid audio-video calls becoming routine business for call centers, the aforementioned existing technical solutions have revealed significant shortcomings and problems in dealing with hybrid call scenarios: The evaluation model is static and singular, and cannot adapt to mixed traffic: Traditional audio stress testing models and the conference stress testing method represented by Comparison Document 2 do not fully consider the huge differences in resource consumption (such as CPU, memory, and bandwidth) between audio and video, and also lack the ability to model the dynamic changes in the mixing ratio of the two, which makes it impossible to accurately evaluate the real server carrying capacity under mixed traffic.

[0027] Relying on runtime adjustments and lacking forward-looking planning: As represented by the solution in Comparison Document 1, its core is to passively expand or shrink capacity based on monitoring metrics during system runtime. While this approach can handle sudden loads, it lacks proactive and forward-looking assessment of capacity requirements and cannot provide accurate data support for resource procurement and deployment during the business planning phase, easily leading to improper initial resource allocation.

[0028] The assessment process is disconnected, leading to significant discrepancies between results and reality: Existing load testing methods (such as SIPP) are typically independent of the final production environment planning. Due to inherent differences in hardware performance and software configuration between the load testing environment and the production environment, directly using load testing data can distort capacity assessment results, resulting in insufficient resource reservation causing system overload, or excessive resource allocation leading to cost waste.

[0029] Over-reliance on experience and lack of standardized processes: Existing capacity assessment work relies heavily on the personal experience of operations and maintenance personnel for parameter estimation and result interpretation. It lacks a set of parameterized and standardized mathematical models and automated processes, resulting in low assessment efficiency, poor consistency, and difficulty in scaling up.

[0030] To address the aforementioned issues in related technologies, this application provides a call center capacity assessment method. First, targeted audio, video, and mixed call load tests are conducted on the target call center. Combined with analysis of its actual business data, baseline performance parameters for a single server are obtained, including maximum audio concurrency support capacity and audio / video resource consumption ratio, as well as core business characteristic parameters such as video call proportion and audio / video call duration ratio. Based on these parameters, on one hand, a first calculation model correlates the baseline single-server capacity, video call proportion, and resource consumption differences to accurately determine the actual capacity limit of a single server under mixed call load. On the other hand, a second calculation model, combining business traffic, call duration, and video call characteristics, accurately predicts the total concurrent demand of the system under mixed call load. Based on this, the total demand is combined with the single-server capacity to calculate the theoretical server demand. Finally, a performance difference correction factor, determined through historical data or linear regression based on hardware differences between production and load testing environments, is introduced to calibrate the theoretical server demand, thereby outputting a server demand that closely reflects the actual needs of the production environment.

[0031] The technical solutions provided by the various embodiments of this application are described in detail below with reference to the accompanying drawings.

[0032] Figure 1 This is a flowchart illustrating a call center capacity assessment method provided as an exemplary embodiment of this application. Figure 1 As shown, the method includes: Step 110: Based on the stress test data of the call center, obtain the maximum audio concurrency support capacity of a single server and the resource consumption ratio of video and audio. Also, based on the business data of the call center, obtain the video call volume ratio and the video and audio call duration ratio of the call center.

[0033] The stress test directly quantified the server's hardware processing capacity limits (maximum audio concurrency support capacity) and the differences in resource consumption among different types of services (audio and video resource consumption ratio), which are objective measurements of physical performance. Simultaneously, based on call center business data, the characteristics of the business (video call volume ratio) and user behavior patterns (audio and video call duration ratio) were extracted. Combining objective performance parameters with subjective business characteristic parameters provides comprehensive and reliable input for subsequently building accurate mathematical models, overcoming the one-sidedness of traditional methods that either only focus on hardware limits or only consider business assumptions, laying the foundation for accurate dynamic evaluation.

[0034] In some exemplary embodiments, the step of acquiring stress test data includes: The SIP protocol testing tool SIPP was used to perform a single-pass call to verify system connectivity, and a large-scale stress test was performed after the verification was successful. During large-scale stress testing, the operational status indicators of a single server are obtained. These operational status indicators include at least one of the following: host performance indicators, call error rate indicators, and media data indicators.

[0035] Single Call Verification refers to initiating a separate, low-complexity call establishment attempt before conducting large-scale stress tests. This confirms the smooth operation of basic signaling interactions and media paths between the system under test (e.g., a call center media server) and the stress testing tool (e.g., SIPP). This step is crucial for ensuring the effectiveness and reliability of subsequent large-scale stress tests, preventing numerous invalid tests and resource waste caused by basic configuration errors or network connectivity issues.

[0036] Large-scale stress testing refers to the process of continuously applying high load to a system by simulating concurrent call requests far exceeding normal business levels after verifying the system's basic connectivity, using stress testing tools. In this application embodiment, it specifically refers to using the SIPP tool to concurrently initiate a large number of audio, video, or mixed call requests according to preset calls per second (CAPS) and call duration, aiming to probe the system's performance boundaries and stability thresholds under extreme or saturated conditions.

[0037] Host performance metrics refer to quantitative data collected from the server hardware level during stress testing, reflecting its resource utilization. In this invention, they primarily refer to CPU utilization and memory usage. Monitoring CPU utilization (such as total percentage utilization and average percentage utilization per throughput) is a core basis for evaluating server computing resource consumption and determining whether it has reached its processing capacity limit.

[0038] The Call Error Rate (CER) metric measures the success rate and quality of signaling layer interactions during stress testing. In this invention, it primarily refers to abnormal events occurring during call setup or release, including call request timeouts without response, abnormal call interruptions (abnormal BYE signaling termination), and signaling message redistribution (such as INVITE and BYE) due to network packet loss or processing delays. This metric directly reflects the stability and reliability of the system's signaling processing under high load.

[0039] Media data metrics are used to evaluate the transmission quality of media streams (audio and video streams) after a call is established. Since signaling interaction may sometimes be normal, but media stream issues (such as silence, stuttering, or high latency) can occur, monitoring only the signaling layer is insufficient for a comprehensive evaluation of the call experience. In this invention, testers periodically initiate real calls to the system under test, subjectively listening to or reviewing the media stream to detect any media-level quality problems, serving as an important supplement to automated signaling monitoring.

[0040] Single-pass verification ensures the basic validity of the test environment, avoiding large-scale invalid testing; large-scale load testing simulates real high-load scenarios. During load testing, three types of indicators are monitored: host performance, call error rate, and media data. This multi-layered, comprehensive monitoring system ensures that the maximum audio concurrency support capacity and resource consumption ratio obtained are not just rough estimates based on a single indicator (such as CPU), but reliable results that comprehensively consider system stability (error rate), user experience (media quality), and resource utilization (host performance). Through standardized load testing tools and comprehensive monitoring indicators, the accuracy, repeatability, and comprehensiveness of basic performance data are guaranteed, fundamentally improving the input quality of subsequent capacity assessment models.

[0041] As an example, a typical SIPP load testing command can be referenced: `. / sipp -sf media_yace.xml -inf callernum.csv -i 192.168.99.168 -p 6666 -s 10086192.168.99.168:5060 -r 30 -d 15000 -rtp_echo -trace_err`. Here, the `-r` parameter sets the calls per second (CAPS), and the `-d` parameter sets the call duration (milliseconds). Before performing large-scale load testing, a single-pass call must be initiated using the `-m 1` parameter to verify connectivity.

[0042] In some exemplary embodiments, the call error rate metric includes at least one of call timeout, abnormal hang-up, or signaling retransmission; the media data metric is obtained by periodically initiating manual detection calls to the server and evaluating call quality.

[0043] The document specifies that call errors include not only explicit timeouts and abnormal hang-ups, but also implicit signaling retransmissions, the latter being a sensitive signal of network congestion or processing delays. Regarding media data metrics, it emphasizes the importance of proactive assessment through timed manual inbound calls, which can capture more subtle signs of system instability. Secondly, it clarifies that the acquisition of media data metrics cannot be entirely automated and reliant on tools; it must be combined with subjective human verification. This method is particularly effective in identifying silent faults where signaling statistics appear normal but the actual user experience is compromised, thus ensuring the authenticity of performance data and a true reflection of user experience.

[0044] Step 120: Based on the maximum audio concurrency support capacity, the proportion of video calls, and the resource consumption ratio, determine the maximum concurrency support capacity of a single server under mixed traffic using the first calculation model; and based on the number of calls per second, audio call duration, video call proportion, and call duration ratio of the call center, predict the total concurrency demand of the call center under mixed traffic using the second calculation model.

[0045] The first calculation model takes server performance parameters and business structure parameters as input, focusing on calculating the actual performance of a single server's supply capacity under mixed traffic conditions, i.e., how many concurrent services a single server can produce, i.e., its maximum concurrent support capacity. On the other hand, the second calculation model takes traffic volume parameters and user behavior parameters as input, focusing on predicting the scale of the entire system's business demand under mixed traffic conditions, i.e., the total amount of concurrent services required, i.e., the total concurrent demand. By separating and coupling supply capacity calculation with business demand prediction, a clear analytical framework is constructed, allowing the theoretical resource demand to be obtained through a simple division (total demand / single-machine supply).

[0046] In some exemplary embodiments, the first calculation model is a call association model, which is used to calculate the upper limit of the carrying capacity of a single server in the mixed call mode based on the maximum concurrent audio support capacity of a single server, the proportion of video calls, and the proportion of audio and video resource consumption.

[0047] In this embodiment of the application, the first calculation model specifically refers to a mathematical model used to solve the problem of calculating the mixed call capacity of a single server, namely, the call correlation model. The core function of this model is to map the known pure audio processing capacity of a single server, under the conditions of a given video service ratio and differences in audio and video resource consumption, to calculate the maximum processing capacity of that server in a mixed call scenario. It establishes a quantitative relationship between audio baseline capacity, service structure parameters, and mixed concurrency capacity.

[0048] The maximum capacity refers to the maximum number of simultaneous and stable audio-video mixed call sessions that a single server can handle while ensuring service quality (such as no excessive CPU usage, signaling errors, or media quality degradation). This maximum value is not fixed but dynamically changes with the proportion of video and the resource consumption ratio in the call traffic model, and is one of the key evaluation outputs of this invention.

[0049] The core technology of the call traffic correlation model lies in introducing the key performance parameter of the resource consumption ratio between video and audio. Video calls with different resource consumption intensities are uniformly converted into equivalent audio calls according to their consumption ratio (b) with audio calls. Combined with the video call traffic proportion (x), the overall equivalent audio concurrency of the mixed call traffic can be calculated. When this equivalent concurrency equals the server's maximum audio concurrency support capacity (…),… When the server reaches full load, it is considered to have reached full capacity, thus allowing the deduction of its actual mixed traffic concurrency (m). This transforms the heterogeneous (audio / video) resource consumption problem into a linear superposition problem of homogeneous (equivalent audio) resources through a measurable scaling factor (b). This allows for the reuse of mature audio capacity assessment experience and the application of formulas... The precise quantification of video services introduces a dilution effect on the capabilities of a single server.

[0050] For example, the server can support 800 audio calls, that is... Video traffic accounts for 10%, i.e., x=0.1; video and audio resource consumption accounts for 16%, i.e., b=16. Therefore, the maximum concurrent support capacity of the server for mixed traffic orders is m=800 / [0.1×(16)]. 1)+1]=800 / 2.5=320 routes.

[0051] In some exemplary embodiments, the second calculation model is a call concurrency statistics model, which is used to predict the change in total concurrency caused by the conversion of some audio calls to video calls based on the number of calls per second in the call center, the duration of audio calls, the proportion of video calls, and the ratio of audio to video call duration.

[0052] The second calculation model in this application specifically refers to a mathematical model used to solve the problem of predicting the total concurrent demand of the system, namely, a call concurrency statistical model. The core function of this model is, under the assumption that the total number of incoming call attempts (CAPS) remains constant, to quantify the impact of some calls changing from audio to video (accompanied by changes in call duration) on the total number of simultaneous calls (i.e., total concurrency) within the system. It focuses on the impact of changes in business structure on the overall resource consumption of the system.

[0053] Total concurrent demand refers to the total number of call sessions that the call center system needs to support simultaneously to meet expected business volume. This demand is determined by both the call volume per second (CAPS) and call behavior (duration, video ratio). In this invention, the total concurrent demand predicted by the second calculation model is a direct input to the subsequent calculation of the number of servers required. Accurately predicting this demand is crucial to avoiding resource planning deviations.

[0054] The core technology of the call concurrency statistical model lies in simultaneously considering the combined impact of changes in business structure (video proportion x) and user behavior (duration ratio a) on the number of sessions residing in the system. Based on the reasonable assumption that the total number of service requests (CAPS) remains constant, the model analyzes how, when a portion of requests change from short-duration audio to long-duration video, the total number of concurrent sessions accumulated within the system inevitably increases due to the longer time each session occupies system resources. The model quantifies this increase using the formula: Total Concurrency = Caps·s·[x(a-1)+1]. This clearly separates and quantifies the different contributions of service volume and service structure / behavior to system concurrency pressure, allowing planners to anticipate that even with the same number of calls, the widespread adoption of video services or longer video calls may necessitate a significant increase in system capacity, thus avoiding short-sighted capacity planning caused by ignoring the evolution of business models.

[0055] For example, if all calls are audio, the total concurrent calls entering the call center are 10,000. Video calls account for 10%, and the ratio of video to audio call duration is 3. Therefore, the total concurrent demand for mixed audio and video calls = 10000 × [0.1 × (3...]]. 1)+1]=10000×1.2= 12000 routes.

[0056] In some exemplary embodiments, the proportion of video calls and the ratio of video to audio call duration are dynamically determined based on historical business statistics, business prediction models, or real-time monitoring data of the call center.

[0057] Dynamic determination means that the two key business parameters—the proportion of video calls and the ratio of audio to video call duration—are not fixed but can be updated and adjusted based on data sources and business changes. Data sources can include analysis and statistics of historical call records, business volume forecasting models based on market trends, or real-time traffic monitoring data from operating systems. This dynamism ensures that the capacity assessment model can continuously track and adapt to actual business changes.

[0058] By not presetting fixed parameter values, but instead allowing and specifying the dynamic acquisition or updating of these parameters from multiple sources, including historical data, predictive models, and real-time monitoring data, this fully considers the complexity and variability of the actual business environment. This transforms the entire capacity assessment methodology from a static, snapshot-based tool into a system capable of continuous learning, adaptation, and evolution. By connecting to historical databases, predictive algorithms, or monitoring dashboards, it can automatically adjust inputs as business develops (e.g., a gradual increase in video coverage) or seasonal fluctuations occur, thereby continuously outputting assessment results that reflect current and near-future realities.

[0059] Step 130: Based on the total concurrent demand and the maximum concurrent support capacity of a single server under mixed call traffic, determine the initial server demand of the call center, and based on the performance difference correction factor, correct the initial server demand to obtain the corrected server demand of the call center.

[0060] The performance difference correction factor is determined by analyzing historical data or using linear regression methods based on the hardware performance differences between the production environment and the stress testing environment.

[0061] The initial server requirement N refers to the minimum theoretical number of servers needed to meet the predicted total concurrent demand, calculated theoretically. The calculation logic is to divide the predicted total concurrent demand of the system in step 120 by the maximum concurrent support capacity m of a single server under mixed traffic, calculated in step 120, i.e., N = total concurrent demand / m. This value is a direct derivation based on an idealized model and stress test data, and does not yet consider the complex factors of the actual operating environment.

[0062] The revised server demand, also known as the actual server demand, is the final planned quantity obtained by calibrating the initial server demand by introducing a performance difference correction factor (k). The calculation formula is as follows: This value aims to bridge the performance gap between theoretical assessment environments and actual deployment environments. It serves as a direct guide for equipment procurement, resource allocation, or cloud resource applications, representing a more realistic and reliable capacity planning conclusion.

[0063] First, based on the precisely quantified total system demand (denominator) and single-machine supply capacity (numerator) from step 120, the initial server demand (N) is determined through simple division. This step transforms the complex hybrid traffic modeling problem into a clear and executable resource allocation quantity, making the evaluation results more intuitive. On this basis, a performance difference correction factor (k) is introduced to further correct the theoretical calculation results. This factor is specifically used to compensate for the inherent performance deviations between the stress testing environment (typically clean and controlled) and the production environment (containing complex interferences such as background processes, network fluctuations, and hardware batch differences). Through this correction mechanism, the final corrected server demand output is not only based on a rigorous model but also incorporates an understanding and compensation for real-world complexity, thereby greatly improving the reliability and practicality of capacity planning results in the production environment and avoiding capacity planning errors caused by environmental differences.

[0064] The performance difference correction factor (k) is determined based on the hardware performance differences between production and stress testing environments, using historical data analysis or linear regression methods. It can be seen that determining this performance difference correction factor (k) involves two implementation paths: one is an empirical path, which involves accumulating long-term operational experience, comparing the performance of similar servers in two environments, and summarizing an empirical k value; the other is a data modeling path, which involves collecting multiple sets of comparative data, establishing a linear regression model between the stress testing evaluation results and the actual required resources, and using the slope or offset as the k value. This provides an operable and repeatable scientific basis for the correction process, enhancing the objectivity and consistency of the entire method. This ensures that different personnel or those performing evaluations at different times can obtain stable and reliable correction results as long as they follow the same data analysis methods.

[0065] In some exemplary embodiments, the revised server requirements are applied to call center resource scheduling, scaling decisions, or automated operation and maintenance systems.

[0066] In this embodiment of the application, resource scheduling specifically refers to the process of allocating, deploying, or adjusting computing, storage, and network resources at the infrastructure level of the call center based on the calculated and corrected server demand. For example, this involves setting up a specified number of servers in a physical data center, or applying for virtual machine instances of the corresponding specifications and number on a cloud platform, and configuring load balancing strategies to distribute call traffic to these new resources.

[0067] Expansion decision-making refers to the judgment and plan made based on dynamic capacity assessment results regarding whether and by how much server resources are needed in the future. In this application's embodiment, the revised server demand output is compared with the existing resource inventory to generate clear expansion guidance: if the demand exceeds the inventory, the expansion process is triggered, and the expansion amount is the difference between the two; otherwise, the status quo can be maintained or downsizing can be considered. This decision-making supports precise resource investment and cost control.

[0068] An automated operations and maintenance (O&M) system refers to a software platform or toolset capable of automatically or semi-automatically performing O&M tasks. Using the actual server demand obtained through this method as input, the system can automatically trigger subsequent resource configuration workflows. For example, it can automatically create virtual machines by calling cloud management platform interfaces via application programming interfaces (APIs), or generate purchase orders with the required server specifications and quantities, thereby achieving closed-loop automation from capacity assessment to resource readiness.

[0069] The revised capacity assessment results are input into resource scheduling, expansion decision-making, or automated operation and maintenance systems. Thus, in resource scheduling, the revised server demand serves as a resource list; in expansion decision-making, it becomes the basis for decision-making; and in automated operation and maintenance systems, it serves as the instruction or parameter to trigger automated actions. This allows for seamless integration with existing or future operation and maintenance management systems, driving subsequent automated operations and more scientific decision-making.

[0070] Figure 2 This is a schematic diagram illustrating an actual process of applying the call center capacity assessment method provided in an exemplary embodiment of this application. For example... Figure 2 As shown, by constructing a complete technical closed loop from basic data collection to production environment correction, accurate and dynamic assessment of the hybrid call traffic capacity of call centers can be achieved. Specifically, this... Figure 2 It may include: (1) Load testing data acquisition: Specifically, the target server can be subjected to load testing of audio, video and audio-video mixed traffic using the SIP protocol testing tool SIPP. The test aims to obtain the key performance parameters of a single server as a benchmark for capacity planning, mainly including: the maximum audio concurrency supported by a single server (Max_audio), and the resource consumption ratio of video calls to audio calls (b).

[0071] (2) Call traffic structure analysis: Specifically, based on the historical operating data of the call center or the prediction of future business development, the core parameters reflecting the current or expected business characteristics can be determined. These parameters mainly include: the proportion of video calls in the total call volume (x), and the ratio of the average duration of video calls to the average duration of audio calls (a).

[0072] (3) Single Server Capacity Calculation: Specifically, the performance parameters and business parameters obtained in the aforementioned steps can be combined and input into the first calculation model, namely the call association model. This model is used to calculate the maximum number of concurrent calls (m) that the same server can support under a specified audio-video mixed call model. Its core calculation logic is: convert the video calls in the mixed call into equivalent audio calls based on their resource consumption ratio (b) with audio calls, and combine this with the proportion of video calls (x) to calculate the total equivalent audio load. When this equivalent load reaches the server's maximum audio concurrency capacity (m), the total equivalent audio load is calculated. When ), the maximum concurrent support capacity m of the server under mixed traffic is obtained.

[0073] (4) Total Concurrency Prediction: Screenshots can be used to input business characteristic parameters into the second calculation model, namely the call concurrency statistics model. This model is used to predict the total number of call sessions that the system needs to process simultaneously after some audio calls are converted to video calls, i.e., the total concurrency requirement, assuming that the total number of incoming call requests (measured by calls per second CAPS) remains unchanged. The core of its calculation lies in considering the session dwell effect brought about by the longer video call duration (a*audio call duration).

[0074] (5) Resource Planning: This step is the preliminary resource allocation calculation. Specifically, the total concurrent demand of the system predicted in step (4) can be divided by the maximum concurrent support capacity (m) of a single server under mixed traffic calculated in step (3) to obtain the theoretical number of servers (N) required to meet the business needs, i.e., the preliminary server demand. The calculation formula is: N = total concurrent demand / m.

[0075] (6) Performance Correction: This step aims to address the performance differences between the stress testing environment and the actual production environment. By introducing a performance difference correction factor (k), the theoretical number of servers (N) is calibrated to obtain a corrected server requirement that more closely reflects the actual production environment. The correction factor k is based on the analysis of hardware performance differences between the production environment and the stress testing environment, and can be determined through linear regression methods using historical data comparison or empirical values. The correction formula is: .

[0076] (7) Dynamic assessment of resources: Through the above steps (1) to (6), an end-to-end solution closed loop is constructed. Starting from standardized stress testing based on SIPP, key performance and business parameters are extracted, and accurate calculations and predictions are performed using call traffic correlation models and call traffic concurrency statistical models. Finally, reliable resource planning quantities are output by correlating environmental difference correction factors with actual production. This closed-loop process enables the capacity assessment work to be standardized and automated, and can be dynamically updated with changes in business parameters, realizing the scientific and dynamic assessment and planning of call center resources.

[0077] As an example, to more intuitively illustrate the above process, assume a call center's business requirement is to support 5000 mixed calls. Analysis determines that video call volume accounts for 0.15%, the resource consumption ratio of video to audio is b=30, and the video to audio call duration ratio is a=1.2. Load testing yields a maximum audio concurrency of 800 per server. Based on experience, the performance difference correction factor k=1.1.

[0078] The single-server mixed traffic capacity m = 800 / [0.15 × (30 - 1) + 1] ≈ 149.

[0079] Total system concurrency requirement = 5000 × [0.15 × (1.2 - 1) + 1] = 5150.

[0080] The theoretical number of servers N = 5150 / 149 ≈ 34.6, which is rounded up to 35.

[0081] Revised server requirements Rounded up, the result is 39 units.

[0082] This example demonstrates the complete calculation process from parameter input to the final server demand.

[0083] The call center capacity assessment method provided in this application first conducts targeted audio, video, and mixed call load tests on the target call center, and analyzes its actual business data to obtain the baseline performance parameters of a single server, namely the maximum audio concurrency support capacity and audio / video resource consumption ratio, as well as core business characteristic parameters, namely the video call ratio and audio / video call duration ratio. Based on these parameters, on the one hand, a first calculation model is used to correlate and calculate the baseline single-machine capacity, video call ratio, and resource consumption differences to accurately determine the actual carrying capacity limit of a single server under mixed call loads; on the other hand, a second calculation model is used to accurately predict the total concurrent demand of the system under mixed call loads by combining business traffic, call duration, and video call characteristics. On this basis, the total demand is combined with the single-machine capacity to calculate the theoretical server demand, and finally, a performance difference correction factor based on the hardware differences between production and load testing environments, determined through historical data or linear regression, is introduced to calibrate the theoretical server demand, thereby outputting a server demand that closely reflects the actual needs of the production environment.

[0084] Figure 3 This is a schematic diagram of the structure of a call center capacity assessment device 300 provided for an exemplary embodiment of this application. Figure 3 As shown, the device 300 includes: an acquisition module 310, a prediction module 320, and an evaluation module 330, wherein: The acquisition module 310 is used to acquire the maximum audio concurrency support capacity of a single server and the resource consumption ratio of video and audio based on the stress test data of the call center, and to acquire the video call traffic ratio and the video and audio call duration ratio of the call center based on the business data of the call center. The prediction module 320 is used to determine the maximum concurrent support capacity of a single server under mixed traffic based on the maximum audio concurrent support capacity, the video call volume ratio, and the resource consumption ratio through a first calculation model, and to predict the total concurrent demand of the call center under mixed traffic based on the number of calls per second of the call center, the audio call duration, the video call volume ratio, and the call duration ratio through a second calculation model. The evaluation module 330 is used to determine the initial server requirements of the call center based on the total concurrent demand and the maximum concurrent support capacity of a single server under mixed traffic, and to correct the initial server requirements based on a performance difference correction factor to obtain the corrected server requirements of the call center; wherein, the performance difference correction factor is determined by historical data analysis or linear regression methods based on the hardware performance differences between the production environment and the stress test environment.

[0085] The call center capacity assessment device 300 provided in this application first conducts targeted audio, video, and mixed call load tests on the target call center, and analyzes its actual business data to obtain the baseline performance parameters of a single server, namely the maximum audio concurrency support capacity and audio / video resource consumption ratio, as well as core business characteristic parameters, namely the video call ratio and audio / video call duration ratio. Based on these parameters, on the one hand, a first calculation model is used to correlate and calculate the baseline single-machine capacity, video call ratio, and resource consumption differences to accurately determine the actual carrying capacity limit of a single server under mixed call loads; on the other hand, a second calculation model is used to accurately predict the total concurrent demand of the system under mixed call loads by combining business traffic, call duration, and video call characteristics. On this basis, the total demand is combined with the single-machine capacity to calculate the theoretical server demand, and finally, a performance difference correction factor based on the hardware differences between production and load testing environments, determined through historical data or linear regression, is introduced to calibrate the theoretical server demand, thereby outputting a server demand that closely reflects the actual needs of the production environment.

[0086] Optionally, the step of acquiring the stress test data includes: The SIP protocol testing tool SIPP was used to perform a single-pass call to verify system connectivity, and a large-scale stress test was performed after the verification was successful. During the large-scale stress test, the operating status indicators of a single server are obtained, including at least one of host performance indicators, call error rate indicators, and media data indicators.

[0087] Optionally, the call error rate metric includes at least one of call timeout, abnormal hang-up, or signaling retransmission; the media data metric is obtained by periodically initiating manual detection calls to the server and evaluating call quality.

[0088] Optionally, the first calculation model is a call traffic association model, which is used to calculate the upper limit of the carrying capacity of a single server in the mixed call traffic mode based on the maximum concurrent audio support capacity of a single server, the proportion of video call traffic, and the proportion of audio and video resource consumption.

[0089] Optionally, the second calculation model is a call concurrency statistics model, used to predict the change in total concurrency caused by the conversion of some audio calls to video calls based on the number of calls per second in the call center, audio call duration, video call proportion, and audio-video call duration ratio.

[0090] Optionally, the proportion of video calls and the ratio of video to audio call duration are dynamically determined based on the historical business statistics, business prediction models, or real-time monitoring data of the call center.

[0091] The call center capacity assessment device 300 can achieve Figures 1-2 For details of the method implementation examples, please refer to [link / reference]. Figures 1-2 The call center capacity assessment method shown in the embodiment will not be described in detail again.

[0092] Figure 4 This is a schematic diagram of the structure of an electronic device provided as an exemplary embodiment of this application. For example... Figure 4 As shown, the device includes a memory 41 and a processor 42.

[0093] Memory 41 is used to store computer programs and can be configured to store various other data to support operation on the computing device. Examples of this data include instructions for any application or method used to operate on the computing device, contact data, phone book data, messages, images, videos, etc.

[0094] Processor 42, coupled to memory 41, is used to execute computer programs in memory 41 for: obtaining the maximum audio concurrency support capacity and the resource consumption ratio of video and audio for a single server based on stress test data from the call center; and obtaining the video call volume ratio and the video and audio call duration ratio of the call center based on the call center's business data; determining the maximum concurrency support capacity of a single server under mixed call volume using a first calculation model based on the maximum audio concurrency support capacity, the video call volume ratio, and the resource consumption ratio; and predicting the total concurrency demand of the call center under mixed call volume using a second calculation model based on the number of calls per second, audio call duration, video call volume ratio, and call duration ratio of the call center; determining the initial server demand of the call center based on the total concurrency demand and the maximum concurrency support capacity of a single server under mixed call volume; and correcting the initial server demand based on a performance difference correction factor to obtain the corrected server demand of the call center; wherein the performance difference correction factor is determined by historical data analysis or linear regression methods based on the hardware performance difference between the production environment and the stress test environment.

[0095] The electronic device provided in this application first conducts targeted audio, video, and mixed call load tests on the target call center, and analyzes its actual business data to obtain the baseline performance parameters of a single server, namely the maximum audio concurrency support capacity and audio / video resource consumption ratio, as well as core business characteristic parameters, namely the video call ratio and audio / video call duration ratio. Based on these parameters, on the one hand, a first calculation model is used to correlate and calculate the baseline single-machine capacity, video call ratio, and resource consumption differences to accurately determine the actual carrying capacity limit of a single server under mixed call loads; on the other hand, a second calculation model is used to accurately predict the total concurrent demand of the system under mixed call loads by combining business traffic, call duration, and video call characteristics. On this basis, the total demand is combined with the single-machine capacity to calculate the theoretical server demand, and finally, a performance difference correction factor based on the hardware differences between production and load testing environments, determined through historical data or linear regression, is introduced to calibrate the theoretical server demand, thereby outputting a server demand that closely reflects the actual needs of the production environment.

[0096] Furthermore, such as Figure 4 As shown, the electronic device also includes other components such as a communication component 43, a display 44, a power supply component 45, and an audio component 46. Figure 4 The diagram only shows some components and does not mean that the electronic device includes only these components. Figure 4 The components shown. Additionally, depending on the implementation of the traffic playback device, Figure 4 The components within the dashed box are optional, not mandatory. For example, when an electronic device is implemented as a terminal device such as a smartphone, tablet, or desktop computer, it may include... Figure 4 The components within the dashed box; when the electronic device is implemented as a server-side device such as a conventional server, cloud server, data center, or server array, it may be excluded. Figure 4 The component within the dashed box.

[0097] The above Figure 4 The communication component is configured to facilitate wired or wireless communication between the device containing the communication component and other devices. The device containing the communication component can access wireless networks based on communication standards, such as WiFi, 2G, or 3G, or combinations thereof. In one exemplary embodiment, the communication component receives broadcast signals or broadcast-related information from an external broadcast management system via a broadcast channel. In one exemplary embodiment, the communication component may further include a Near Field Communication (NFC) module, Radio Frequency Identification (RFID) technology, Infrared Data Association (IrDA) technology, Ultra Wideband (UWB) technology, Bluetooth (BT) technology, etc.

[0098] The above Figure 4The memory in the memory can be implemented by any class of volatile or non-volatile storage devices or combinations thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk or optical disk.

[0099] The above Figure 4 The display includes a screen, which may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen can be implemented as a touchscreen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touches, swipes, and gestures on the touch panel. The touch sensors can sense not only the boundaries of the touch or swipe action, but also the duration and pressure associated with the touch or swipe operation.

[0100] The above Figure 4 The power supply component provides power to the various components of the device in which it resides. The power supply component may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to the device in which it resides.

[0101] The above Figure 4 The audio component can be configured to output and / or input audio signals. For example, the audio component includes a microphone (MIC) configured to receive external audio signals when the device containing the audio component is in an operating mode, such as call mode, recording mode, or voice recognition mode. The received audio signals can be further stored in memory or transmitted via a communication component. In some embodiments, the audio component also includes a speaker for outputting audio signals.

[0102] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0103] Accordingly, embodiments of this application also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, enables the processor to implement the steps in the above-described call center capacity assessment method embodiments.

[0104] Accordingly, this application also provides a computer program product, which stores instructions that, when executed by a computer, cause the computer to perform the steps in the call center capacity assessment method embodiment provided in this application.

[0105] 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, and 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.

[0106] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0107] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0108] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0109] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0110] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other classes of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0111] 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 a process, method, article, or apparatus. Without further limitation, 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 said element.

[0112] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for assessing call center capacity, characterized in that, include: Based on the stress test data of the call center, the maximum audio concurrency support capacity of a single server and the resource consumption ratio of video and audio are obtained. Based on the business data of the call center, the video call volume ratio and the video and audio call duration ratio of the call center are obtained. Based on the maximum audio concurrency support capacity, the video call volume ratio, and the resource consumption ratio, the maximum concurrency support capacity of a single server under mixed call volume is determined by a first calculation model. Based on the number of calls per second of the call center, the audio call duration, the video call volume ratio, and the call duration ratio, the total concurrency demand of the call center under mixed call volume is predicted by a second calculation model. Based on the total concurrent demand and the maximum concurrent support capacity of a single server under mixed call traffic, the initial server demand of the call center is determined, and the initial server demand is corrected based on a performance difference correction factor to obtain the corrected server demand of the call center; wherein, the performance difference correction factor is determined by historical data analysis or linear regression methods based on the hardware performance difference between the production environment and the stress test environment.

2. The method according to claim 1, characterized in that, The steps for obtaining the stress test data include: The SIP protocol testing tool SIPP was used to perform a single-pass call to verify system connectivity, and a large-scale stress test was performed after the verification was successful. During the large-scale stress test, the operating status indicators of a single server are obtained, including at least one of host performance indicators, call error rate indicators, and media data indicators.

3. The method according to claim 2, characterized in that, The call error rate metric includes at least one of call timeout, abnormal hang-up, or signaling retransmission; the media data metric is obtained by periodically initiating manual detection calls to the server and evaluating call quality.

4. The method according to claim 1, characterized in that, The first calculation model is a call traffic association model, which is used to calculate the upper limit of the carrying capacity of a single server in a mixed call traffic mode based on the maximum concurrent audio support capacity of a single server, the proportion of video call traffic, and the proportion of audio and video resource consumption.

5. The method according to claim 1, characterized in that, The second calculation model is a call concurrency statistics model, which is used to predict the change in total concurrency caused by the conversion of some audio calls to video calls based on the number of calls per second in the call center, audio call duration, video call proportion, and audio-video call duration ratio.

6. The method according to claim 1, characterized in that, The proportion of video calls and the ratio of video to audio call duration are dynamically determined based on the call center's historical business statistics, business prediction models, or real-time monitoring data.

7. A call center capacity assessment device, characterized in that, include: The acquisition module is used to acquire the maximum audio concurrency support capacity of a single server and the resource consumption ratio of video and audio based on the stress test data of the call center, and to acquire the video call traffic ratio and the video and audio call duration ratio of the call center based on the business data of the call center. The prediction module is used to determine the maximum concurrent support capacity of a single server under mixed traffic based on the maximum audio concurrent support capacity, the video call volume ratio, and the resource consumption ratio through a first calculation model, and to predict the total concurrent demand of the call center under mixed traffic based on the number of calls per second of the call center, the audio call duration, the video call volume ratio, and the call duration ratio through a second calculation model. The evaluation module is used to determine the initial server requirements of the call center based on the total concurrent demand and the maximum concurrent support capacity of a single server under mixed call traffic, and to correct the initial server requirements based on a performance difference correction factor to obtain the corrected server requirements of the call center; wherein, the performance difference correction factor is determined by historical data analysis or linear regression methods based on the hardware performance differences between the production environment and the stress test environment.

8. An electronic device, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the steps of the method as claimed in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the method as described in any one of claims 1 to 6.

10. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the steps of the method according to any one of claims 1 to 6.

Images