A network performance testing method and system based on 5G communication terminal

Abstract

The present invention belongs to the field of network testing and provides a network performance testing method and system based on 5G communication terminals, including configuring N+1 5G communication terminals and N hardware relay devices; selecting one from the N+1 5G communication terminals as the current test terminal, and corresponding the remaining N 5G communication terminals to the N hardware relay devices one by one; utilizing the N hardware relay devices to amplify and multi-channel distribute the signals of the corresponding 5G communication terminals; controlling the N+1 5G communication terminals to perform data transmission tests; obtaining test data of the current test terminal and adding it to a first test data group; obtaining test data of the remaining N 5G communication terminals and adding it to a second test data group; and analyzing and determining the network performance based on the data of the first test data group and the second test data group.

Description

Technical Field

[0001] The present invention belongs to the field of network testing, and specifically relates to a network performance testing method and system based on 5G communication terminals. Background Art

[0002] With the rapid development of fifth-generation mobile communication technology (5G), 5G networks, with their high speed, low latency, and massive connectivity, are widely used in fields such as the Industrial Internet of Things, smart cities, telemedicine, and autonomous driving. However, the complexity of 5G networks also brings challenges to performance testing. Accurate network performance assessment is not only related to the optimized design of 5G networks but also directly affects user experience and the stability of business applications.

[0003] Existing 5G network testing methods often fail to fully reflect network performance in multi-user scenarios based on single-terminal test results. For example, in real-world 5G applications, multiple user terminals often access the network simultaneously, leading to performance interference due to resource competition between terminals. Furthermore, the dynamic nature of the network (such as resource scheduling and slicing) causes network performance to vary over time and under load conditions. Therefore, traditional single-terminal static testing methods cannot effectively capture these complex performance characteristics.

[0004] To address these issues, multi-terminal joint testing has become an important method for evaluating 5G network performance. By configuring multiple terminals and hardware relay devices to simulate large-scale access and diverse load conditions, it can more realistically reflect the network's actual performance. However, existing testing methods still have the following shortcomings in handling multi-terminal background traffic and dynamic performance analysis:

[0005] Traditional testing methods primarily focus on static performance data at a single moment in time, lacking the ability to analyze dynamic changes in network performance over time. This limitation makes it impossible to capture the performance transition patterns of the network under different load conditions.

[0006] 5G network performance involves multiple key indicators (such as throughput, latency, packet loss rate, signal quality, etc.). There are complex dependencies between different indicators, and existing methods find it difficult to extract meaningful information from multi-dimensional data.

[0007] In multi-user access scenarios, background load (such as traffic generated by auxiliary terminals) has a significant impact on network performance, but traditional methods usually fail to effectively integrate background load data for joint analysis. Summary of the Invention

[0008] In order to solve the problems in the prior art, the present invention provides a network performance testing method based on a 5G communication terminal, comprising the following steps:

[0009] Step 1: Configure N+1 5G communication terminals and N hardware relay devices;

[0010] Step 2: Select one of the N+1 5G communication terminals as the current test terminal, and assign one-to-one correspondence between the remaining N 5G communication terminals and the N hardware relay devices;

[0011] Step 3: Use the N hardware relay devices to amplify and demultiplex the signals of the corresponding 5G communication terminals;

[0012] Step 4: Control the N+1 5G communication terminals to perform a data transmission test;

[0013] Step 5: Obtain the test data of the current test terminal and add it to the first test data group;

[0014] Step 6: Obtain test data of the remaining N 5G communication terminals and add them to the second test data group;

[0015] Step 7: Repeat steps 2 to 6, taking each of the N+1 5G communication terminals as the current test terminal in turn, obtaining test data, adding the test data of the current test terminal to the first test data group, and adding the test data of the remaining N 5G communication terminals to the second test data group;

[0016] Step 8: Analyze and determine the network performance based on the data of the first test data group and the second test data group.

[0017] Furthermore, the 5G communication terminal is one or more of a smart phone, a fixed wireless access device, and an industrial-grade 5G module.

[0018] Furthermore, the hardware relay device refers to a hardware device that can extend, amplify and distribute wireless signals.

[0019] Furthermore, the current test terminal serves as an active test device to simulate user behavior; the remaining N 5G communication terminals are used to generate background traffic to simulate a multi-user scenario.

[0020] Furthermore, the multi-path distribution includes: for each distributed signal, the relay device adjusts its directivity to ensure that the distributed signal coverage areas do not overlap, and sets different powers for each distributed signal to simulate user access environments of different strengths.

[0021] Furthermore, controlling the N+1 5G communication terminals to perform a data transmission test includes:

[0022] The current test terminal sends large data packets to the target server or network to test the network's uplink throughput and stability;

[0023] The current test terminal receives large data packets from the target server or network to test the network's downlink throughput and stability;

[0024] The current test terminal uses small data packets to measure end-to-end delay and counts the changes in the time interval between the arrival of statistical data packets;

[0025] The current test terminal sends a preset number of data packets, calculates the proportion of lost data packets, and evaluates the reliability of the network;

[0026] Each auxiliary terminal sends and receives data according to a preset traffic model, simulating background user traffic in real scenarios;

[0027] Different auxiliary terminals are set with different service types and data rates;

[0028] The auxiliary terminal simulates traffic changes in different time periods.

[0029] Furthermore, the test data includes: throughput, delay, jitter, packet loss rate, and signal quality.

[0030] Furthermore, after each round, the data of the current test terminal is classified into the first test data group, and the data of the auxiliary terminal is classified into the second test data group;

[0031] The data is marked by terminal number and test round. After all terminals are used as current test terminals, all rounds of testing are completed.

[0032] On the other hand, the present invention also provides a network performance testing system based on a 5G communication terminal, which performs network performance testing using the aforementioned method.

[0033] The present invention proposes a network performance testing method based on 5G communication terminals. By combining a first test data set and a second test data set, the method comprehensively analyzes the dynamic changes of network performance under different load conditions. Compared with the existing technology, the method has the following advantages:

[0034] This method incorporates background load conditions into the second test data set, dynamically adjusting the hidden state transition probabilities and observation probability distributions. This allows it to reflect the true impact of background load on network performance in scenarios with multiple concurrent terminals. This improvement makes the evaluation results more representative and provides data support for optimizing resource allocation in multi-user scenarios.

[0035] By combining multiple 5G terminals and hardware relay devices, this method can simulate large-scale user access and complex network environments under limited physical terminal conditions. It is suitable for performance testing of various 5G application scenarios (such as industrial Internet of Things and autonomous driving), and the test results can truly reflect the network performance in the actual operating environment.

[0036] By incorporating a hidden Markov model (HMM), this method reveals the hidden states and transition patterns of network performance over time, thereby capturing the dynamic performance changes of 5G networks under varying load conditions. This dynamic analysis capability can help identify performance bottlenecks and abnormal transitions, providing guidance for network optimization. This method utilizes an improved HMM to jointly model multi-dimensional performance indicators such as throughput, latency, and packet loss rate. By capturing the potential correlations between these indicators, it can more accurately reflect the comprehensive characteristics of network performance.

[0037] Based on the state transition rules of the hidden Markov model, this method can not only analyze the current network performance, but also predict the future performance status, help discover potential risks in advance, and guide the optimization of resource scheduling and configuration strategies of 5G networks.

[0038] In summary, the present invention achieves a comprehensive evaluation of 5G network performance by introducing an improved hidden Markov model and a multi-terminal joint testing method, which can provide a scientific basis for the optimized design, stable operation and improved user experience of 5G networks, and has significant technical value and application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0040] Figure 1 It is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0041] Below, the invention is preferably described with reference to the accompanying drawings and specific embodiments.

[0042] This embodiment solves the above problem through the following steps:

[0043] In one embodiment, reference Figure 1 The present invention provides a network performance testing method based on 5G communication terminals, which involves comprehensively testing and analyzing various performance indicators of the 5G network by deploying and using multiple 5G communication terminal devices and auxiliary hardware relay devices to evaluate the network's throughput, latency, jitter, packet loss rate, signal coverage, and switching performance, thereby accurately measuring the operating status of the 5G network and providing data support for network optimization and user experience improvement. Specifically, the method includes the following steps:

[0044] Step 1: Configure N+1 5G communication terminals and N hardware relay devices.

[0045] This step aims to provide the necessary equipment configuration and basic conditions for the implementation of subsequent network performance testing.

[0046] 5G communication terminals refer to user terminal devices that support fifth-generation mobile communication technology (5G), including but not limited to smartphones, fixed wireless access equipment (such as CPE devices), and industrial-grade 5G modules. These terminals can connect to the 5G network and support 5G characteristics such as high speed, low latency, and large-scale connectivity. They are core equipment for network performance testing. In this step, N+1 5G communication terminals are configured, where N is the number of terminals participating in the signal amplification and distribution test. This number can be determined based on the test scenario or the number of existing terminals, such as 10 or 15. The additional terminal is used to collect actual test data.

[0047] Hardware relay devices are devices that extend, amplify, and distribute wireless signals, including RF repeaters, signal amplifiers, or protocol splitters. These devices receive signals transmitted by 5G terminals, amplify them, and distribute them across multiple virtual signal channels to simulate larger-scale terminal access scenarios. In this step, N hardware relay devices are configured to correspond to N 5G communication terminals to ensure signal transmission quality and distribution.

[0048] In practice, when selecting N+1 5G communication terminals, ensure that their hardware performance and software configuration meet test requirements, including support for the target test frequency band, high data throughput, and compatibility with the target test network's protocol stack (such as the 5G NR protocol). Furthermore, the configured N hardware relay devices should feature high-gain amplification characteristics and low-noise interference to ensure no significant loss of signal quality during the amplification process. Furthermore, basic network connectivity should be configured for all devices to ensure that the terminal devices can properly access the test network and that signal transmission between the hardware relay devices and the terminals is stable, thus creating a complete test environment.

[0049] In addition, the layout location of the terminal and relay affects the test results, but the layout location is not the focus of the present invention. Those skilled in the art can determine the layout location according to the test scenario. For example, indoor dense scenarios, outdoor wide-area scenarios, high-speed mobile scenarios, etc. are all commonly used test scenarios.

[0050] Step 2: Select one of the N+1 5G communication terminals as the current test terminal, and correspond the remaining N 5G communication terminals to the N hardware relay devices one by one.

[0051] The core of this step is to clarify the division of labor between the test terminal and the auxiliary terminal, and to establish signal extension and distribution paths through hardware relay equipment to provide basic support for the collection of subsequent test data.

[0052] From the configured N+1 5G communication terminals, select one as the current test terminal for collecting and analyzing actual test data.

[0053] The current responsibilities of the test terminal are:

[0054] Establish a direct connection to the target network.

[0055] Receive test data from the network (such as throughput, latency, and other indicators).

[0056] During the entire testing process, it acts as an active testing device to simulate user behavior.

[0057] The remaining N 5G communication terminals are designated as auxiliary terminals. The main responsibilities of these terminals are:

[0058] Establish a connection with the corresponding hardware relay device to complete the signal receiving and sending tasks.

[0059] Cooperate with relay equipment to complete signal amplification and distribution, and provide support for simulating user behavior on a larger scale.

[0060] When the current test terminal is testing, these auxiliary terminals are mainly responsible for generating background traffic and simulating multi-user scenarios.

[0061] For N auxiliary terminals, they are mapped and bound to N hardware relay devices one by one to establish corresponding relationships:

[0062] Each auxiliary terminal is connected to only one relay device to ensure the independence of the signal path.

[0063] After receiving the signal from the auxiliary terminal, the hardware relay device amplifies and distributes it so that the signal can cover the target area or extend to other test areas.

[0064] Step 3: Use the N hardware relay devices to amplify and multi-channel the signals of the corresponding 5G communication terminals.

[0065] Before signal amplification and distribution, the hardware relay equipment needs to be initialized and configured to ensure that it can correctly process the signals from the corresponding 5G communication terminals.

[0066] Ensure that each hardware relay device corresponds to a 5G communication terminal one-to-one, configure the device identifier (such as device ID or MAC address) and establish a connection relationship.

[0067] Configure the operating frequency range of the relay device based on the actual operating frequency band of the test target network (such as Sub-6GHz or millimeter wave band) to match the signal characteristics of the 5G communication terminal.

[0068] Configure the signal gain parameters (such as the amplification factor) of the relay device to ensure that signal quality is not degraded due to insufficient amplification and noise is not introduced due to excessive amplification during signal transmission.

[0069] Signal amplification is a key step in hardware relay equipment to enhance the received 5G signal, which is used to improve the signal strength and coverage.

[0070] The hardware relay device captures the radio frequency signal from the corresponding 5G communication terminal through the built-in receiving module.

[0071] Ensure that the received signal quality reaches a certain level (for example, RSRP is not less than -110dBm).

[0072] The relay equipment amplifies the received signal to increase the signal power so that it can meet the needs of longer distance transmission.

[0073] During the amplification process, a low-noise amplifier (LNA) is used to reduce noise interference and ensure that the amplified signal retains the integrity of the original data.

[0074] Use signal testing tools to verify the quality of the amplified signal to ensure that the signal strength and signal quality (such as signal-to-noise ratio (SNR)) meet the target requirements.

[0075] After the signal amplification is completed, the relay equipment will distribute the enhanced signal into multiple channels to support multiple virtual terminals or signal coverage requirements in different directions.

[0076] The relay device divides the enhanced signal into multiple channels through the built-in signal distribution module. Each signal channel can be used independently or sent to a specific direction.

[0077] The number of allocated paths is set according to the test requirements, for example, it can be divided into 2 paths, 4 paths or more.

[0078] For each distributed signal, the relay device can adjust its directionality (such as controlling the signal coverage range through a directional antenna) to adapt to the specific needs of the test scenario.

[0079] Ensure that the assigned signal coverage areas do not overlap to reduce possible signal interference.

[0080] Different powers are set for each distributed signal to simulate user access environments with different strengths, such as strong signal areas and weak signal areas.

[0081] After signal amplification and demultiplexing are completed, the signal needs to be verified to ensure that the distribution process does not introduce errors or interference.

[0082] Use a signal test tool to measure the strength of each distributed signal to ensure that its power level meets the set value.

[0083] Avoid problems such as excessive signal strength attenuation or uneven power distribution during the distribution process.

[0084] Measure the quality of each signal (such as SNR, RSRP, and RSRQ) to verify whether it meets the expected signal coverage requirements.

[0085] This step uses hardware relay equipment to amplify and distribute signals across multiple channels, effectively expanding the signal coverage of 5G communication terminals and creating a realistic and complex network environment for multi-terminal performance testing. Through appropriate signal amplification and distribution configuration, we can effectively simulate large-scale user behavior in real-world network scenarios, ensuring the representativeness and accuracy of test results.

[0086] Step 4: Control the N+1 5G communication terminals to perform data transmission tests.

[0087] Data transmission of the current test terminal

[0088] The current test terminal selected from N+1 5G communication terminals will undertake the main data transmission tasks:

[0089] Uplink test: The current test terminal sends large data packets to the target server or network to test the network's uplink throughput and stability.

[0090] Downlink test: The test terminal receives large data packets from the target server or network to test the downlink throughput and stability of the network.

[0091] Latency and jitter testing: Use small data packets (such as ICMPPing) to measure end-to-end latency and measure the variation in the time interval between packet arrivals (jitter).

[0092] Packet loss rate test: Send a certain number of data packets, count the proportion of lost data packets, and evaluate network reliability.

[0093] Auxiliary terminal cooperation tasks

[0094] The remaining N auxiliary terminals communicate with the target network through relay devices, simulating a multi-terminal concurrent environment:

[0095] Background traffic generation: Each auxiliary terminal sends and receives data according to a preset traffic model (such as continuous video streaming, large file downloads, and periodic communication of IoT devices), simulating background user traffic in real scenarios.

[0096] Terminal diversity: Different auxiliary terminals are set with different service types and data rates. For example, some terminals run high-bandwidth services (such as high-definition video streaming) while others run low-bandwidth services (such as sensor data uploading).

[0097] Dynamic load changes: Simulate traffic changes in different time periods, such as sudden high loads in a short period of time, to test the network's dynamic adjustment capabilities.

[0098] Step 5: Acquire the test data of the current test terminal and add it to the first test data group.

[0099] Test data refers to the performance indicators collected through the interaction between the current test terminal and the target network during the data transmission test. It mainly includes:

[0100] Throughput: The maximum data transmission rate of the network upstream or downstream, usually measured in Mbps or Gbps.

[0101] Latency: The time required from when a terminal sends a data packet to when it receives a response data packet, measured in milliseconds (ms).

[0102] Jitter: Fluctuations in packet transmission time, reflecting network stability.

[0103] Packet Loss Rate: The ratio of data packets lost during transmission, usually expressed as a percentage.

[0104] Signal quality (RSRP, SINR, etc.): Network signal strength and interference.

[0105] The first test data group is used to store the test data generated by the current test terminal. This data group contains the network performance data recorded for each terminal in all rounds when it served as the current test terminal. This data group aggregates the independent performance indicators of the current test terminal to facilitate subsequent analysis of performance differences between different terminals.

[0106] Step 6: Obtain the test data of the remaining N 5G communication terminals and add them to the second test data group.

[0107] The remaining N 5G communication terminals refer to the N terminal devices not selected as "current test terminals" in the current test round. They establish a connection to the target network through hardware relay devices and serve as background traffic terminals or auxiliary terminals to simulate a multi-user environment. Signals are amplified and distributed by the relay devices before being connected to the target network. Different types of user behavior are simulated, such as high bandwidth requirements (video streaming), low latency requirements (IoT), or sudden load changes. Background traffic is generated to increase the load on the target network. Performance data during the signal amplification and distribution process is recorded.

[0108] The second test data group is used to store the collection of test data generated by all auxiliary terminals in the current test round. Data group content: Contains test data of each auxiliary terminal, such as signal quality, throughput, delay, jitter, packet loss rate and other performance indicators.

[0109] This is used to analyze the effects of signal amplification and distribution and to evaluate the performance differences of auxiliary terminals in a multi-user environment.

[0110] Step 7, repeat steps 2 to 6, take each of the N+1 5G communication terminals as the current test terminal in turn, obtain test data, add the test data of the current test terminal to the first test data group, and add the test data of the remaining N 5G communication terminals to the second test data group.

[0111] This step rotates each of the N+1 5G communication terminals as the current test terminal, repeating steps 2 through 6 to fully collect performance data for all terminals across different test rounds and storing it in the first and second test data groups, respectively. This step ensures that every terminal can participate in the main test task while simulating the impact of other terminals acting as background load on the target network, providing comprehensive data support for network performance analysis in a multi-terminal environment.

[0112] The complete testing process is:

[0113] Test round initialization

[0114] Define N+1 test rounds. In each round, select one 5G communication terminal as the current test terminal, and the remaining N terminals as auxiliary terminals.

[0115] During the round switching, the roles of all terminals are swapped in sequence, ensuring that each terminal can participate in independent testing and background traffic generation.

[0116] Repeat steps 2 to 6

[0117] Select the current test terminal (step 2):

[0118] In the i-th test round, the i-th terminal (Ti) is selected as the current test terminal, and the remaining terminals (T1 to Ti-1 and Ti+1 to TN+1) are used as auxiliary terminals and connected to the relay equipment to generate background traffic.

[0119] Signal amplification and distribution of repeater equipment (step 3):

[0120] The relay equipment amplifies and distributes the signals of all auxiliary terminals, simulating signal expansion in a multi-terminal environment.

[0121] Perform a data transfer test (step 4):

[0122] The current test terminal communicates directly with the target network to complete the testing of key indicators such as throughput, latency, jitter and packet loss rate.

[0123] The auxiliary terminal generates background traffic to simulate actual multi-user scenarios.

[0124] Get the test data of the current test terminal (step 5):

[0125] All test data are collected from the current test terminal and stored in a first test data group.

[0126] Get test data for the auxiliary terminal (step 6):

[0127] The test data is collected from the auxiliary terminal and stored in a second test data group.

[0128] After each round, the data of the current test terminal is classified into the first test data group, and the data of the auxiliary terminal is classified into the second test data group.

[0129] Data is labeled by terminal number and test round to ensure accuracy and consistency in subsequent data analysis.

[0130] After all terminals have been used as current test terminals, all rounds of testing are completed.

[0131] For example, assuming there are three 5G communication terminals (T1, T2, T3) and two hardware relay devices, the specific rounds of testing are as follows:

[0132] Round 1: T1 is the current test terminal

[0133] Current test terminal:

[0134] T1 communicates directly with the target network, recording performance data such as throughput (250 Mbps), latency (15 ms), and packet loss rate (0.5%).

[0135] The data is stored in a first test data group.

[0136] Auxiliary terminal:

[0137] T2 and T3 were connected to the network through a relay device to generate background traffic, and the throughput (120 Mbps and 150 Mbps) and other indicators were recorded respectively.

[0138] The data is stored in the second test data group.

[0139] Round 2: T2 is the current test terminal

[0140] Current test terminal:

[0141] T2 direct test records performance data such as throughput (200Mbps), latency (20ms), and packet loss rate (1.2%).

[0142] The data is stored in a first test data group.

[0143] Auxiliary terminal:

[0144] T1 and T3 act as auxiliary terminals, generating background traffic, and recording throughput (140 Mbps and 170 Mbps) and other metrics, respectively.

[0145] The data is stored in the second test data group.

[0146] Round 3: T3 is the current test terminal

[0147] Current test terminal:

[0148] T3 direct test, recording performance data such as throughput (230Mbps), latency (18ms), and packet loss rate (0.8%).

[0149] The data is stored in a first test data group.

[0150] Auxiliary terminal:

[0151] T1 and T2 serve as auxiliary terminals, recording throughput (160 Mbps and 140 Mbps) and other indicators, respectively.

[0152] The data is stored in the second test data group.

[0153] Step 8: Analyze and determine the network performance based on the data of the first test data group and the second test data group.

[0154] In this step, by comprehensively analyzing the performance data of the first test data group and the second test data group, the hidden Markov model (HMM) is used to mine the hidden states in the time series data, revealing the state transition rules of network performance under different load conditions, and providing a basis for evaluating network adaptability and optimization strategies.

[0155] Traditional hidden Markov models (HMMs) are primarily suitable for univariate time series modeling, such as analyzing the temporal variation of a single performance metric (such as throughput). However, in this approach, network performance evaluation relies on multiple key metrics (such as throughput, latency, and packet loss rate) as well as background load conditions (performance data from auxiliary terminals). These metrics often have complex nonlinear relationships and dependencies, making it difficult for univariate HMMs to capture this interactive information.

[0156] To adapt to multivariate time series data, the core of the improved hidden Markov model in this invention lies in:

[0157] Jointly model multiple observed variables (such as throughput, latency, and packet loss rate) to capture the correlation between indicators.

[0158] Dynamically integrate background load information to make the relationship between hidden states and observed variables closer to real network scenarios.

[0159] Specifically, the following steps are included:

[0160] Step 8.1: Prepare and preprocess the data

[0161] Extract the time series data of the current test terminal (such as performance indicators such as throughput, delay, and packet loss rate) from the first test data group.

[0162] The performance data of the auxiliary terminal (such as background load flow, signal strength, etc.) is extracted from the second test data group as a reference for the background load condition.

[0163] Indicators such as throughput, latency, and packet loss rate are normalized to make their ranges consistent for ease of modeling.

[0164] If the acquisition times of the first test data group and the second test data group are not synchronized, an interpolation method or an alignment operation is used to ensure that the timestamps are consistent.

[0165] Step 8.2: Define the Hidden Markov Model Structure

[0166] Divide network performance into three hidden states:

[0167] : High performance state (high throughput, low latency, and low packet loss rate).

[0168] : Moderate performance state (moderate throughput, moderate latency, and moderate packet loss rate).

[0169] : Low performance state (low throughput, high latency, and high packet loss rate).

[0170] Define the observed variables:

[0171] The observation vector includes the throughput ( ), delay ( ) and packet loss rate ( ), forming a multidimensional time series:

[0172] Introducing background load conditions:

[0173] Use background load data (throughput or traffic intensity of auxiliary terminals) as a dynamic influencing factor: , which is the background load at time t.

[0174] Step 8.3: Initialize model parameters

[0175] Initial state distribution:

[0176] Set the initial hidden state probability. The initial state probability can be estimated based on the observed data distribution at the first time point, such as:

[0177] State transition probability matrix:

[0178] The state transition probability matrix A describes the transition rules between hidden states. It can count possible state transitions based on the performance change pattern of each two consecutive time points in the observation sequence. The transition probability matrix of the initialized hidden state is as follows:

[0179]

[0180] The sum of each row is 1, which represents the total probability of transitioning from one state to another.

[0181] Observation probability distribution:

[0182] Assuming that the observed variables follow a normal distribution, initialize the mean and covariance matrix:

[0183] in, represents the state at time t, j represents the state value, represents the mean vector of each hidden state; Represents the covariance matrix of each hidden state.

[0184] Step 8.4: Train the model

[0185] Model input:

[0186] Input multi-dimensional observation data (throughput, latency, packet loss rate) and background load data.

[0187] Training algorithm:

[0188] Use the Baum-Welch algorithm to train model parameters and update the state transition probability matrix , considering the dynamic impact of background load: in represents the probability of state i transitioning to state j, , is the transfer probability function under the influence of background load conditions, Represents the state transition probability matrix at time t, and updates the observation probability distribution parameters ( , ).

[0189] Convergence conditions:

[0190] Iterate the training until the log-likelihood value converges (e.g., the change is less than the set threshold).

[0191] Step 8.5: State decoding

[0192] Input data:

[0193] Use the observation sequence (test data) and the trained model parameters.

[0194] Decoding algorithm:

[0195] Apply the Viterbi algorithm to infer the hidden state sequence corresponding to each time point.

[0196] Result output:

[0197] Generate hidden state sequences that reveal the dynamics of network performance over time.

[0198] Step 8.6: Analyze hidden states and state transition patterns

[0199] State distribution analysis:

[0200] Count the proportion of each hidden state and evaluate the distribution of the network in high performance, medium performance and low performance states.

[0201] Example results:

[0202] High performance state : 60% of the time.

[0203] Medium performance state : 30% of the time.

[0204] Low performance state : 10% of the time.

[0205] State transition analysis:

[0206] Use the trained transition probability matrix to analyze the transition rules between states:

[0207] high performance( ) → Medium performance ( ): probability 0.15.

[0208] Moderate performance ( ) → Low performance ( ): probability 0.2.

[0209] Impact of background load on state transition:

[0210] When the background load is at its peak, → The transition probability of increases significantly, indicating that the network is prone to enter a low-performance state under high load.

[0211] Through these steps, a comprehensive time series analysis of test data based on an improved Hidden Markov Model (HMM) can be performed, revealing the hidden state of network performance and its dynamic transition patterns. Furthermore, embedding background load conditions makes the model more realistic, providing deep insights for resource optimization and performance improvement in 5G networks.

[0212] On the other hand, the present invention also provides a network performance testing system based on a 5G communication terminal, which uses the aforementioned method to perform network performance testing.

[0213] The prior art mentioned in the above background technology section and specific embodiments section of the present invention can be regarded as part of the present invention and used to understand the meaning of some technical features or parameters.

Claims

1. A network performance testing method based on a 5G communication terminal, characterized in that: The method comprises the following steps: Step 1: Configure N+1 5G communication terminals and N hardware relay devices; Step 2: Select one of the N+1 5G communication terminals as the current test terminal, and assign one-to-one correspondence between the remaining N 5G communication terminals and the N hardware relay devices; Step 3: Use the N hardware relay devices to amplify and demultiplex the signals of the corresponding 5G communication terminals; Step 4: Control the N+1 5G communication terminals to perform a data transmission test; Step 5: Obtain the test data of the current test terminal and add it to the first test data group; Step 6: Obtain test data of the remaining N 5G communication terminals and add them to the second test data group; Step 7: Repeat steps 2 to 6, taking each of the N+1 5G communication terminals as the current test terminal in turn, obtaining test data, adding the test data of the current test terminal to the first test data group, and adding the test data of the remaining N 5G communication terminals to the second test data group; Step 8: Analyze and determine the network performance based on the data of the first test data group and the second test data group.

2. A network performance testing method based on a 5G communication terminal according to claim 1, characterized in that: The 5G communication terminal is one or more of a smart phone, a fixed wireless access device, and an industrial-grade 5G module.

3. A network performance testing method based on a 5G communication terminal according to claim 1, characterized in that: The hardware relay device refers to a hardware device that can extend, amplify and distribute wireless signals.

4. A network performance testing method based on a 5G communication terminal according to claim 1, characterized in that: The current test terminal serves as an active test device to simulate user behavior; the remaining N 5G communication terminals are used to generate background traffic to simulate a multi-user scenario.

5. A network performance testing method based on a 5G communication terminal according to claim 1, characterized in that: The multi-path distribution includes: for each distributed signal, the relay device adjusts its directivity to ensure that the distributed signal coverage areas do not overlap, and sets different powers for each distributed signal to simulate user access environments of different strengths.

6. A network performance testing method based on a 5G communication terminal according to claim 1, characterized in that: Controlling the N+1 5G communication terminals to perform the data transmission test includes: The current test terminal sends large data packets to the target server or network to test the network's uplink throughput and stability; The current test terminal receives large data packets from the target server or network to test the network's downlink throughput and stability; The current test terminal uses small data packets to measure end-to-end delay and counts the changes in the time interval between the arrival of statistical data packets; The current test terminal sends a preset number of data packets, calculates the proportion of lost data packets, and evaluates the reliability of the network; Each auxiliary terminal sends and receives data according to a preset traffic model, simulating background user traffic in real scenarios; Different auxiliary terminals are set with different service types and data rates; The auxiliary terminal simulates traffic changes in different time periods.

7. A network performance testing method based on a 5G communication terminal according to claim 1, characterized in that: The test data includes: throughput, delay, jitter, packet loss rate, and signal quality.

8. A network performance testing method based on a 5G communication terminal according to claim 1, characterized in that: The step 7 comprises: After each round, the data of the current test terminal is classified into the first test data group, and the data of the auxiliary terminal is classified into the second test data group; The data is marked by terminal number and test round. After all terminals are used as current test terminals, all rounds of testing are completed.

Images