CN100372269C - A method for performance testing of wireless communication products - Google Patents

Abstract

The invention relates to a method for performance testing of wireless communication products. Including the following steps: (1) Select a test tool that meets the requirements according to the test requirements and the characteristics of the tested object; (2) Compile the network diagram and description of the performance test; (3) Determine the traffic model; (4) Write the test suite and (According to the well-written test suite) to debug; (5) Conduct comprehensive performance test on the tested object. The business processing capability and capacity of the switching subsystem can be tested from different levels by concurrent multiple requests, or the performance of the system can be tested by creating anomalies under the load background, so the effect of perfect testing can be achieved. In addition, due to the adoption of normative The traffic model makes the test results standard and comparable.

Description

A method for performance testing of wireless communication products

technical field

The invention relates to a method for performance testing of wireless communication products.

current technology

Wireless communication systems often provide multiple call services for multiple mobile users macroscopically, and handle multiple services or processes at the same time. Therefore, in order to ensure the normal operation of the wireless communication system, an effective method is needed to detect the parallel processing capabilities of the system. , including the capacity of the system, the stability of the system under a certain load, etc.

In recent years, the switching subsystems used in my country's GSM and CDMA mobile communication systems have established corresponding technical standards and test specifications based on these standards. Among them, there are test items for testing switching subsystem performance such as "800MHz CDMA Digital Cellular Mobile Communication Network Equipment General Test Specification: Switching Subsystem Part", but because the test method is only to use the analog pager to make repeated and cumulative calls, the test method has the disadvantages that the test is not perfect, and the test results are not standard and comparable , In addition, the measurement and comparison of system performance is not sufficient, and the processing capability test does not really test the actual operation of the system under test.

Contents of the invention

The present invention aims to solve the technical problems that the existing method for performing performance testing on wireless communication products is not perfect, and the test results are not standardized and comparable.

In order to solve the above technical problems, the technical scheme that the present invention takes is:

A method for performance testing of wireless communication products, characterized in that it comprises the following steps:

1. Select test tools that meet the requirements according to the test requirements and the characteristics of the measured object;

2. Compile performance test network diagram and description;

3. Determine the traffic model;

4. Write a test suite and debug it according to the written test suite;

5. Carry out a comprehensive performance test on the tested object.

The comprehensive performance test in the fifth step includes six aspects: performance index test, stability test, reliability test, boundary value test, system function test and maintainability and testability test.

The performance index tests described include

a. A variety of traffic models are commonly seen on the Internet, and normal simulation tests of various networking methods;

b. The ultimate capacity and ultimate performance test of the whole system and each subsystem;

The stability test includes a long-term stable operation test and a long-term performance index test;

The reliability test includes a system failure test under a load background and an abnormal flow test under a load background;

The boundary value test is a load test carried out under the conditions of the boundary conditions to achieve the boundary conditions configured by means of data configuration;

The system function test includes the test of routing selection, dynamic resource allocation and load sharing optimal algorithm under the background of traffic, the optimization protection measures of the system under the background of traffic and the test of the functions that need to be tested under load;

The maintainability and testability tests described include

a) Data configuration, maintenance, and alarm testing;

b) Testing of debugging information and software parameter testing platform;

c) Data recovery and version upgrade testing;

d) The main single-process test.

Described step one comprises the following steps:

(1) According to the characteristics of the test requirements, select a test tool that can meet the characteristics of the test requirements. The test tool must be able to realize the main process of the client and be able to count the number of successes and failures;

(2), calculate the capacity of the measured object and the peripheral nodes of the measured object, select the test tool that can meet the capacity of the measured object from the test tools that can meet the characteristics of the test requirements, and configure the peripheral node equipment that meets the capacity of the measured object.

Described step 2 comprises the following steps: (1) determine office direction number, size, carry out the signaling point numbering and circuit of each office direction or network node, number section, speech path, the allocation of link resources; (2) Draw a test network diagram and indicate the network nodes simulated by the test tool; (3) Realize the content of the test network diagram design in the data configuration of the test tool, match the data configuration of the device under test, and prepare for the docking test .

The method for determining the traffic model in the described step 3 is: total call times L=W×T1; total location update times M=W×T2; total switching times N=W×T3; The number of call attempts during busy hours = L×E1/(E1+E2+2×E3); the number of call attempts from mobile to fixed when busy = L×E1/(E1+E2+2×E3); the number of call attempts from mobile to mobile when busy Number of times = L×E1/(E1+E2+2×E3), where W is the system user capacity, T1 is the average number of trials when mobile users are busy, T2 is the number of location updates, T3 is the number of handovers between BSCs, and E1 is the fixed E2 is the proportion of mobile calls to fixed traffic, and E3 is the proportion of mobile calls to mobile traffic.

Described step 3 when determining the traffic model, the load of average busy hour of mobile user is: 0.018-0.03Er1/user; The average number of times T1 of trial call times when mobile user is busy is in the scope of 1-3.6, and the scope of location update times T2 is 2- 6. The range of switching times T3 between BSCs is 0.5-1; the range of traffic distribution ratio for fixed calls to mobile E1 is 25%-40%, the range of mobile calls to fixed E2 is 30%-50%, and the range of mobile calls to mobile E3 is 10 %-45%.

After adopting the above technical solution, it is possible to test the business processing capability and capacity of the switching subsystem from different levels by concurrent multiple requests, or to test the performance of the system by creating anomalies under the load background, so that the test can achieve the effect of perfection , In addition, due to the adoption of a normative traffic model, the test results are standardized and comparable.

Description of drawings:

Fig. 1 is an overall flow chart of a specific embodiment of the present invention.

Fig. 2 is a detailed test item of a certain product utilizing the present invention.

Detailed ways

The meaning of the abbreviations used in this article:

BHCA Busy Hour Call Attempts Call attempts when busy

ERL Ireland

3G 3rd Generation The third generation of mobile communication system is referred to as WCDMA in this article

2G 2rd Generation The second generation of mobile communication system is referred to as GSM in this article

CDMA Code Division Multiple Addressing code division multiple access mobile communication system

BSC BaseStationController 2G base station controller

MSC A second generation Mobile Services Switching Center that only supports the A interface 2G Mobile Switching Center

RNC Radio Network Controller 3G Wireless Network Controller

Iu Interface between the RNS and the core network Iu interface

VMSC Visit Mobile Switching Center Visit Mobile Switching Center

GMSC Gateway Mobile Switching Center Gateway Mobile Service Switching Center

HLR Home Location Register Home Location Register

SCP Service Contral Point Business Control Point

MAP Mobile Application Part Mobile Application Part

TUP Telephone User Part Telephone User Part

BAM Back Administration Module Back Administration Module

BAU Bill Administration Unit Bill Management Unit

1. Select the test tool that meets the requirements according to the test requirements and the characteristics of the measured object

1. Analyze and select the characteristics of tools that can meet the testing requirements

The performance testing tool can simulate various client entities concurrently running multiple services, that is, a message transceiver. The test tool must be able to realize the main process of the client and support the corresponding protocol. It should be possible to customize the process and realize the configuration of each process data. It can realize flexible configuration of various process ratios. It can realize the statistics of each checkpoint of each load process, at least the number of successful failures can be counted. It can track and record exceptions to specified processes or users. In addition, the performance and capacity of the load tester itself must exceed at least three times the expected capacity of the entity under test. The load tool must have strong stability, be able to handle abnormal situations, and send and receive messages following standard protocols. The device should be as compact as possible, and the man-machine interface should be friendly. Currently available commercial testers include MGTS (TEKELEC Company), Pager (AMERITEC Company), and 8610 (ACTERNA Company). In order to reduce the cost, the test tool can be developed by itself. Its basic functions must realize the simulation of real switching network nodes to send and receive a large number of messages. There are many achievements in this regard.

2. Calculate the capacity of the measured object and the surrounding nodes of the measured object, and select the test tools that can meet the capacity requirements of the measured object and the surrounding nodes of the measured object from the test tools that can meet the characteristics of the test requirements.

In the wireless network, the switching nodes are not independent, and must be connected with other nodes such as BSC, VMSC, GMSC, HLR, SCP, etc. Strictly speaking, if the tester can completely simulate these nodes and the call volume can meet the design capacity of the tested object If the call volume is more than twice that, it is best to use a tester. But in most cases, the tester cannot completely replace other nodes, because sometimes the call volume, links, and configurations are not enough, so the surrounding environment needs to be built separately during the test. Then it is necessary to consider in advance whether the performance of these environments can meet the test needs. The calculation method is whether the number of users supported by each node matches the object under test, or whether its configuration can meet the maximum message flow. If you encounter this situation when testing smart services, the tester may not be able to provide SCP nodes that meet the test capacity, and the SCP must be real. The user data of the smart business is stored in the SCP server. During the load test, the SCP must provide the batch account opening function, otherwise the test cannot be performed. Similar to HLR, before using real database-type equipment to form a network, it is necessary to estimate its performance, how many users it can support, and how much call intensity (BHCA). When testing, always pay attention to the changes of its CPU, so as not to affect the test due to performance problems. If it becomes a test bottleneck due to the small capacity of surrounding nodes, measures should be taken immediately to expand the capacity of node equipment.

2. Compile performance test network diagram and description

The principle of networking is to be as close as possible to the actual deployment situation. The device will encounter various complex networks on the Internet, and it is impossible to be completely consistent with the Internet in the laboratory, and the data is too complicated to write test cases and cause trouble in maintaining data. At this time, it is necessary to analyze the characteristics of the online data as much as possible, and simulate the online situation in a targeted manner. The number, size, and especially capacity of office routes must be greater than or equal to the maximum configuration for online deployment. Sometimes it is necessary to consider the changes in the initial stage, the first stage and the second stage of the project. For example, it is impossible to complete the transmission circuit in one step at the beginning of the initial deployment. Generally, all office routes are only opened in the first three modules, and then some circuits are added. When networking, it is not possible to make the data uniform for the convenience of writing test suites (tester data), but it is almost impossible to meet in practice, and large-capacity local and small local directions are also required when performing extreme tests. Therefore, to draw the performance test network diagram and number it, the following tasks must be completed: A. Determine the number and size of office directions, and carry out the numbering of signaling points and circuits, number segments, voice paths, links and other resources of each office direction or network node. Assignment; B. Draw a test network diagram and indicate the network nodes simulated by the test tool; C. Realize the content of the test network diagram design in the data configuration of the test tool, and match the data configuration of the device under test, for the docking test be prepared.

3. Determine the traffic model

The traffic model comes from the traffic statistics data on the Internet. Only long-term statistics on user behavior can truly characterize the characteristics of an office user. For the laboratory, we basically adopt the standard user model provided in the technical specification. For example, the traffic model of an end office VMSC1 with 240,000 users is as follows, and the average hourly load of mobile users: 0.03Er1/user

Average call time: 60 seconds

Average call attempts: 1.8 times/busy hour

Location update: 2 times/busy hour

Handover (between BSCs): 0.5 times/busy hour

Traffic distribution: fixed to mobile 40%, mobile to fixed 50%, mobile to mobile 10%

Proportion of mobile subscribers calling in long-term city: 1:3

According to the actual operation experience of the telecommunication network, the traffic models are different in different countries, different regions, different development stages, and different mobile networks for different purposes. Various characteristics will change within a certain range. The average busy-hour load of commonly used mobile users is generally 0.018-0.03Er1/user. This index only refers to the index of calls, and has the following relationship with the average trial call:

Average load of mobile users during busy hours = average call attempts × average call duration (seconds)/3600.

If the user capacity of the system is W, the average call attempt of mobile users when they are busy is T1, the location update is T2, the handover (between BSCs) is T3, the traffic allocation ratio is E1 for fixed calls to mobile, E2 for mobile calls, Mobile hit Mobile as E3. So how should we implement this traffic model when it comes to testing?

First, the number of mobile users determines the capacity of the switching subsystem. If the user capacity of the system is W users, it means that there are W mobile users with the above standard features. The standard features are determined by the main user services that occur when the mobile network is normally busy. , the number of occurrences of this service during busy hours is much greater than that of other services, so it should be counted. It should be converted into the number of BHCA calls to different call types from BSC and PSTN offices during implementation. First calculate the total BHCA, the calculation formula is as follows:

The total call BHCA order is L, L=W×T1

The total location update BHCA is set to M, M=W×T2

The total switching BHCA order is N, N=W×T3

Taking the end office VMSC1 as an example, the calculation result is

The total call to BHCA is L=240000×1.8=432000

The total position update BHDA is M=240000×2=480000

The total switching BHCA is N=240000×0.5=12000

Since there are different call situations under the call service, L is allocated according to the proportion of traffic flow, and the calculation formula is as follows,

Fixed call and mobile BHCA＝L×E1/(E1+E2+2×E3)

Mobile and fixed BHCA＝L×E1/(E1+E2+2×E3)

Mobile call mobile BHCA＝L×E1/(E1+E2+2×E3)

Taking the end office VMSC1 as an example, the calculation result is:

Fixed and mobile BHCA＝432000×4/11＝157091

Mobile and fixed BHCA＝432000×5/11＝196364

Mobile and fixed BHCA＝432000×1/11＝39273

Then M, N, L×1/11, L×5/11 are distributed into respective call volumes according to the ratio of call relays (GSM, CDMA) or RNC occupied by each BSC to call resources (WCDMA); ×4/11 Allocate the call volume according to the proportion of each PSTN office to the trunk. If the proportion of long-city calls is required, the allocation of resources to bureaus of long-city calls should also be proportionate to ensure that the resources are sufficient. When calculating the traffic model, you need to pay attention to two points:

1. In different regions, different stages, and mobile networks for different purposes, the services are very different, and new services will continue to increase in the future. It is possible that T1, T2, and T3 may not fully represent the main user services. T4, T5, etc. appear, but in the test of the mobile gateway office, T2 and T3 do not exist, and only need to calculate the call service T1. The general range of T1 is 1-3.6, the general range of T2 is 2-6, and the general range of T3 is 0.5-1.

2. The L that represents the total call BHCA refers to all mobile users, including the calling user and the called user. As long as the mobile calls the mobile phone, the call of two mobile users will be generated, so that the total call BHCA of all mobile users can be guaranteed. It is L. If there is no movement and movement, the score will be directly proportional. E1 generally ranges from 25-40%, E2-50%, and E3 generally ranges from 10-45%.

The traffic model calculated by the above method is a standard traffic model, that is, the general situation when the device network is running normally and busy. It is representative and comparable between products, and is of great significance for performance testing.

4. Writing test suites and debugging

After the load test is started, the estimated time for the system test plan should be determined first. If there is no network that can stably call, then the preparation of environmental materials, data and tester test data will take at least 1/5 of the entire test process. Tester test data is also called a test suite. The test suite determines the load call model, call type, and call volume. The test suite should be clearly described in the test plan. All test suite details should be documented.

Debugging the written test suite should start with a single process, and then debug the load process after debugging, and also debug one by one, and run all the test data until there is no problem, then you can call the global.

5. Carry out comprehensive performance test on the tested object

The comprehensive performance test includes the following six aspects, and the specific operation is different according to the characteristics of different products. The execution sequence of the six test methods of performance index test, stability test, reliability test, boundary value test, system function test, maintainability and testability test can be determined according to actual needs, as a complete test should all be Execute it again. The following describes the test purpose and checkpoints of various test methods to help understand the test methods. And use the test sub-item as an example to illustrate the specific content, as shown in Figure 2, it is a detailed test item for a certain product.

1. Performance index test

1) A variety of traffic models are commonly seen on the Internet, and normal simulation tests of various networking methods.

Purpose:

Investigate whether the system detection system can support normal traffic services, and whether various resources can be allocated and managed normally. Whether it can work normally under various traffic models, and whether its performance is up to standard. Investigate the performance differences of the system under various models and networking structures, thereby analyzing system defects, and giving suggestions for system optimization networking and data configuration.

checking point:

Resource occupancy and release, performance indicators, communication traffic, and monitoring and result information generated under load, such as bills, alarms, and common traffic statistics. Internal message sending and receiving statistics.

Test subproject:

Standard traffic model test, networking mode test

2) The ultimate capacity and ultimate performance test of the whole system and each subsystem.

Purpose:

Investigate the problem of subsystem communication coordination and whether the communication protocol is reasonable.

Test the system as a whole to find system bottlenecks.

Investigate the efficiency of system process scheduling and other scheduling algorithms.

Examine loop control problems, sorting, and implementation of search algorithms.

Examine resource management efficiencies such as speed and space redundancy. In the case of load impact, whether the system resources will be abnormal.

Investigate the self-protection ability of the system, such as whether the congestion control is normal, resource inspection and other issues.

Investigate the release and leakage of buffers (memory, timers and other resources that can be recycled and reused).

Investigate errors in concurrency processing, such as mutual exclusion of critical resource access and synchronization issues.

Increases the probability of certain random issues appearing.

checking point:

Resource occupancy and release, memory occupancy, performance indicators, communication traffic speed, correct rate of data transmission, alarms, traffic statistics, and major exceptions in bills. CPU and space usage of BAM and BAU etc.

Test subproject:

Service processing limit test, signaling processing limit test, communication flow test, resource box limit test, table resource limit test, user plane limit test, switching capacity limit test, global limit test, flow control test

2. Stability test

1) Long-term stable operation.

It is suggested that the test can be carried out according to the traffic law of the online statistics, the traffic curve has peaks and valleys, or the traffic load according to the standard busy hour, and the time is more than 72 hours.

Purpose:

Investigate the stability of the long-term operation of the system, that is, accumulate small probability events through time and load.

checking point:

Resource occupancy and release, call loss indicators, connection rate, failure process and reasons.

Test subproject:

Long time test, combined call test

2) Long-term performance index test.

Purpose:

One is the system's call loss and other important indicators under the standard traffic model, compared with similar products. The second is the performance stability index of the system under the super-large traffic model.

checking point:

Call loss, link load, subsystem CPU usage, connection delay, bill traffic.

Test subproject:

performance index test

3. Reliability test

1) System failure test under load background.

Purpose:

Investigate the anti-abnormal characteristics and self-healing ability of the system. Detect whether the software system can perform fault-tolerant processing when various hardware faults occur, and minimize the loss caused by the fault. Detect the ability of the system to prevent catastrophic accidents, such as backup speed capacity, switching speed, recovery ability and other issues.

checking point:

Resource occupancy and release, performance degradation or improvement, recovery time, and major exception alarms. Test subproject:

System failure and switching, traffic switching test

2) Abnormal flow test under load background.

Purpose:

Investigate release leaks of buffers such as memory and timers that can be recycled and reused.

Loop control problem.

Investigate process concurrency issues, such as whether it will affect normal processes.

1 Checkpoint:

Resource release, performance degradation or improvement, memory usage, major exceptions.

Test subproject:

TUP message exception flow, MAP message exception flow, Iu interface message exception flow, RNC exception

4. Boundary testing

1) Reach the configured boundary conditions by means of data configuration, or reduce the boundary value, and carry out load testing under these conditions.

Purpose:

Expose loop control issues.

Investigate whether the size of resource allocation is reasonable and other issues.

Validation of boundary values.

checking point:

Implementation of border-related functions, search speed, resources, performance degradation or improvement, major anomalies.

Test subproject:

Main design indicators of IU-CS interface

5. System function test

1) Tests on the implementation of optimal algorithms such as routing selection, dynamic resource allocation, and load sharing under the background of traffic.

Purpose:

Investigate the rationality and error-freeness of various optimization algorithms of the system, combined with abnormal test, such as routing.

checking point:

Function realization, performance improvement, resources, abnormal data configuration and processing.

Test subproject:

Signaling load sharing

2) The optimization protection measures of the system under the traffic background and the tests of other functions that need to be tested under load, such as calibration, voice quality, echo suppression, etc.

Test subproject:

Resource verification, voice quality, echo suppression

6. System maintainability and testability test

1) Tests on data configuration, maintenance, alarms, etc.

Since the performance test environment is often the most complete in the experimental environment and the closest to the online running configuration, this feature can be used to find problems in data configuration, maintenance, and alarms during the construction of the environment and operation and maintenance. However, the discovery of these problems requires It depends on the testers' mastery of the basic principles of data configuration, maintenance, and alarms and their sensitivity to problems.

Test subproject:

Query resource CPU usage

2) Testing of test platforms such as debugging information and software parameters

The testing methods provided by the system to users are limited, and at the same time, it also provides more in-depth testing methods for testers, such as debugging information and software parameters, but the correctness and practicality of these methods themselves also need to be tested.

Test subproject:

track

3) Data recovery and version upgrade test

Purpose:

The data configuration of the performance environment is large and involves many aspects, which can comprehensively examine the system version upgrade and data recovery functions.

checking point:

Function realization, abnormal data configuration processing.

Test subproject:

Data and Upgrades

4) Main single-process test

The prerequisite for the correct execution of the test item is whether the single process of the tester is correct. This part of the work cannot be omitted. The test item is always executed first, and different testers are debugged separately.

Test subproject:

Performance Tester Single Process

As long as the operation steps are followed, the tester can conduct a complete simulation test. The above description of a complete set of methods enables those of ordinary skill in the art (especially for testers) to use the present invention.

Claims (8)

1. A method for performance testing of wireless communication products, characterized in that it comprises the following steps: 1. Select the test tool that meets the requirements according to the test requirements and the characteristics of the tested object; 2. Compile performance test network diagram and description; 3. Determine the traffic model; 4. Write a test suite and debug it according to the written test suite; 5. Carry out a comprehensive performance test on the tested object. 2. The method for performing a performance test on a wireless communication product as claimed in claim 1, wherein the comprehensive performance test in said step 5 includes performance index test, stability test, reliability test, boundary value test, There are six aspects of system function test and maintainability and testability test. 3. the method that wireless communication product is carried out performance test as claimed in claim 2, it is characterized in that, Described performance index test comprises a. common multiple traffic models on the Internet, the normal simulation test of multiple networking modes; b. the limit capacity and limit performance test of the whole system and each subsystem; The stability test includes a long-term stable operation test and a long-term performance index test; The reliability test includes a system failure test under a load background and an abnormal flow test under a load background; The boundary value test is a load test carried out under the conditions of the boundary conditions to achieve the boundary conditions configured by means of data configuration; The system function test includes the test of routing selection, dynamic resource allocation and load sharing optimal algorithm under the background of traffic, the optimization protection measures of the system under the background of traffic and the test of the functions that need to be tested under load; The maintainability and testability tests include a) tests on data configuration, maintenance, and alarms, b) tests on debugging information and software parameter test platforms, c) tests on data recovery and version upgrades, and d) tests on main unit Process testing. 4. The method for performance testing of wireless communication products as claimed in claim 1, 2 or 3, wherein said step 1 comprises the following steps: (1) According to the characteristics of the test requirements, select a test tool that can meet the characteristics of the test requirements. The test tool must be able to realize the main process of the client and be able to count the number of successes and failures; (2), calculate the capacity of the measured object and the peripheral nodes of the measured object, select the test tool that can meet the capacity of the measured object from the test tools that can meet the characteristics of the test requirements, and configure the peripheral node equipment that meets the capacity of the measured object. 5. The method for performance testing of wireless communication products as claimed in claim 1, 2 or 3, characterized in that said step 2 includes the following steps: (1) determine the number and size of office directions, and carry out each office direction or The signaling point numbers of network nodes and the distribution of circuits, number segments, voice paths, and link resources; (2) draw a test network diagram and indicate the network nodes simulated by the test tool; (3), in the test tool In the data configuration, the content of the test network diagram design is realized, which matches the data configuration of the device under test, and is ready for the docking test. 6. The method for performing a performance test on a wireless communication product as claimed in claim 4, characterized in that said step 2 comprises the following steps: A, determine the number and size of office directions, and carry out the signaling of each office direction or network node Point numbers and circuits, number segments, call paths, and allocation of link resources, B. Draw a test network diagram and indicate the network nodes simulated by the test tool; C. Realize the test network diagram in the data configuration of the test tool The content of the design matches the data configuration of the device under test to prepare for the docking test. 7. as claimed in claim 1 or 2 or 3 described wireless communication products are carried out the method for performance testing, it is characterized in that, the method for determining traffic model in the described step 3 is: total number of calls L=W*T1; The total number of location updates M=W×T2; the total number of switching N=W×T3; the number of call attempts when the fixed-to-mobile call is busy in traffic distribution = L×E1/(E1+E2+2×E3); the mobile-to-fixed call The number of call attempts during busy hours = L×E1/(E1+E2+2×E3); the number of call attempts when the mobile phone is busy = L×E1/(E1+E2+2×E3), where W is the system user capacity , T1 is the average number of attempts of mobile users when they are busy, T2 is the number of location updates, T3 is the number of handovers between BSCs, E1 is the proportion of fixed-to-mobile traffic allocation, E2 is the proportion of mobile-to-fixed traffic allocation, and E3 is The proportion of mobile-to-mobile traffic allocation. 8. The method for performance testing of wireless communication products as claimed in claim 7, characterized in that, when determining the traffic model in said step 3, the average busy hour load of mobile subscribers is: 0.018-0.03Erl/user; The average number of call attempts T1 ranges from 1-3.6 for mobile users during busy hours, the range of location update times T2 ranges from 2-6, and the range of T3 switching times between BSCs ranges from 0.5-1; the range of traffic allocation ratio for fixed calls to mobile E1 is 25% -40%, the range of mobile vs fixed E2 is 30%-50%, and the range of mobile vs mobile E3 is 10%-45%.

Images