Anyone who sees the "pass" message displayed on the tester screen will face a problem behind these two words: data management. Because we know that a passing result is obtained after a series of complicated measurements and calculations, and this process is more complicated year by year. With the launch of the first portable cable tester in 1988, the link test and test report began to change greatly. Microtest's cable scanner has the function of locating cable open circuit, which was mainly used for coaxial Ethernet at that time. Storing data and generating test reports are not supported. When a problem is found, it is impossible to show exactly what was tested. When you need to record, you can only manually copy the problems or faults you care about. In 1990, a pair of scanners are introduced to test "new" and "high-speed" 4Mbps token ring and 10 Mbps 10 Base-TTP network. This is the earliest product that can support preliminary data storage and text result overview. However, there is still a lack of time stamps, and cables can only be numbered with simple numbers to provide a basic summary of attenuation and length. NEXT Scanner, a crosstalk scanner introduced later, increases and enhances the testing ability, such as near-end crosstalk, wiring diagram, resistance, etc., but the reporting ability remains basically unchanged. The report to the user can be brought out, or it can be an electronic file in text form, and the test location and user information can be input after the test results. Unfortunately, this method cannot avoid "occasionally" tampering with the test results. With the increasing demand for certification testing, 1993 appeared the first PentaScanner in the industry. Soon after, Fluke DSP- 100 came into being. These instruments have brought many new tests with time-stamped results, and can store up to 500 linked detailed test results in the form of text tables. These results show whether the test follows the existing standards at that time, and ensure that each link can meet the requirements of the standards. The application software attached to the instrument is used to record the test results in the database. You can classify, view, report overview and detailed report the database. Regardless of the additional functions brought by these softwares, the test results are still arranged by links. The integrity of the data has been improved, because the real data used to prove compliance with the standards has been submitted to the users. The application software does not support editing test data. For some users who need data analysis functions other than basic pass/fail, or generate personalized report formats, the ability to output data in CSV format is added later. Because this data format provides the greatest flexibility, the "accidental" problem of modifying data may occur as before. That's not enough. All manufacturers quickly developed optical cable test adapters for the tester. These adapters are usually connected to the top of the tester to test the attenuation and length of multimode or single-mode optical cables according to existing standards, and mark the test results with pass/fail information. OMNIScanner introduced by Fluke Network in 1998 and DSP-4000 greatly increased the amount of data to be stored after it supported six standards that were still in the draft at that time. It is necessary to speed up data transmission and express data better. For measurement data, before category 5, less than 50 sets of independent data are needed, while before category 5, less than 200 sets of data are needed for testing, and more than 400 sets of data are needed for testing in categories 5 and 6, which is more than 8 times that before category 5. For these data, pass/fail is no longer the only concerned result. Now we are concerned about the margin, not only at the worst point, but in the whole measurement range. In order to effectively store a large number of line pair data and drawing data, Fluke Networks DSP-4300 and OMNIScanner2 both support removable memory cards. In 2002, Fluke Networks launched LinkWare, a comprehensive data reporting and management software, which is compatible with existing testers in more than 80% markets. LinkWare greatly simplifies the data management of integrators by reducing the number of software needed to support various testers. LinkWare provides color graphic report, which is greatly improved on the basis that only the worst gross profit result is given in each measurement. It is much easier to read the information provided by graphic reports than to read text ASCII reports or CSV reports that require great patience. In any case, the challenge of managing large amounts of data remains. When all the measurement results are stored, the header information of some columns is also stored. Testers before category 5 usually store some basic information to mark connections and testers. Due to the massive installation of Class 5 connecting roads and the requirements of TIA 606, the records of who, what, place and time have been added. This includes operator, serial number of tester, software version, floor, building, cable type and NVP value, test adapter, time and date, etc. With the development of TIA 606 standard, the test result information at both ends of each link is also increasing. Table 1: continuous data recording-it has been about 10 years since the amount of information to be recorded in each report was never recorded until a large number of results needed to be stored. For example, 1000 test reports require more than 4 19000 pieces of data! The data is overwhelming. After completing a project, users have to face a stack of printed reports or a CD. Who wants to recite a lot of reports? Who wants to read hundreds of test results carefully with unfamiliar software? Faced with so much data, it is difficult to look at the trend or find out the abnormal problems: is the next margin consistent or is the special component better?

what's up These reports can not only prove that the installed links meet the standard requirements of Category 5 or Category 6. They are also tools to help you get more business, improve efficiency, improve product quality, improve service level, check the level of installers and improve and maintain customer satisfaction. There are four main user groups of field test information: installers, consultants, end users and cable manufacturers. Every user has his own special requirements. Installers/integrators are in a competitive environment. For attractive industries, it is a continuous challenge to provide educated and trained installers and products and processes with quality assurance at reasonable profit prices. Integrators may be the users who have the greatest demand for on-site testers, and usually use them for 6-7 days a week. They are also particularly concerned about the following questions: 6? 1 Will I win more projects?

6? 1 Are the termination levels of all installers consistent? In many projects, the integrator can only receive the project payment after completing all the installation work and the user obtains the warranty certificate from the original manufacturer. Sometimes it takes four weeks just to prove the warranty. Consultants usually specify construction requirements, including mandatory tests and report formats. For some important links, it may also be required to ensure that the next margin exceeds a certain level of the standard limit, to ensure the minimum ACR (margin), or to exceed the frequency range specified by the standard. At the same time, the frequency of consultants using test instruments is not as high as that of other users. They are in an influential position because they usually ask for what is necessary. They also pay special attention to the following issues:

6? Are there any unknown anomalies in the data of 1? Manufacturers manufacture and supply cables and connectors, and grant and undertake quality assurance requirements. The needs of manufacturers are different from those of other users, and usually need more detailed analysis in order to maintain and improve product quality and efficiency. Their needs include:

6? 1 When fault diagnosis is needed, how to quickly find the root of the problem and separate normal and abnormal data? There are several alternatives to solve the problems listed above. Including manually checking the test results, analyzing the detailed test results output in CSV format, or using special software to automatically analyze the results, such as Fluke Networks LinkWare Stats statistical reporting software. The traditional way is manual analysis. In the field, small projects (below 100) usually print and bind test reports. For large-scale projects, the outcome documents will be stored on CD-ROM in the form of electronic files. It is unrealistic to analyze the data comprehensively in an acceptable time. Check randomly selected reports whenever possible. If the anomaly is detected first, we should make an in-depth analysis. If nothing unusual is found, all we can do is hope that everything is normal. Another manual method is to analyze the detailed data exported in CSV format. Most cable record management software includes the option of exporting all data items and test records in CSV format: each test record is output in the form of rows or columns. CSV may be difficult to read, but it is very easy to analyze with standard applications. Such as spreadsheet programs or internally developed software. Using this method, it goes beyond simple pass/fail analysis and can meet the requirements of different types of users for field test information:

6? 1 Integrator can determine whether the installer has tested all the links according to the required test settings.

6? 1 Consultants can analyze parameters not required by TIA standard (such as insertion loss in actual copy length).

6? 1 The manufacturer can sample and detect the long-term average.

6? End users of 1 can view the average and worst margins of important links. Although CSV can be used for valuable in-depth analysis, there are still many problems. As mentioned in the first part of this paper, a large amount of stored data will cause the number of data items to exceed the maximum limit that a standard spreadsheet program can import (255). LinkWare test result management software, such as Fluke Networks, allows editing to a certain extent to select items to be exported. However, if you need to export all the data items to get all the information, you can only export multiple files and analyze them many times, which makes the process more complicated. Another potential problem is that CSV data items may be changed (such as adding new data items) due to the requirements of testing new versions of software or standards. These changes will have a certain impact on the output format. This will force the process and user-developed software to be modified accordingly. The result is that it can be used normally now, and the tester software may not work normally after upgrading. CSV data can also be edited and tampered with at will. When the data is output in text format such as CSV, it is easy to modify the test results intentionally or unintentionally, and the authenticity of the data will be doubted. Automatic analysis programs, such as LinkWare Stats report statistics software of Fluke Networks, can provide more convenient and powerful test data analysis ability, and also bypass the shortcomings of using CSV output. LinkWare Stats can give a performance overview of10,000 points on a piece of paper. It can quickly point out the differences of performance parameters and settings in the whole installation project (see Figure 3). Figure 3: Look at the whole test result from two aspects. In addition to the overview information, looking at the data in the whole range of storage results will find valuable information. The commonly used overview tool is a column chart, which shows the number of tests on the Y axis and the required test result margin on the X axis. In Figure 4, you can see the distribution of the next margin of the 1674 permanent link. Note that the results are on the normal distribution curve, and it can be considered that the product and the termination process are consistent. Link headroom less than 1dB is displayed in yellow. Any failed links will be displayed in red. Figure 4:4: The next margin histogram-normal distribution is shown in Figure 5, and we see the result moving to the right. Obviously, the average profit rate in this example is much better, and it is a higher quality link. Figure 5: Next Tolerance Bar Chart-Better Average Tolerance Figure 6 shows evidence of inconsistent termination process, as it is known that the same cables and connectors are used throughout the project. This is a signal to check the next allowance according to the installer or location (floor, building, etc.). Someone may need further training. Figure 6: Next margin bar chart-Inconsistent margins are based on line pair combination or frequency. It is also meaningful to further analyze the performance test results at both ends of the link. Fig. 7 plots the worst return loss tolerance according to frequency. We see that most of the worst return losses occur in the high frequency band (the results of more than 350 lines are between 225 and 250MHz). It is recommended to consult the experts of connectors, because the mismatch between connectors and cables affects the return loss margin in this example. Figure 7: Histogram of return loss versus frequency. A complicated problem surrounding management data mentioned in the first part of this paper is "false pass". When invalid settings are used to create the illusion of passing, "false passing" will occur, which should not happen. A typical example is to use the settings of Class 5 basic links (adapters, test standards) to test super Class 5 permanent links. To find such problems, we need to look at the test results according to the tester, software version, connection adapter and so on. This can check the test settings to ensure that all conditions are legal and valid. There are many problems in the example provided in Figure 8: using the wrong tester, outdated software version, wrong adapter and inconsistent NEXT margin. Without such a simple overview method, it will be a long time to manually find out these problems from hundreds of test reports. Figure 8: Tester overview shows a contradictory summary. With the continuous development of cables, connectors and test equipment, the demand for analyzing the growing test data is also increasing. This trend continues. The Standards Committee is discussing the new performance and test requirements of the next generation wiring system. The data storage requirements of installed links are getting higher and higher. Automatic analysis methods such as LinkWare Stats provide a brand-new scheme for checking data. Through the statistical analysis of the test results of the whole project, the data is easier to understand, the workload is reduced, and the shortage of using CSV format data is avoided. Integrators, consultants, end users and manufacturers all have different needs. Automatic statistical analysis can find "hidden" information in depth test results, ensure the correct use of the tester, and gain the necessary in-depth understanding of link performance and installation technology. LinkWare report management software can be downloaded from the Chinese website of Fluke Network Company.