With the advent of new technologies such as high-speed downlink packetaccess and high-speed uplink packet access, RF verification has rapidlyevolved to address the complex RF interfaces found in today's telecomstandards.
Intensive mathematical analysis is required to correctly understand andverify these interfaces. Most sectors of the telecoms industry havedefined validation processes and procedures that laboratories mustadhere to.
In the mobile-phone segment, groups such as the
Note that a validation organization must be
Validation of a new technology begins as soon as an appropriateindustry body decides that the technology is stable, and that there areenough manufacturers and operators ready to use it.
The validation organization is required to write a specificationbefore starting any verification, after which it produces a validationreport.
Similarly, in case of problems during test implementation, themanufacturer must be alerted immediately. Generally, a laboratoryverification procedure entails the following steps:
1. Confirm thatthe manufacturer uses the correct version of the industryspecifications.
2. Check that theprotocol message sequences comply with relevant test specifications.
3. Ensure thatthe correct pass, fail or inconclusive verdicts are given.
4. Verify thatthe implementation is stable and will run reliably in differentscenarios.
5. Assess thecalibration methodology.
6. Evaluate themanufacturers' test-system uncertainties.
7. Assess theaccuracy of the test system's RF path-compensation procedures.
8. If a newhardware is added to the platform, perform rigorous testing to ensurethat its RF characteristics suit its intended purpose.
Meanwhile, a validation report must detail all software, firmwareand hardware used, along with the version of the industry specificationand any amendments to it.
Any deviations done and the rationale for such must also be noted.However, a detailed description of the device-under-test (DUT) mayprove difficult to make due to commercial confidentiality agreements.
Non-compliance with a test case is something that manufacturerswould not want to admit to their competitors. The test platformmanufacturer identifies a target test case. Several DUTs from differentmanufacturers that can run appropriate test cases are then acquired.
Often, this means approaching the manufacturers directly becauseoff-the-shelf devices do not support special test modes required fortesting. Nonetheless, non-compliance is not an issue as it helpsdetermine how a test platform behaves with failing user equipment. Infact, large failure data can crash a test platform if the manufacturerhas not allotted sufficient memory for it.
Several technologies require the manufacturer to complete animplementation conformance statement to determine the suitability of atest sample. However, the surest and fastest way to do so is to run thetest case on a test sample. A protocol trace log records the full test.
Cellular technologies use DUT and test system logs, while WLANtechnologies often use over-the-air “sniffers.” Basic RF parameters areeasily verified using standard test measurement instrumentation.Meanwhile, some test sample modes and settings are only verified usingdebug facilities within the DUTs or with trace logging tools.
To reduce test times, some technologies use statistical analysis.Keeping the test time at a minimum without compromising the accuracy ofthe overall pass-fail verdict is a paramount consideration.
A simple but time consuming method of assessing the accuracy of thestatistical test method requires comparing the measurements done withand without statistical analysis. Generally, the industry committeeswrite and develop the standards.
The task of validation laboratories is to ensure compliance withthese specifications. However, labs can also point out errors toindustry representatives.
Reviewing the manufacturer's system uncertainty calculations isvital. Often, the applicable industry will have issued guidelines onhow to calculate uncertainties.
For mobile radio equipment, an industry body such as the
System uncertainties are a combination of systematic and randomcomponents, the suitability of detectors to measure modulated signals,the operation of test instrumentation in linear measuring range andmismatch of RF connections.
A huge amount of data can be gathered from manufacturers' datasheets and from calibration certificates, while repeatability ofmeasurements can be quickly checked by practical measurements.
Independent validation laboratories help manufacturers keep abreastof evolving applicable standards. They develop and maintain relevantskills in both consultancy and verification testing, while advisingmanufacturers on test specifications and generating accurate testreports.
More importantly, they help expedite the launch of RF products inthe market, matching new technologies with consumer needs andprevailing standards.
Stuart Thomas is Manager ofCellular Technology Development at RFIGlobal Services Ltd.