Nonsignaling technique improves RF test
A new approach eliminates the need for real-time emulation of base-station or network functions during wireless-device production testing to speed the alignment process.
By Henry Gropp, Rohde & Schwarz -- Test & Measurement World, 3/1/2008

Many wireless devices now combine GSM, WCDMA, Bluetooth, WLAN, GPS, and FM technologies while also operating at the high data rates required for mobile Internet. And because customers demand that mobile radio service be available on any continent, many devices also operate in multifrequency bands and support multimode operation.
While all these features are great for the customer, they create challenges for the test engineer. Each additional technology and frequency band that is added increases the test effort, leading to longer alignment times in production—clearly an impediment for meeting demands for low-priced products. Cost containment requires that manufacturers adopt completely new test approaches. Fortunately, non-signaling test concepts and predefined test sequences can reduce test time by up to a factor of 10.
Two-step approach aligns RF devicesRF chips and components that are manufactured under cost-saving measures exhibit variations in their frequency and level characteristics. To ensure that the devices will function properly in a real network, manufacturers use complex alignment procedures during production to eliminate the variations. Most manufacturers prefer a two-step approach for this purpose (Figure 1):
- The first step involves measuring how much the actual transmit and receive parameters of a wireless device deviate from their ideal values. The correction values are then determined and stored in the device. In most cases, this requires the manufacturer to calibrate the transmit power stages and the receive signal strength indication (RSSI) for various mobile radio bands and technologies.
- The second step involves verifying that the fully calibrated device works as intended. The manufacturer measures transmit parameters such as modulation quality, spectrum, and power and compares them with the parameters for the technology for which the device is designed (GSM, Bluetooth, WLAN, etc.). In most cases, the absolute sensitivity of the receiver is determined by means of a bit-error-ratio (BER) test.
![]() Figure 1. Typical steps in the production of wireless devices involve, first, measuring the deviations of the actual transmit and receive parameters from their ideal values, and second, checking the fully calibrated device. |
Because each additional technology and each additionally supported frequency band prolongs the two-step test alignment process, some manufacturers are turning to approaches that involve nonsignaling techniques for aligning wireless devices. In such an approach, the first step—the calibration step—is carried out in a nonsignaling mode in which the device under test (DUT) is operated in a special test mode. The measuring equipment includes RF analyzer and generator functions but does not perform real-time emulation of base-station or network functions. The wireless device is not yet aligned and therefore does not yet perform as it will in later network operation.
In the second test step, verification is carried out by means of the signaling mode. In this mode, the tester simulates various functions of the base station and of the network in real time. Signaling procedures are used in the wireless device to switch through all the technologies, bands, levels, and so on, to be tested.
It is the second step—the verification step—that offers the potential for sizeable savings, since the signaling, as compared to the RF measurements alone, takes up to four times longer. The main reason for this difference is that signaling procedures were actually developed for real network operation, not for extremely fast production test. In contrast, the nonsignaling mode is the test mode that is especially speed-optimized for production. As a result, the DUT can quickly activate the required test signals (level, frequencies, technologies).
![]() Figure 2. The signaling sequences for driving a GSM/GPRS/WCDMA mobile phone typically require 160 s; RF tests add another 40 s. |
The signaling sequences for driving a GSM/GPRS/WCDMA mobile phone typically require 160 s (Figure 2). When you also add in the RF tests, the time is typically increased by a further 40 s. In the classic signaling method, the alignment can thus typically be expected to take 200 s. If you use the nonsignaling approach, the DUT could be aligned in 40 s—a time savings as high as 80%. Communications testers like our R&S CMW500 employ a nonsignaling technique for both steps. Predefined test sequences reduce data communications
At Rohde & Schwarz, our analyses of test times in the nonsignaling mode have shown that a further increase in the speed of the individual RF measurements does not necessarily lead to a drastic reduction in alignment time. Many radio communication testers already attain a high measurement speed at which they measure GSM signals in real time, for example, so they can carry out a modulation analysis on one time slot for each GSM frame without any problem. Thus, the largest potential for improvement in data transmission is found within the production test system itself.
At present, the most commonly applied test concept is to carry out individual measurements one at a time. The system controller requests each measurement separately—that is, both the DUT and the tester are reinitialized at each test step (Figure 3). At the end of each single measurement, the result obtained is returned to the system controller from the tester.
![]() Figure 3. (a) Test has typically been carried out one measurement at a time, with reinitialization of the DUT and tester required after each step. (b) Predefined test sequences allow test results to be accumulated and transferred to a system controller all at once. |
To shorten the cycle, mobile communications testers with a nonsignaling mode make use of predefined test sequences instead of single measurements. First, the complete sequence of all tests to be carried out is transferred from the system controller to the DUT and tester in one operation. After receiving a start trigger, the DUT and the tester process the sequence simultaneously. The DUT activates the signals to be measured, and the tester starts the appropriate measurements in sync. The results are stored in transient memory. At the end of the sequence, the list of all results is transferred to the system controller.
Therefore, the software layer structure must only be traversed a total of two times: once from top to bottom for initialization and then back again in order to transmit the list of results. This approach allows a further 50% reduction in test time in the nonsignaling mode.
In the past, chipset and wireless device manufacturers focused on customizing their solutions to meet the end customer’s requirements. New features such as higher data rates, better ease of operation, and smaller dimensions should offer a more attractive solution.
To implement the test modes that must be used for nonsignaling production concepts, chip designers must include special test modes on their devices. For established technologies such as GSM and WCDMA, fulfilling this task should not be any problem. The extra development effort will quickly pay for itself.
New technologies such as LTE, however, represent a notable challenge. To ensure they present a stable and functioning solution to the customer on schedule at rollout, manufacturers use a production test setup that matches how the device will operate in a real network as precisely as possible. As a result, signaling concepts are employed for new product designs. Measuring equipment that employs nonsignaling techniques must, therefore, be able to accommodate signaling tests as necessary.
It is possible for the test time in wireless device production to be reduced by a factor of up to 10. For this to happen, chipset manufacturers must integrate the required test modes, and the production line must incorporate a tester with a nonsignaling mode and predefined test sequences.
| Author Information |
| Henry Gropp has been product manager for mobile radio testers at Rohde & Schwarz in Munich since 1997. His main tasks are product definition and marketing of mobile radio test assemblies. After earning a degree in avionics with emphasis on communications/navigation from Riga Aviation University in 1989, he worked in test equipment development at the Dresden Aircraft Factory. From 1993 to 1997, he served mobile radio customers as a sales engineer at Rohde & Schwarz. |





















