In-Field Testing

The number of digital handsets is increasing, and the digital market is maturing. Although this is great for business, it creates challenges for comprehensive in-field handset testing.

In today's market, digital phones are being held to higher expectations. There are increasing standards to determine what level of failures and/or performance degradation should render a phone out of service. As carriers continue marketing digital as higher quality than analog, ensuring proper handset operation becomes increasingly important.

Often performance degrades slowly in the field, so users do not detect degradation in the handset's performance. Through documentation and testing, you can find and report such trends to the manufacturer. In addition, by determining problems and documenting them before you send phones to the manufacturer's service center, you can keep track of phone quality and ensure manufacturers are correcting problems and improving quality continually.

Digital handsets also require new test instruments, which often translates into a substantial capital investment. In addition, manufacturers are increasing the cost of handset returns -- regardless of fault or no fault -- driving carriers to implement their own comprehensive testing plans in the field.

A point-of-sale test program has many advantages. You can save money by cutting out no-fault-finds early in the process, provide customers with a quick response, and put forth a high-quality, technically competent and personalized image. In an industry with a high churn rate, image is vital.

But testing is a confusing and complex process. This is true even for carriers that just want to determine if a handset is working or not before they send the phone to a high-level service center for repair.

The average technician, or sales- person, who will perform most of the testing at the service desk or store counter has limited knowledge of technical theory and testing. It is essential that he be able to test phones quickly, accurately and easily. You could try to hire high-level technicians to perform in-field testing, but more than likely you won't find them because of the high demand throughout the industry. With that in mind, you will want to find testing equipment that can make the process efficient, meet your needs, and ensure that you can offer customers the quality products and services they expect.

FIELD TESTS When testing any communications system at a low level, the basic test plan should determine whether the problem is related to power, audio, modulation/timing/frequency or software (call processing). This allows the technician to zoom in on a particular block of the handset. When testing digital handsets, there are several groups of tests that can isolate a problem into one of these areas:

*Power and frequency tests

*Timing and modulation quality (EVM in TDMA, phase error in GSM and Rho in CDMA)

*Receiver tests (dynamic range, sensitivity)

*Handoff functionality

*Call processing (mobile origination/termination, base originated call/termination, registration tests, authentication and short messaging functionality)

*Voice loopback (to test the microphone and speaker)

*Analog tests (dual-mode operation)

POWER-CONTROL TESTING Power is important in any communications system. With TDMA and various GSM systems, handsets burst power on and off during assigned time slots. Therefore, technicians must ensure the correct power level is being transmitted, that the burst ramps up and down quickly, and that the top of the burst is flat. It also is important to make sure the handset responds to power-level change requests.

In CDMA systems, power control is the most important test, but also one of the more difficult. The two forms of reverse-channel power control in a CDMA system, closed and open loop, must be thoroughly tested, even in the field. First, the open-loop system should be tested. The test station performs this by sending alternating 1s and 0s as the power-control bits and varying the base station power. The base-station power is set at -25, -65 and -104dBm, with the mobile changing its output power according to table 10.4.1.3-1 in IS-98.

Following any change in base-station mean output, the mean output of the mobile must respond in a certain amount of time. Test this by varying the base-station output at different levels over 100ms and measuring the output of the mobile according to a specific mask.

Technicians also will need to perform access-probe power tests. They need to verify parameters such as nominal power offset, initial power offset, power increment between consecutive probes, number of access probes in one sequence and the number of probe sequences in one access attempt. Technicians can test these by setting the MAX_RSP_SEQ in the Access Parameter Message to one and setting the test set to ignore all of the access attempts. Then, the test set can measure the output power for each probe and the number of probes, the number of sequences. Then, the Access Parameter Message is changed, and the probes are measured again. This time, measure the first access probe of each sequence and make sure it is 6+/-1. 2dB above the power in the first access scenario. Measure the increment between consecutive access probes in each sequence and ensure it is 1+/-0. 5dB. The number of access probes in each sequence should be five, with three sequences. The probes should be randomized.

Technicians also should test the range of the closed-loop power control. The test set will send 100 consecutive 0s and 1s as the power control bit, and measure the range and functionality of the closed loop process. This is done at all of the data rates. The range of the closed-loop power control must be +/-24dB around the open-loop estimate.

Two other tests that combine the operation of the closed- and open-loop system are the maximum and minimum RF output power tests. For the minimum test, a -104dBm base-station signal is sent, thus maximizing the open-loop power control and all 0s on the power control bit to ensure the closed loop is at its ceiling. The specification will vary depending on the mobile station class from 1W to 6.3W. The minimum output power test is essentially the same. The maximum base-station signal is sent (-25dBm) and all 1s on the power-control bit. Then, the mobile should be transmitting less than -50dBm.

Standby output power is the output power of the mobile when transmit functions are not active. A measurement is made between access probes and must be below -61dBm. Gated output power is a measurement of the time response of a mobile during variable data rate transmissions, when the mobile is only transmitting during gated-on periods. This time response must fit in a given envelope mask.

TIMING & MODULATION QUALITY Although today's phase-locking technology reduces frequency errors, you should measure frequency on all handsets. Timing is important with digital systems. In TDMA systems, the handset must have excellent timing in order to use the correct time slot, and CDMA handsets must have the correct timing for its code sequences to be aligned properly.

GSM systems use GMSK modulation, so technicians need to be concerned primarily with any phase errors in the modulated signal. Large phase errors can result in poor voice quality as well as dropped calls. Generally, this results from a problem in the modulator section of the handset.

IS-136 uses pi/4DQPSK modulation. This means that both amplitude and phase errors can contribute to modulation problems. Error vector magnitude measures the combined phase and amplitude errors.

Many handset test instruments use a constellation diagram to graphically represent EVM. Phase is represented around the unit circle, and amplitude is represented outward from the origin. Usually, the symbols will appear in tight clusters at their appointed phase and amplitude positions, and a spreading of points indicates the problem. Problems related to amplitude will spread away or toward the origin, while phase errors spread around the unit circle.

This constellation analysis can be used as a form of fingerprinting handsets. Each manufacturer's handset will have a unique pattern. In this way, a technician can recognize a problem with a particular handset quickly.

Because IS-95/J-Std 8 handsets use O-QPSK modulation, EVM and constellation analysis can be used in the same way as in IS-136. However, CDMA uses another waveform quality measurement known as Rho. Essentially, Rho is a correlation of a mathematically ideal waveform with the waveform that is received. A correlation of 1.0 is perfect, meaning the waveform is identical to what it should look like. A primary cause of bad Rho is modulation quality, so if troubleshooting, a technician should look at a constellation diagram, which can graphically display the phase and magnitude errors that make up the error-vector-magnitude measurement.

RECEIVER TESTS Compared to traditional analog systems, digital communications systems have substantially different methods for measuring how well the receiver works. The basic idea is to transmit a known level modulated signal to the handset, allow the receiver to demodulate and then determine how many bit errors or frame errors were in reception.

Bit error rate (BER) is used in IS-136. In CDMA, frame error rate (FER) is used in determining this level. GSM uses BER and FER to determine how well the receiver works. Sensitivity is the lowest level that the handset can receive and demodulate a signal correctly. Dynamic range is the maximum and the minimum levels.

END RESULT Field testing is quite different from the full-fledged, complete testing done on the manufacturing line. Equipment costs, ease of use and setup, and speed requirements vary. You should invest in specialized equipment to meet all of these needs.

Effective field tests will ensure that you do not send phones back to the manufacturer unnecessarily. It also ensures the network is not degraded because of rogue phones. More importantly, it ensures customers receive the level of performance they expect.

Because CDMA handsets all use the same frequency channels, the CDMA network is a delicate balance of mutually interfering mobile phone users. The actions of one phone can interfere with the performance of another phone. Even if a CDMA handset seems to be working, it needs further testing to ensure overall network optimization.

In a CDMA network, the total level of interference (and thus, power control) is directly related to the quality of the network. The power-control factor is figured into the equation when determining capacity of the system. A malfunctioning power-control system could increase the overall level of interference in a sector, increasing the main limiting factor of a spread-spectrum system. All tests that involve power levels are critical in testing CDMA phones, including the closed- and open-loop power-control system and the access probing mechanisms. These functions should be tested thoroughly in the field.

Softer handoffs require the phone to perform high-level and split-second measurements and calculations. Too many handoffs can cause network problems; not enough handoffs decreases quality and increases dropped calls. Basic handoff functionality needs to be tested in the field to ensure they will operate correctly. Thorough testing of handoff ability is complex and expensive, requiring multiple base-station signals as well as extremely accurate synchronization. The complexity and cost make such extensive testing impractical in a field environment. Therefore, you should check basic functionality of the handoff system by using a softer (intersector) handoff.