Testing the performance of digital technologies

Different digital standards result in different test methods and specifications. Field test equipment that generates and receives digital signals can improve maintenance of emerging radio systems.

The new digital systems are changing the way radio performance is specified and measured. Ensuring the best system operation and user satisfaction requires that new tests and measurements be made to properly maintain the equipment. Here are some of the similarities and differences in performance measurement between existing FM radios and some of the new digital systems.

Digital radio standards differences Anyone who has serviced two-way radios is familiar with the tests and measurements used to confirm proper operation of analog FM radios. The most basic measurements commonly made in the field are likely to be power output and receiver sensitivity. Simple talk-and-listen tests verify that the radio is ready to be put back into service.

The new digital standards have changed things quite a bit. Different digital standards result in different test methods and specifications, which further complicates the issue. The method of specifying and measuring the two basic parameters of power output and receiver sensitivity varies from system to system.

A digital radio or system is one in which the desired intelligence (either voice or data) is transported by digital, rather than analog, means. Systems using digital signaling_such as trunking control channels or those transmitting low-speed digital data_have been around for years. Voice systems using a 12kbps data rate have been in operation on the land mobile bands since at least 1976. In this case, some audio quality is sacrificed for security and simplicity.

Two trends have resulted in the development of new radio technology:

* The increasing ability to pack higher and higher data rates into existing radio channels using improved modulation techniques. * New voice compression technology allowing increasingly lower data rates without sacrificing good audio quality.

Low-power-consumption microprocessors make it possible to use both of these computation-intensive trends.

Understanding the basic layers of a digital system helps in understanding the issues raised in measuring performance. The most basic layer is the over-the-air connection. This layer establishes a stream of data between two radios: The type of RF modulation is specified, and the means for establishing synchronization and data framing is determined. (See Figure 1 below.) At this point, the traditional radio begins to evolve into a digital device. Once the radio-to-radio data connection is established, data are then passed on to be processed for signaling information, user verification, error correction and other radio control functions. Channels of data, including voice data, are prepared to be passed on to their appropriate applications. In the case of voice operation, data pass on to the vocoder.

For voice transmissions, the vocoder converts analog voice information into a stream of digital data that can then be transmitted on the established data channels. On the receiving end, the reverse process is carried out. Many types of vocoders are available today, each with good and bad attributes. For two-way land mobile communications, some of the desirable characteristics are:

* Good audio quality under noisy conditions. * Good data compression. * Immunity to degradation of voice quality caused by received bit errors.

On a voice channel, the resulting intelligibility is of key importance.

As in analog systems, power output and receiver sensitivity are important measurements of digital radio performance. How power output is measured depends on the type of digital system being measured. For an FM system such as an APCO Project 25-compliant system using C4FM modulation, the measurement is identical to that for an analog FM system. In the case of TDMA systems such as iDEN*, where the transmitter is pulsed on and off, a more complicated method that takes into account the duty cycle must be used.

Measuring receiver sensitivity is more difficult. A common method of measuring sensitivity in a digital data system is the bit-error rate (BER). This measurement compares a known transmitted pattern with the received data stream and expresses the result as a percentage of the bits in error.

The problem with using BER to indicate receiver sensitivity is that its effect on radio voice performance depends, to a great extent, on where in the radio layers the measurement was taken. For instance, a single bit error that results in the loss of synchronization could result in a series of incorrect bits being sent to the vocoder. How well the vocoder reacts to this series of incorrect bits determines the effect of this single bit error on intelligibility. On the other hand, various levels of error correction may be used before data are sent to the vocoder. This would correct for a single over-the-air bit error and, therefore, would have absolutely no effect on voice quality. Measuring only the over-the-air BER before framing, error correction and vocoding may not always be the best indication of actual voice performance.

Standards have been developed for C4FM digital radios compliant with APCO Project 25. In this case the desired test signal is a four-level, digitally modulated signal that produces a 1,011Hz tone when fed into the proper vocoder. Because this bit pattern is known, observing it at the receiver and comparing it to the expected result yields a BER measurement. The reference level has been set at a 5% BER. Sensitivity of these digital radios is expressed as the test signal level that results in a 5% BER.

A test signal for a silent tone has also been established for these same radios. This signal allows hum and noise measurements to be made in a manner similar to analog radios. The silent tone delivers a series of bits, which when delivered to the specified vocoder, produces silence. The test tone signal is used by the radio to produce an audio reference level. The signal is then replaced with the silence tone, enabling hum and noise measurements to be made.

Interference testing of APCO Project 25-compliant radios follows the same pattern. In this case, the standard test tone signal is used to set the reference sensitivity level. The desired signal level is then increased by 3dB. An interfering signal on the adjacent channel is then added. This signal is a C4FM-modulated signal that has a bit pattern consisting of a specified, repeating pseudo-random sequence. The level of this signal is then increased until the reference BER increases back to 5%. This process is similar in nature to the familiar analog FM measurement.

Also, as with analog FM, several additional laboratory tests ensure acceptable radio performance within the digital system. Tests such as attack time, adjacent channel splatter and timing accuracy tend to be complex. These tests are generally more suited for the laboratory or, possibly, a production setting.

New test equipment requirements The new digital radio systems pose new requirements both for laboratory and field testing. Test equipment that supports maintenance of radios used in the new emerging systems must contain modulators and demodulators capable of generating and receiving the specialized physical signals each system requires. Each system has unique, exclusive framing and error correction methods. In addition, each system has its own type of vocoder and they are generally incompatible. To make testing matters worse, these digital systems lend themselves well to encryption. Because the voice message is already converted to digital form, it is a natural extension to add digital encryption; however, it produces yet another compatibility challenge for test equipment and radios.

To keep field equipment portable and reasonably priced, a distinction must be made between field and laboratory measurements. The interference measurements described for both analog and digital systems are typically laboratory measurements because of their complexity and the requirement for specialized equipment. The more basic types of power and sensitivity measurements are useful in the field. Because many of the digital radios also operate in the analog mode, the field test equipment must also be capable of performing the traditional analog FM measurements.

A communications analyzer used in the field to maintain digital radio equipment should be able to accurately measure power output. For APCO Project 25-compliant radios using C4FM modulation, a standard wattmeter used to measure power in analog FM systems should be adequate. For other digital systems, more complex power meters may be necessary. In TDMA systems, the wattmeter must be capable of handling the pulsed nature of the transmitter.

Measuring receiver sensitivity requires the generation of a specialized test signal. This signal must have physical characteristics matching the system being tested. Additionally, depending on the system and test method being used, the correct framing, error correction and voice or data bits must be included. The ability to generate other test signals, such as the silence pattern and the interfering signal pattern specified in the C4FM digital radio standard, may also be useful.

One traditional test, commonly used for maintaining most analog radios, is a talk-and-listen test. The ability to pick up a microphone and hear your voice come out of the radio under test goes a long way in gaining confidence that the radio is working properly. Because the digital radios use a vocoder, special capability must be added to allow this test. The test equipment could be made to store voice data transmitted to it by a working transmitter and then retransmit this data to the receiver under test. If the test is unsuccessful, some uncertainty is left as to whether the receiver or transmitter was at fault. A more complete way of performing the test would be to include a complete vocoder within the test equipment itself.

Looking ahead The new digital radios definitely create a new challenge to field testing and maintenance. Along with the usual analog measurements, new methods must be used to perform basic tests in the digital mode. In the case of APCO Project 25-compliant radios, the test methods are well-developed in the C4FM performance recommendation and measurement methods documents.

Many of these tests are familiar to those who have tested analog. Field test equipment capable of generating and receiving these specialized signals can help to make maintenance of these new radios more productive and efficient.

References EIA/TIA TSB102.CAAA Digital C4FM/CQPSK Transceiver Measurement Methods EIA/TIA TSB102.CAAB Digital C4FM/CQPSK Transceiver Performance Recommendations

Mach is engineering manager for two-way products at Motorola Communication Test Equipment, Scottsdale, AZ.