New cellular wireless technologies commonly called third-generation (3G) aim to increase capacity and provide new services for customers. To meet these goals, the cdma2000 technology builds upon IS-95 technology and provides a relatively easy upgrade.
cdma2000 Differences and Test Implications
Complex reverse transmission coding enables simultaneous data and voice. The reverse channel-coding method surfaces as one of the most significant differences between IS-95 and cdma2000.
An IS-95 phone transmits one reverse channel of either voice traffic or access data. During a call, a cdma2000 phone handles at least two reverse channels, a pilot channel and a voice or data channel, using different codes. A cdma2000 phone can be configured to transceive one voice and two data streams simultaneously.
An IS-95 phone applies convolutional encoding to the data to be transmitted. A cdma2000 phone may use either convolutional encoding or turbo coding. Turbo coding is used for supplemental data channels only.
Depending on the data rate in use, turbo coding and decoding provide about 0.5 dB to 2.5 dB more efficient error correction than convolutional encoding and Viterbi decoding. As a result, about 0.5 dB to 2.5 dB less RF transmit power is required for an equivalent received frame error rate (FER).
An IS-95 phone uses offset quadrature phase shift keying (OQPSK) modulation to transmit the same data on both the I and Q paths. A cdma2000 phone transmits different data on the I and Q paths using hybrid phase shift keying (HPSK) modulation (Table 1) (see below).
Table 1. cdma2000 SR1 Reverse Link Channel Characteristics for RC3, RC4, and RC5
Channel
Abbreviation
Channel
Name
Obviously, it is important to test these reverse-link functions for characteristics such as relative power levels, modulation quality, noise levels, and channel-to-channel characteristics. TIA/EIA-98-D defines both the tests and the performance limits.1
• Relative Code Channel Power Levels and Inactive Channel Noise
TIA/EIA-98-D specifies certain combinations of code channel levels during testing. The phone must generate the correct reverse code channels at the correct power levels (within ±0.25 dB). A test instrument named a code domain power analyzer is specified to measure power in each transmitted Walsh code. See Figure 1 for an example of a code domain power measurement.
Uncorrelated power in inactive or active code channels is undesirable, wasted transmitter (TX) power. This noise also is measured with the code domain power analyzer, indicated in yellow in Figure 1.
• Turbo and Convolutional Coding Functionality
Both turbo and convolutional coding can be verified during Receiver (RX) FER testing, with data rates up to 307.2 kb/s.
• Waveform Quality
According to the current version of TIA/EIA-98-D, the waveform quality measurement is performed during a call handoff with only the R-Pilot transmitting preamble data. Rho is the figure of merit for waveform quality. Rho indicates how well the CDMA signal’s power distribution correlates with the ideal power distribution. See the sidebar Multi-Code Rho for a discussion of an alternative measurement.2
• Reverse TX Code Channels Must Be Orthogonal
For the base station RX to decode the multiple code channels transmitted by the phone, the phone’s code channels must be orthogonal; that is, the data streams must be time and phase aligned. Adherence to orthogonality is verified by measuring the reverse pilot channel to code channel time tolerance and phase tolerance in TIA/EIA-98-D.
Multiple radio configurations enable voice and data in many combinations. For SR1 reverse transmissions, there are four radio configurations (RCs) to verify. Depending on the RC chosen, R-FCH voice traffic transmits at 9.6 kb/s or 14.4 kb/s, and R-SCH data traffic operates from 9.6 kb/s up to 307.2 kb/s. Proper operation in all RCs must be verified on each new phone design.
IS-95 uses a simple open-loop power control formula with a constant of either -73 dBm or -76 dBm. In cdma2000, different RCs use different channel structures and data rates, so variable power levels must be transmitted to provide the same energy per bit at the base station RX. Consequently, a table and multivariable formula define open-loop power control operation for cdma2000. Proper open-loop power control of multiple configurations should be tested on a phone’s TX.
Following is an example of how open-loop power levels change vs. time while connecting a voice call between a phone and base station. If the total base-station cell power is -73 dBm/1.23 MHz, then the first access probe transmitted by the phone will be at approximately 0 dBm. The phone’s TX power will transition until it connects at approximately -1.9 dBm (assuming R-Pilot and full-rate R-FCH).
A quick paging channel allows a phone to stay inactive longer, extending standby and talk time. IS-95 phones monitor a specific time slot on a paging channel on a periodic basis. Demodulating the paging channel in the slotted mode requires power for the RX and signal-processing circuits for a relatively long period.
cdma2000 base stations transmit two page indicators per frame before the assigned time slot on the forward quick paging channel (F-QPCH). If no page indicators are present, the phone skips the associated time slot and quickly shuts off power for the RX and signal-processing circuits.
The phone will sleep until the next F-QPCH is due, conserving battery energy. Testing must confirm that a cdma2000 phone awakens and sleeps at the proper times when quick paging functions are enabled and disabled by the base station.
Receiver quality must be measured for both data and voice transmissions. For voice transmissions, cdma2000 receivers can be tested for sensitivity and dynamic range without interference and demodulation with additive white Gaussian noise (AWGN). These tests, performed during a call, are similar to those of IS-95.
For cdma2000 data transmissions, new tests are added using the Test Data Service Option (TDSO).3 According to TIA/EIA-98-D, the phone shall support TDSO if it implements a forward dedicated control channel (F-DCCH) or forward supplemental channel (F-SCH). Also, the phone shall support the Loopback Service Option if it implements a forward fundamental channel (F-FCH).
Table 2 (see below) illustrates a selected subset of these tests for RCs 1 to 5. The test instructions and data tables for cdma2000 RX testing cover many pages of TIA/EIA-98-D. There are hundreds of combinations of parameters that must be verified for a new phone design. However, only a small subset of these normally would be verified during phone manufacturing.
Table 2. Subset of Common cdma2000 SR1 Reverse Link RX FER Configurations
(TIA/EIA-98D
Section 1.3)
Radio
Configuration
Radio
Configuration
A manufacturer might choose to test the RX based on worst-case scenarios for that phone design or on common usage scenarios from service providers. For example, two service-provider scenarios might be a Loopback Service Option 2 call using RC1 at 1/8th rate, demodulating in the presence of AWGN, and a Test Data Service Option 32 call using RC3 at a 38.4 kb/s data rate, demodulating in the presence of AWGN.
The worst-case RX test scenarios often are determined for each phone during the development phase, usually called design verification test. In this phase, the test time primarily is dictated by how many 20-ms frames must be tested. Test time can be minimized if the test instrument allows rapid parameter changes between tests.
TIA/EIA-98-D also specifies receiver testing with real-world signal impairments such as fading and Doppler shifting.
Practical Considerations for Testing cdma2000 Phones in Manufacturing
Table 3 (see below) includes many tests commonly performed only during the R&D or design verification phases of phone development. Many different instruments are required to perform this extensive list of tests.
Table 3. Key cdma2000 Phone Test Differences
- Check F-FCH
- Check F-SCH using TDSO if supported
- Check convolutional and turbo coding functions
- Many more test combinations
- Rho for cdma2000 measures only R-Pilot during a hard handoff
- An alternate Multi-Code Rho measurement is available to check multiple active channels.2
Table 4 (see below) lists tests that might be performed in the practical environment of phone manufacturing. It also focuses on one type of instrument, often called a mobile station test set. Many cellular-telephone manufacturers around the world currently use a mobile station test set for phone calibration and test. This type of flexible test set acts like a miniature base station with internal instrumentation. Typically, the test set performs call processing and makes the key RX and TX parametric measurements needed in production.
Table 4. Summary of Recommended cdma2000 Test Set Capabilities for Manufacturing
- Calibrate TX power from -50 dBm to +23 dBm over each band class; this can be the most time-consuming calibration process for a phone
- Verify access probe power
- Verify that TX power tracks complex open-loop formulae
- Wide dynamic range (to -68 dBm desired) to allow external loss from cables or attenuators
- Measure rapidly for fast calibration
- Calculate open-loop formulae and display correct expected power
- Adjust and test TX relative code channel levels
- Test inactive code channel noise
- Test RX sensitivity and dynamic range
- Test RX demodulation with AWGN
- Accurate CDMA signal generator output level for RX measurements with low uncertainty
- Include convolutional and turbo coding capabilities
One example of this type of test set is the Agilent 8960 (E5515C) Wireless Communications Test Set with E1962B cdma2000 Test Application software. This test set is optimized for the calibration and final-test steps of phone manufacturing. It also can be useful in R&D and design verification applications.
A recent trend in mobile station test sets often is called multiformat capability. In the case of the Agilent 8960 Test Set platform, it also can be configured to test other cellular technologies, including GSM, GPRS, TIA-136 TDMA, and AMPS. All or a subset of these technologies can reside in the test set simultaneously.
References
- TIA/EIA-98-D, Recommended Minimum Performance Standards for cdma2000 Spread Spectrum Mobile Stations, Nov. 27, 2000, Ballot Version, Release A.
- Agilent Technologies Product Note [for E5515C with E1962B], Multi-Coded Waveform Quality and Code Domain Measurements for cdma2000, Agilent Literature Number 5988-1989EN, 2/01.
- TIA/EIA/PN-4877, Test Data Service Option (TDSO) to cdma2000 Standards for Spread Spectrum Systems, Ballot Version 11/13/2000.
Additional Information
- 3GPP2 documents: www.3gpp2.org/
- Telecommunications Industry Association documents: www.tiaonline.org/
- Agilent Technologies documents: www.tm.agilent.com/
About the Author
George Brandle is an application engineer at the RF Communications Product Generation Unit of Agilent Technologies. He received a B.S.E.E. in 1976 from Stanford University, then joined Hewlett-Packard as a product-marketing engineer. During his career, Mr. Brandle also has been a production engineer and held management positions in custom engineering, instrument and system production, metal fabrication, printed-circuit production, and learning products. Agilent Technologies, RF Communications Product Generation Unit, 24001 E. Mission Ave. MS 3WU-322, Liberty Lake, WA 99019, 509-921-3494, e-mail: [email protected].
Return to EE Home Page
Published by EE-Evaluation Engineering
All contents © 2001 Nelson Publishing Inc.
No reprint, distribution, or reuse in any medium is permitted
without the express written consent of the publisher.
June 2001