Introduction to WiMAX Transmitter Measurements

Overview

As the deployment of WiMAX devices continues to rise, test engineers face an increasing challenge to reduce test costs. Thus, automated test systems used in either design validation testing or production testing must be designed to perform fast, accurate, and repeatable WiMAX measurements. Read this white paper to understand the types of measurements typically required for characterization of Fixed WiMAX (IEEE 802.16-2004) and Mobile WiMAX (IEEE 802.16e-2005) devices.

Overview of the WiMAX Frame Structure

Each of the two WiMAX classifications, Fixed and Mobile, are based on a subset of the IEEE 802.16 standards and are defined by the WiMAX forum. More specifically, Fixed WiMAX is based on the orthogonal frequency division multiplexing(OFDM) physical layer of the 802.16-2004 specifications, which are sometimes called IEEE 802.16d. Mobile WiMAX is based on the orthogonal frequency division multiplexing access (OFDMA) physical layer of the 802.16e-2005 standard, which is a revision of the original Fixed WiMAX standard. Mobile WiMAX provides added functionality such as base station handoffs, multiple input multiple output (MIMO) transmit/receive diversity, and scalable fast Fourier transform (FFT) sizes [1]. Table 1 shows a high-level side-by-side comparison of the Fixed and Mobile WiMAX standards.

	Fixed WiMAX	Mobile WiMAX
Standard	IEEE 802.16-2004 (also called « d »)	IEEE 802.16e-2005
Multiplexing	OFDM	OFDMA
FFT size	256	Scalable (512, 1024, and so on)
Duplexing mode	TDD, FDD	TDD
Modulation scheme	BPSK, QPSK, 16-QAM, and 64-QAM	QPSK, 16-QAM, and 64-QAM
Subcarrier spacing	15.625, 31.25, 45 kHz	10.94 kHz
Signal bandwidths	3.5, 7, and 10 MHz	5, 7, 8.75, and 10 MHz
Spectrum	3.5 and 5.8 GHz	2.3, 2.5, and 3.5 GHz

Table 1. Fixed Versus Mobile WiMAX [1]-[3]

Fixed WiMAX Frame Structure

Fixed WiMAX supports both time division duplex (TDD) and frequency division duplex (FDD) modes. In addition, you can configure Fixed WiMAX frames to support multiple bursts that use various modulation schemes. In the following sections, learn about the structure of TDD and FDD frames, subframes, and duplex modes.

In FDD mode, transmit-and-receive signals from both the base and subscriber stations are on different channels. As shown in Figure 1, the base station transmits on the downlink channel frequency and the subscriber stations transmit on the uplink channel frequency. The WiMAX standard is also explicitly designed to concurrently support half-duplex and full-duplex subscriber stations, as shown in Figure 1.

Figure 1. Uplink and Downlink of FDD Mode [2]

From Figure 1, note that the base station uses a time-division multiple access (TDMA) transmit scheme. Under this configuration, the FDD downlink frame is used by multiple subscriber stations. Also observe that subscriber station 0 is supported with subframe 0 in both the uplink and downlink channels. This subscriber station (SS) in full-duplex mode is able to support simultaneous transmit and receive. By contrast, half-duplex subscriber stations are not able to transmit and receive simultaneously. Thus, the uplink and downlink subframes from subscriber stations 1 and 2 do not occur at the same time.

Time Division Duplex Mode

In TDD mode, the same channel is used for both uplink and downlink traffic. As a result, a TDD frame consists of an uplink and downlink subframe, each of which is separated by a timing transition gap (TTG). In addition, each subframe contains a preamble and one or more bursts. Note that each burst within a subframe must be arranged in decreasing order of robustness [3]. In Figure 2, observe that the QPSK (most robust) burst precedes the 16-QAM burst that precedes the 64-QAM burst (least robust).

Figure 2. TDD Downlink Subframe Structure

Mobile WiMAX Frame Structure

While the Mobile WiMAX profile allows for TDD and FDD operation, most deployed systems use TDD mode. Unlike Fixed WiMAX, which uses OFDM frames, Mobile WiMAX uses OFDMA frames. With OFDMA, multiple users can receive data from the base station at the same time. In this type of system, downlink bursts from the base station are divided by both time and frequency (subchannel) offset. Thus, each burst has unique time slot and subchannel allocations. An example of a TDD OFDMA burst used in Mobile WiMAX is shown in Figure 3.

Figure 3. OFDMA Downlink Subframe

In Figure 4, note that each burst in a Mobile WiMAX TDD subframe has a unique symbol and subcarrier offset. This burst allocation method allows for multiple subscriber stations to receive information from the base station simultaneously. One common visualization technique used to create Mobile WiMAX bursts is with the Zone map. Figure 4 illustrates an example map from the NI Mobile WiMAX Generation Soft Front Panel.

Figure 4. Soft Front Panel Zone Map

In Figure 4, each burst can have a unique modulation and coding scheme. In addition, Mobile WiMAX does not have the same “order of robustness” restrictions that apply to Fixed WiMAX signals.

WiMAX Transmitter Measurements

This white paper focuses exclusively on transmitter measurements and methods to reduce measurement time. In the following sections, explore measurements such as transmit power, error vector magnitude (EVM), subcarrier flatness, and spectral mask margin.

Transmit Power

The transmit power measurement is one of the simplest measurements for validating the RF front end. While you can measure transmit power with an RF power meter as well, using a vector signal analyzer to measure power provides benefits such as faster measurement times and more granular power measurements on the preamble and individual bursts. One common vector signal analyzer trace is “power versus time,” which is shown in Figure 5.

Figure 5. Power versus Time for a Mobile WiMAX Burst

Note that in both Fixed and Mobile WiMAX, the preamble is boosted by either +3 dB (Fixed WiMAX) or +9 dB (Mobile WiMAX) higher power than the data symbols.

Also observe that the OFDM and OFDMA signal structures used in Fixed and Mobile WiMAX fundamentally produce an inherently large peak-to-average-power ratio (PAPR). In fact, a Mobile WiMAX uplink signal can have a PAPR of up to 12 dB. When measuring RF power, it is important that the peak power of the burst does not exceed the peak overload power of the vector signal analyzer. Note that this is not the same as analyzer reference level. For example, the NI PXIe-5663 vector signal analyzer allows for a 10 dB peak-to-average headroom. When the analyzer is configured with a 0 dBm reference level, the instrument can accept signals up to +10 dBm before overloading  the instrument. Thus when measuring a burst that has an average power of +10 dBm, the vector signal analyzer reference level must be set to +12 dBm or higher.

Error Vector Magnitude (EVM) and Relative Constellation Error (RCE)

EVM and RCE are some of the most important metrics of a Fixed WiMAX transmitter’s performance because these measurements capture error due to a variety of impairments including quadrature skew, I/Q gain imbalance, phase noise, clock recovery, and nonlinear distortion. Note that the EVM and RCE are nearly interchangeable terms.In general, RCE describes an EVM measurement that is calculated over an entire Fixed WiMAX frame.

For a modulated signal, the EVM measurement compares the measured phase and amplitude of a signal with the expected phase and amplitude. The NI Fixed and Mobile WiMAX analysis toolkits calculate it by dividing the error vector|E| by the magnitude vector |V|, shown in Figure 6.

Figure 6. Graphical Representation of an EVM Measurement

The IEEE 802.16-2004 standard (Section 8.3.10.3) prescribes that a Fixed WiMAX transmitter must have a minimum RCE per each modulation scheme, as shown in Table 2.

Burst Type	SS RCE (dB)	BS (RCE)
BPSK-1/2	-13	-13
QPSK-1/2	-16	-16
QPSK-3/4	-18.5	-18.5
16-QAM-1/2	-21.5	-21.5
16-QAM-3/4	-25.0	-25.0
64-QAM-2/3	-29.0	-29.0
64-QAM-3/4	-30.0	-31.0

Table 2. 802.16d Minimum RCE for Various Modulation Types [3][4]

The IEEE 802.16e-2005 standard (8.4.12.3) prescribes that a Mobile WiMAX transmitter must have a minimum RCE per each modulation scheme, as shown in Table 3.

Burst Type	SS RCE (dB)	BS (RCE)
QPSK-1/2	-15.0	-15.0
QPSK-3/4	-18.0	-18.0
16-QAM-1/2	-20.5	-20.5
16-QAM-3/4	-24.0	-24.0
64-QAM-1/2	-26.0	-26.0
64-QAM-2/3	-28.0	-28.0
64-QAM-3/4	-30.0	-30.0

Table 3. 802.16e-2005 Minimum RCE for Various Modulation Types [4]

When characterizing a transmitter’s performance, your choice of modulation scheme at a given power level often has little effect on EVM performance. Thus, in a validation or production test environment, it is most common to simply measure the EVM or RCE for only a 64-QAM burst.

Visually, you can inspect EVM in several dimensions. A constellation plot provides the measured phase and amplitude of each recovered symbol. You can use this plot to identify signal impairments. In many cases, you can use this plot to identify which factor (AWGN, nonlinearity) contributes the most error to an EVM measurement. An example constellation plot showing QPSK, 16-QAM, and 64-QAM signals is shown in Figure 7.

Figure 7. Visualization Using the Constellation Plot

Also, the EVM versus subcarrier trace helps you identify if spurs or other in-band distortion is affecting the modulation quality of a given subcarrier.

Figure 8. EVM versus Subcarrier for a Mobile WiMAX Signal with a 1024 FFT Size

A final EVM visualization trace is the EVM versus symbol. With this trace, you can detect whether the modulation quality is consistent throughout the burst.

Figure 9. EVM versus Symbol for a Mobile WiMAX Burst

Subcarrier Flatness

Both the IEEE 802.16-2004 and the IEEE 802.16e-2005 specifications place requirements on the maximum channel-to-channel power offset between OFDM subcarriers. As shown in Table 1, the number of subcarriers varies between Fixed and Mobile WiMAX. While Fixed WiMAX signals have 256 subcarriers, Mobile WiMAX bursts can have 128, 512, 1024, or 2048 subcarriers. In both flavors of WiMAX, the maximum allowable subcarrier-to-subcarrier power difference is ±0.1 dB. In addition, the 802.16-2004 specifications for Fixed WiMAX mandate maximum and minimum power levels relative to the average power level as well. This is shown in Table 4.

Spectral Lines	Flatness Specification
–50 to -1 and +1 to +50	±2 dB of average power level
–100 to -1 and +1 to +100	+2 dB/-4 dB of average power level

Table 4. Tolerance for Fixed WiMAX Subcarrier Flatness

You can observe the spectral flatness of the NI PXIe-5673 vector signal generator and NI PXIe-5663 vector signal analyzer in Figure 10.

Figure 10. Vector Signal Analyzer and Vector Signal Generator with Maximum Flatness Less Than ±0.4 dB over the Entire Spectrum for a Mobile WiMAX Signal

Spectral Mask Margin

Spectrum mask, the final key transmitter measurement, provides a method to characterize transmitter nonlinearity and check for spurious signals. In general, you use the spectrum mask trace as a diagnostic tool. However, the spectrum mask margin measurement is a pass/fail test that is defined by the IEEE 802.16-2004 specifications in Section 8.5.2. According to these specifications, a transmitted signal in licensed bands must meet the mask specified in Table 5.

Bandwidth	A	B	C	D
20 MHz	9.5 MHz	10.9 MHz	19.5 MHz	29.5 MHz
10 MHz	4.75 MHz	5.45 MHz	9.75 MHz	14.75 MHz

Table 5. 802.16-2004 Spectral Mask Limits

Observe from Table 5 that the 802.16-2004 specifications provide unique spectral mask limits for the 10 and 20 MHz signal bandwidths. Note that for unlicensed band, the mask limits are determined by the local regulatory authority and are not prescribed by the IEEE 802.16-2004 specifications. Also, while the IEEE 802.16e-2005 specifications do not offer additional guidance on spectral mask measurements, the NI Signal Analysis Toolkit for Mobile WiMAX provides an adjacent channel power measurement.

With the NI WiMAX analysis toolkits, you can perform spectral mask either “instantaneously” on a nongated acquisition or on a burst using the gated spectrum measurement. Note that a gated spectrum measurement requires the transmit signal to be bursted and cannot be performed on a continuously modulated signal. Figure 11 illustrates a spectrum mask measurement on a Fixed WiMAX signal at 2.5 GHz.

Figure 11. Spectrum Mask of a Mobile WiMAX Signal

Conclusion

Testing today’s WiMAX devices requires accurate RF instrumentation and a variety of measurement techniques. For transmitter testing, measurements such as power, EVM, subcarrier flatness, and spectral mask margin are some of the common metrics used to characterize device or component performance. To learn more about the products used to perform the measurements in this white paper, view Preconfigured NI WiMAX Test System.

References

[1] Andrews, Ghosh, and Muhamed, Fundamentals of WiMAX: Understanding Broadband Wireless Networking. Prentice Hall, 2007.

[2] Mobile WiMAX – Part 1: A Technical Overview and Performance Evaluation, by WiMAX Forum, August, 2006.

[3] IEEE Standard for Local and Metropolitan Area Networks: IEEE Std. 802.16-2004.

[4] IEEE Standard for Local and Metropolitan Area Networks: IEEE Std. 802.16e-2005.