Making measurements involving MIMO applications requires adaptation of the algorithms used in the analysis software.

By Johan Nilsson and Dr. Wolfgang Wendler, Rohde & Schwarz

click to enlarge Figure 1. MIMO in WiMAX system.

Due to ongoing growth in demand for channel capacity, more and more radio standards such as WiMAX, 3GPP HSPA, 3GPP LTE and WLAN are adopting MIMO technology. Using MIMO it is possible to boost the data rate and achieve better transmission quality through simultaneous usage of multiple transmit and receive antennas. This technology takes advantage of the multi path propagation that occurs in a radio channel, which in earlier radio standards was regarded as interference. In MIMO, every additional path between the transmitter and the receiver improves the signal-to-noise (S/N) ratio. For mobile applications in particular, multipath reception considerably reduces the necessary receive level compared to a single transmission channel. Most communication standards define MIMO modes for two-, three- and four-antenna systems.

There is a distinction between space coding and spatial multiplexing (true MIMO). In space coding, the same information is transmitted on both antennas but with different coding. This increases the signal-to-noise (S/N) ratio and thus increases the capacity at the cell edges. However, the data rate increases only indirectly as a result of the improveed signal quality. This mode is also known as transmit diversity and the coding is based on the Alamouti technique.

click to enlarge Figure 2. The information is then multiplied with a matrix to perform the actual pre-coding.

With spatial multiplexing, different information is transmitted simultaneously on both antennas. This increases the throughput and improves the bandwidth efficiency. In more advanced systems the base station continuously adapts the coding of the signal to make optimum use of the multi path propagation.

To reduce the correlation of the propagation paths, the transmitter may introduce delays on all but one of the transmitted signals. This is called cyclic delay diversity and is used in combination with spatial multiplexing.

In all of the described modes, the data of the transmit antennas contain pilot or reference sequences. These sequences are transmitted on different frequencies or carriers as a function of the antenna in use so that there is no mutual interference. Based on the sequence, the receiver can thus clearly differentiate the data transmitted by the different antennas.

Due to space restrictions and power consumption in the mobile devices the so called collaborative MIMO is commonly used for transmission from the mobile device to the network. The principle is similar to Spatial Multiplexing but instead of two transmit antennas one device, two subscribers can send (collaborate) at the same frequency resource. By using collaborative MIMO, the system throughput is increased, not the throughput for each individual user.

OFDMA

click to enlarge Figure 3. The figure shows a typical measurement of a 2 antenna WiMAX MIMO signal. As result the spectrum flatness of the measured channel is displayed. Single antenna measurements can be carried out on antenna 0, where the preamble can be seen, and antenna 1 or in case of spatial multiplexing both signals can be combined and measured as well.

The OFDM transmission method is the basis for most wide band applications. In contrast to single-carrier methods, an OFDM signal consists of many orthogonal carriers, each of which is separately modulated. Since parallel data transmission is used, symbol duration is much longer than in single-carrier methods of the same transmission rate. In OFDMA several orthogonal physical carriers are combined, and each subscriber is assigned a specific number of carriers, depending on the bandwidth required. This makes OFDMA well suited for MIMO, as the necessary pre-coding can be adapted to the user's situation individually.

WiMAX MIMO

In the standard IEEE 802.16e-2005 MIMO is defined for 2 or 4 antenna systems, whereas first applications focus on 2 antennas only. WiMAX uses transmit diversity (matrix A) and spatial multiplexing (matrix B).

The burst structure is different on antenna 0 and antenna 1. The first zone is a DL PUSC zone with a preamble, which is always transmitted at antenna 0 only, no signal is present at antenna 1. In the following zone both antennas are active transmitting the MIMO signal. For MIMO precoding diagonal matrices are used and therefore the transmitted symbol is not distributed between the antennas, there is always one symbol at antenna 0 the next one at antenna 1 and so on (see Figure 1.)

For analysis of the TX signal it is therefore not really necessary to capture both antennas at the same time. The transmitted MIMO signals can be analyzed separately and only one signal analyzer is needed, which reduces the cost for T&M equipment. However, for measurements at antenna 1 a different synchronization algorithm has to be used, as no preamble is available.

UMTS Long Term Evolution MIMO

click to enlarge Figure 4a and 4b shows the constellation diagram at the antenna of a LTE signal where the off diagonal elements in the precoding matrix are ¹ ≠ 0. By combining the signal from the two analyzers, the signal can be correctly demodulated and the transmitted signals with 64 QAM (Orange colored), 16 QAM (blue) and QPSK (Green) can be identified.

To ensure the competitiveness of UMTS for the next 10 years and beyond, UMTS Long Term Evolution (LTE) is being specified in 3GPP release 8. LTE, which is also known as Evolved UTRA and Evolved UTRAN, is based on the OFDMA technology.

LTE uses the both transmit diversity and spatial multiplexing. The spatial multiplexing can be combined with delay diversity.

The user data (code words) is scrambled, and then modulated with the appropriate modulation format QPSK, 16QAM or 64 QAM. The information is then mapped onto layers. The number of layers is smaller or equal to the number of antennas in the system. The information is then multiplied with a matrix to perform the actual pre-coding (see Figure 2.)

Depending on the channel conditions the matrix is filled with different content. There are a large number of possible predefined matrixes. These are in the standard called code book entries. Table 1 gives an overview of the number of codebook entries.

click to enlarge Figure 4b.

When the off diagonal elements in the matrix are ≠ 0, the user information will be distributed among the transmit antennas. For a signal analyzer to perform signal analysis on such a signal, the RF signals from all the transmitters has to be captured at the same time to be able to reconstruct the content of the signal. WLAN IEEE 802.11n MIMO.

WLAN-n (IEEE 802.11n), the standardized "pre n draft" expansion of the IEEE 802.11a / g Wi-Fi mobile radio standards, is going to ensure a net data throughput of up to 100 Mbit/s in wireless LANs. Channel bandwidths of 20 MHz and 40 MHz are supported to enable high throughput. In the standard MIMO applications are defined with up to four spatial streams. Due to the fact, that spatial mappings of the different data streams to the different antennas is more similar to LTE, means matrices with off diagonal elements are used, in general for analysis more analyzers than one have to be used. For more details please refer to Ref 1.

Measurements on MIMO Transmitters

Making measurements involving MIMO applications requires adaptation of the algorithms used in the analysis software.

click to enlarge Figure 5. Supporting all MIMO modes.

Many test applications designed for verification work or for production are intended primarily to determine whether the transmitted signals comply with the relevant standard and whether the physical characteristics lie within specified limits. Here, the different transmitter paths do not need to be measured simultaneously. Instead, they can be tested one after the other. This means that one single analyzer is sufficient for this application. Results such as EVM, power and I/Q imbalance can be retrieved for each transmitter path.

click to enlarge Figure 6. Supporting up to 4 transmit antennas.

Much more extensive results are required, of course, during the development phase and for approval tests. For example, if you need to fully reproduce the data in the transmit signal or analyze the crosstalk between the antennas, you will have to perform simultaneous measurements on both antennas (Figure 3). In transmit diversity mode, this is still possible with one signal analyzer. The transmit antennas are combined and connected to the input of the analyzer (MISO: multiple input single output). Figure 3 shows this example for WiMAX MIMO measurement using one signal analyzer R&S FSQ or R&S FSG equipped with option R&S FSQ-K94 for MIMO measurements on WiMAX signals. Due to the special coding that is used, it is possible to separate the signals and fully demodulate the data. Both channels can be measured.

In spatial multiplexing mode, however, this measurement requires two or more analyzers in order to compute the channel matrix, and demodulate the signal. In the Rohde & Schwarz solution, one spectrum acts as the master and the additional spectrum analyzer(s) as the slave(s). They are triggered by the first and is used only to record the data that is collected at a central location. Figures 4a and 4b show a typical setup for measurement of a LTE MIMO signal.

Conclusion

Figure 7. Rohde & Schwarz's FSQ Signal Analyzer.

Rohde and Schwarz provides flexible and scalable solution for testing MIMO transmitters for WiMAX, LTE and WLAN with the R&S FSG and R&S FSQ. For receiver tests a wide range of signal generators is available as well (see Ref[2] and Ref[3]).

Johan Nilsson is a product manager for spectrum analyzers for Rohde & Schwarz in Munich. He can be contacted at Johan.Nilsson@rohde-schwarz.com. Dr. Wolfgang Wendler joined Rohde & Schwarz in 2004 as a product manager for spectrum analyzers. Dr. Wendler can be contacted at Wolfgang.Wendler@rohde-schwarz.com.

References
[1] IEEE 802.11n: "All Signals for Development, Production, Service", news 195, Rachid El Assir and Simon Ache
[2] MIMO receiver tests using only one signal generator, news 193, Dr. Jan Prochnow
[3] From SISO to MIMO – taking advantage of everything the air interface offers (2), news 194, Josef Kirmaier