The variability of MIMO has spawned a variety of test solutions

The diverse and evolving nature of multiple-input, multiple-output (MIMO) antenna systems forces test companies to strive to stay ahead of industry requirements. Depending on where the testing is done, from academic and industry R&D labs to product qualification and manufacturing, or from ICs to base stations to handsets, the specific requirements vary considerably. The latest test and measurement techniques to verify cutting-edge performance in R&D labs, as well as cost-effective production testing for the United States and other regions, have recently changed.

Different MIMO Types

A lot has changed in the MIMO space in 2012. As part of smart antenna technology, MIMO provides higher data rates and better spectral efficiency (bits/second/Hz) without requiring additional bandwidth or increased transmit power. Of course, using multiple antennas and increasing data processing on both the transmit (Tx) and receive (Rx) sides increases system complexity and associated test requirements.

Perhaps the first issue when implementing MIMO testing is understanding the variations that exist. The different types of MIMO and their confusion with other multi-antenna technologies are common (see Table 1).

MIMO is part of 4G/Long Term Evolution (LTE). The main difference between the two different types of MIMO, time division (TD) LTE or LTE TDD (TD duplex) and frequency division (FD) LTE or LTE FDD (FD duplex), is the type of spectrum required to deploy each protocol, said Erik Org, senior marketing manager at Azimuth Systems.

"When using frequency division LTE, you need paired spectrum, one band for uplink and another band for downlink, because uplink and downlink work at the same time," said Org. Each radio transmits and receives signals at the same time. When using time division (TD), transmission occurs in one time slot and reception occurs in another time slot. Therefore, paired spectrum is not required to implement the TD protocol.

When using multi-user or MU MIMO (802.11ac), rather than doubling the data rate for a single user, two users share the double data rate. "It's not actually increasing the available data rate, it's increasing coverage," said Mike McKernan, product marketing manager at Spirent Communications. "You can use a single MIMO channel to support another user in this cellular system."

“With MU MIMO, you can have multiple receivers, so you can have a single transmitter with four antennas and multiple receivers (up to four), each with its own antenna, or even a combination of three receivers, one with two antennas,” said Raajit Lall, product marketing manager for RF and wireless test at National Instruments (NI) (Figure 1).

Figure 1: Single-user MIMO allows multiple antennas to be connected to just one device. Multi-user MIMO allows multiple users to act as spatially distributed transmission sources, thereby increasing efficiency.

The difference between lab test results obtained by handset manufacturers, which typically use conducted or wired test methods, and field test results by service providers has led to the development of MIMO over-the-air (OTA) testing. The 3rd Generation Partnership Project (3GPP) standards organization is working to clarify how to bridge the gap between lab and field testing through MIMO OTA testing, said Jung-ik Suh, wireless marketing program manager for Agilent Technologies' electronic measurement division.

Test equipment vendors offer different OTA test methods. Agilent's two-stage MIMO OTA test method attempts to maximize the user's return on investment (ROI) and improve OTA test results. "There are roughly three recommendations for MIMO OTA testing, and 3GPP is working with test vendors such as Agilent to determine which is the best approach," Suh ​​revealed.

Beamforming does not necessarily require MIMO. "Beamforming, which uses multiple antennas, can improve wireless system performance by directing beams, which not only increases data rates and coverage, but also causes less interference to the intended receiver," Suh ​​said. However, when combined with MIMO, MIMO beamforming has become a planned part of TD-LTE and LTE-Advanced technology.

"The initial LTE-Advanced R&D work focused on carrier aggregation, with the goal of providing wider frequency bands and achieving higher data rates of up to 1Gbit/s. However, some leading R&D teams are also developing up to 8×8 MIMO," Suh ​​pointed out.

Issue: Specification Status

Current testing needs to take into account many standards that are not yet finalized and approved. As part of the MIMO LTE-Advanced verification process, CTIA, the international wireless industry association, is conducting reference antenna testing. CTIA uses known good devices and known bad devices to compare test equipment.

"These known good and known bad devices will be distributed so that CTIA can verify that people are getting the results they expect, and there will be a lot of work in this area," said Spirent's McKernan.

TD LTE MIMO is also continuing to improve. Although the industry has defined a wide range of frequency bands for the use of FDD and TDD, there are many bands that have not been used so far. "Last year, people's attention was more focused on the TDD 38-41 band." Azimuth's Org pointed out. He hopes that some changes will occur in the 2.5GHz to 2.7GHz band. Operators around the world, especially Chinese operators, will consider adopting TD LTE because this technology has the advantage of asymmetrically allocating spectrum or allocating capacity.

In the wireless LAN (WLAN) space, the Wi-Fi Alliance has largely determined which MIMO versions are mandatory and which are optional. For the latest WLAN standard, 802.11ac, testing 3×3 is becoming a mandatory requirement for many chipset vendors. "3×3 is definitely a requirement for the latest WLAN standards," said Lall of National Instruments. "LTE does not require 8×8, but many R&D labs have conducted testing for it."

Some experts believe that 802.11ac may not be fully completed by the end of 2013. Jan Whitacre, LTE market planning manager for Agilent Technologies' electronic measurement division, believes that finalization is at least a year away. No vendor recommends waiting until the specification is finalized to purchase test equipment.

Testing Strategy

While many of the MIMO changes that are happening are uncertain, they are temporary. Now is the time to implement the changes in testing. The expected changes to MIMO have been announced for several years. In many cases, they will eventually or urgently be approved. In addition, when test companies develop test equipment to address known or planned changes, the ease of use (for setup and use) of the equipment or platform is greatly simplified.

In some cases, only software changes are required. For example, National Instruments has prepared its platform design to meet the requirements of 10×10 and higher systems for future implementation. Therefore, customers who purchase such equipment now can still meet changing test requirements many years later.

“That was a conscious decision and the primary idea behind the PXI platform,” said National Instruments’ Lall. “There is no technical limitation to scaling it to 20x20, and we have certainly tested a 10x10 system.”

Granted, companies that cannot decide to purchase test equipment right away because all the conditions may not be in place are also facing a difficult decision today. If a company waits too long to purchase equipment to test to the final specification, it is likely to fall behind its competitors. If equipment providers are closely connected to specification development and consider specification requirements in their platform and product development, the risk of purchasing test equipment in advance can be minimized. [page]

The test equipment supplier's approach to adaptability can be to use built-in functions or easily upgrade to minimize costs and disruptive impact on customers. Of course, if the decision is wrong, the user may have to start from scratch to get an effective test device, so this is not an easy decision to make. In addition, if the current test is 2×2 or 4×2, it may seem premature to discuss 8×8 or even 4×4 for some test customers. But test equipment suppliers have already felt the increase in user interest and orders in 2012.

Even if they participate in standards discussions, customers who wait for specifications to get closer to finalization may be reaching their limits. "I thought probably last year that they were getting desperate, like we can't wait," said Nigel Wright, vice president of wireless marketing at Spirent Communications. "The specifications are definitely starting to converge on one approach, so it's a lower-risk strategy."

The timing of purchases is an ever-present question in the test equipment business. “When will people actually invest? Will they wait until the standards are finalized and take the risk that their competitors have already come up with a solution and are working on developing equipment that performs better?” Wright asked.

Other vendors also acknowledged the increased customer interest. “8x8 MIMO is still in the R&D labs. We have been involved in a lot of activities around 802.11ac this year,” Lall said.

The question is how users can avoid buying equipment now that will limit their flexibility and ability to implement the next phase of the standard.

"Today, my existing solutions can support any feature that will be commercially deployed in the next few years," said Org of Azimuth. It takes several years for feature changes and new topology options to propagate through the ecosystem, so the current commercial equipment can fully support it. "The highest order MIMO required by LTE today is 8×2 (8 base station antennas/2 user equipment (UE) antennas), which can support up to two spatial layers," Org pointed out.

Azimuth released a test strategy in 2012 that was built with buyers in mind. The company's planned future products, platforms and solutions will meet real-world test coverage and automated functional requirements. In addition to meeting user requirements for test equipment power consumption, noise, size and weight, Azimuth's platform will support the test challenges of the latest protocols and network deployments, supporting up to 200MHz channel bandwidth and 16×16 bidirectional MIMO topology.

NI's Lall encourages customers who are considering test equipment to narrow their choices. "They should look for flexibility or upgradeability," he says. If the current requirement is a 2x2 system, users need to properly evaluate potential equipment to ensure that it can support a 4x4 or 8x8 system in the future. While this may seem obvious, Lall has a way to determine whether a potential choice is flexible enough.

“If you invest in a certain type of test equipment and your generator or analyzer doesn’t output an analog-to-digital converter (ADC) clock or a reference clock, that in itself is a red flag,” Lall said. Simply outputting the sampling clock is not enough to implement a complete MIMO system; the reference clock must also be output. “For the customer, that’s the first time they realize that the equipment I’ve invested in may not be upgradeable in the future.”

Agilent's Suh pointed out another problem that can occur if test equipment decisions are delayed: lack of data correlation. For example, chipset vendors and RF customers may blame each other when test results are inconsistent. Suh emphasized that a consistent and widespread platform, from the oldest stages to the current MIMO R&D labs, helps solve this problem.

Delaying solution completion due to testing issues also means delaying technology development and product launches. Therefore, suppliers and customers need to focus on finding the real problem areas and solving them quickly. Test equipment is an important part of this process.

Availability and evolving products

To support all MIMO types, the range of product options for implementing MIMO at any stage or for any terminal application continues to expand. For example, Agilent's products targeting LTE-A, LTE-FDD and LTE-TDD and the R&D cycle from design, baseband, RF, integration, protocol to verification/pre-conformance testing can fully meet MIMO test requirements.

Some systems cover up to 8 channels, others up to 2 or 4 channels. X-Series SG receivers have up to 16 channels. Some products have full functionality, others have only general functionality, especially in the transmit area.

Agilent's solution for TD LTE MIMO and beamforming is the N7109A multi-channel signal analyzer and the latest 89600 vector signal analyzer software (Figure 2). The N7109A already has WiMAX and LTE MIMO measurement functions. The new functions added to these functions can complete TD LTE beamforming and MIMO signal analysis for up to 8 channels.

Figure 2: Agilent Technologies’ N7109A multichannel signal analyzer can meet emerging multichannel LTE, LTE-Advanced, and MIMO RF measurement requirements.

In addition, the Agilent MIMO PXI vector signal analyzer developed for MIMO 802.11ac can provide analysis of signals up to 800MHz. This capability, coupled with the accuracy and speed of the instrument, can help R&D and test engineers validate their MIMO 802.11ac designs.

To support the company's latest test strategy, Azimuth Systems launched its first product, the ACE MX2 wireless channel emulator (Figure 3). Designed to simplify the complexity of the test lab, this channel emulator also meets power consumption, size, weight and noise constraints. For example, the size and weight are 60% less than previous solutions. An ACE MX2 device can handle both TDD and FDD with bidirectional operation, supporting up to 8×4 MIMO systems.

Figure 3: Azimuth Systems’ ACE MX2 and ACE MX channel emulators are both designed to solve MIMO test problems. With built-in real-time fading capabilities, the ACE MX2 MIMO channel emulator supports the implementation of complex multi-user MIMO testbeds and the evaluation of carrier aggregation.

China Telecom Technology Laboratory (CTTL) is an organization that conducts wireless testing and certification under the leadership of China Academy of Telecommunications Research (CATR). In 2012, the laboratory selected Spirent VR5 HD spatial channel simulator for TD LTE device testing, including advanced MIMO beamforming implementation. VR5 also provides the basis for other Spirent test equipment.

“The ability to create realistic MIMO beamforming channels, automate the associated rigorous phase alignment process, and simplify testing were key design considerations in our recently released Spirent VR5 solution,” said Spirent’s Wright. Built on the VR5 foundation, the MB5 beamforming test system supports testing of 8×2 and 8×4 MIMO beamforming systems and other applications requiring the advanced phase alignment specified in 802.11ac (Figure 4).

[page]

Figure 4: Spirent Communications’ MB5 MIMO beamforming test system can be used for base station, mobile device and research testing as well as development and design validation.

National Instruments demonstrated an 8×8 LTE test solution based on the PXI platform. The synchronization results were very impressive (Figure 5). “We have tested 10×10 systems daisy-chained across multiple chassis, and the phase offset between each channel is within 0.1°, which is really excellent even if you look at all the different wireless standards. This solution will exceed your needs.” Lall assured.

Figure 5: The NI test solution for 802.11ac WLAN supports 4×4 MIMO on both the Tx and Rx sides.

In addition to testing 802.11a/b/g/n devices, NI's 802.11ac WLAN test solution also provides the flexibility needed to test 802.11ac devices. The solution can meet various signal bandwidths in the transmit and receive directions in up to 4×4 MIMO configurations, including 20, 40, 80 and 160 (80+80) MHz. The company revealed that it is working with several early contact partners to test the latest 802.11ac devices, including silicon suppliers, OEM manufacturers and electronic manufacturing service (EMS) providers.

Time is an issue

MIMO is clearly making a lot of noise in the test world, and 2012 seems to be a turning point in terms of implementation. "We think the biggest change in MIMO is definitely on the wireless LAN side, as engineers in that space are working to implement 3x3 multi-user MIMO," said NI's LaLL. But as the number of specifications grows, time and technology issues remain.

"In the long term, MIMO in user equipment is one of the major test challenges in the future," said Agilent's Suh. He does not expect this challenge to be encountered in the short term because the industry is focusing on carrier aggregation first. Once this problem is solved, the attention of those organizations will shift to MIMO on the user equipment side.

Even though 802.11ac will not be finalized for at least a year, R&D companies and academia need to start measuring it now. "There will definitely be a lot of disagreements and arguments because it's not a simple issue," said Agilent's Whitacre.

Spirent's Wright expressed concern about the handset problem. "There are already physical spacing issues with the antennas because you need to separate the antennas to achieve decoupling, so you need a certain level of physical space related to the wavelength," he said. The operating frequency range from 700MHz to up to 3.5GHz will cause key RF implementation challenges.

However, there are signs that the industry may already have a solution from innovative antenna manufacturers. “Maybe there is a solution, but it seems to us that there are still daunting challenges to solve,” said Wright. Even if suppliers believe there is a solution, they still need to verify it through testing.