Electronic Design

Will Your 802.11n Products Have Next-Generation Performance?

The forthcoming IEEE 802.11n standard will specify next-generation Wi-Fi products with performance that greatly exceeds current solutions. One of the key technologies used in 802.11n is multiple-input multiple-output (MIMO) technology.

MIMO has the potential to boost throughput beyond that of traditional wired Ethernet connections, significantly increase the range of Wi-Fi devices, and dramatically improve quality of service (QoS). The advances in the draft 802.11n specification also include beam-forming techniques, new medium-accesscontrol (MAC) technology, and new power-save modes.

The resulting products will be able to handle new bandwidthintensive applications at greater distances with improved reliability. In addition, 802.11n products will expand the capabilities for Wi-Fi platforms and applications already in widespread use. Realizing the promise of 802.11n, however, hinges on the availability of interoperable products that can deliver next-generation wireless applications.

The performance, conformance, and certification testing methodologies that are standard fare for legacy Wi-Fi products face new challenges as the industry transitions to next-generation technology. Specifically, using MIMO technology adds layers of complexity that require new testing techniques.

For vendors to truly deliver the performance promises offered by the 802.11n standard with new applications like video over wireless, they must be particularly diligent in testing their products—not only during the quality assurance (QA) process, but also throughout the design, development, and verification cycles.

Imagine a busy workplace filled with cubicles, office equipment, furniture, and employees. Signals from the wireless local-area network (LAN) fill the room and bounce off these obstacles, causing the transmissions to take different paths before arriving at their destination (Fig. 1). This phenomenon is known as multipath.

Multipath causes each of the reflected signals to arrive at the receiver(s) at different times and with different strengths. Typically, each path is characterized by a delay in time, a change in amplitude, and other more subtle factors (such as the angle of arrival, angle of departure, or angular spread).

Multipath isn't easy to predict and control because it's affected by everything from building construction to the movement of people. The performance of Wi-Fi products based on single-input single-output (SISO) technology is degraded by multipath, which affects signal quality and the connection's robustness.

On the other hand, the MIMO technology in the draft 802.11n specification takes advantage of multipath to improve network performance. Multiple transmitters send independent data streams at the same frequency (channel) at the same time.

Because of multipath, these multiple transmissions arrive at the multiple receivers at slightly different times, amplitudes and phases. MIMO algorithms use these differences to decorrelate the original signals. This technique, known as spatial multiplexing, can double or even triple a system's throughput.

MIMO's multiple receivers distinguish between multiple signals and allow parallel signals to be received (Fig. 2). Besides providing greater throughput, MIMO receiver chains can better reconstruct weak signals that have traveled a greater distance. This will yield a distinct range advantage over SISO systems. While multiple antennas and multiple simultaneous data streams improve MIMO-system performance, they create complexities in MIMO-product performance.

Controlled, repeatable testing of Wi-Fi products requires an isolated RF environment, motion emulation, and a test executive that ensures consistent test execution and results recovery via automated methods and archivability. Thorough performance characterization should include the controlled emulation of all environmental variations.

Therefore, to properly evaluate the performance of MIMO-based products in a real-world environment, multipath must be incorporated into the testing process. A high-level plan for functional testing of 802.11n-based products has four components:

  • Basic functionality
  • Compliance with the 802.11 specification
  • Verification of product interoperability and backward-compatibility with 802.11a/b/g
  • Characterization of product performance

The first three of these tests can be performed with good results in an almost entirely unimpaired RF environment (an environment with no multipath reflections). Since MIMO relies on multipath and other channel impairments to increase throughput, throughput performance should be measured under well-defined channel conditions with impairments. This requires an additional test tool—a channel emulator—to characterize the performance of these products in impaired environments.

Functional tests prove the product design across the full set of features and functionality. For example, one basic functional test will test device throughput over distance. The shape of the resulting plot is determined by the algorithm used by the product to select its transmission data rate. Other important Wi-Fi functional tests include:

  • Roaming and end-user mobility tests that verify a client device's ability to successfully roam between access points
  • Security testing, which focuses on authentication and encryption and measures the efficiency with which an access point manages simultaneous authentication requests
  • QoS protocols, which ensure that the network properly prioritizes voice traffic and accounts for roaming speed, jitter, and network delay
  • Network behavior tests that measure performance under abnormal network conditions, such as congestion, overload, and errors
  • AP packet forwarding rate, which measures the processing power of the AP and its affect on data throughput

Basic functionality testing requires equipment that measures results in an isolated RF environment. It also must be able to emulate devices in motion and control the performance parameters of interest.

Table 1 summarizes a few of the 77 Modulation Coding Schemes (MCS) from the latest draft of 802.11n specification. The throughput/range performance of a product will depend on its implementation of coding schemes. The mandatory schemes must be implemented for the product to be specification-compliant. There are 576 possible data-rate configurations in the current draft. Products will only interoperate at these configurations when vendor implementations of these configurations match.

Compliance tests ensure that a product has been implemented per the established standard. They verify that all mandatory features and functionality (data rates, qualityof-service mechanisms, MAC functionality) are incorporated in the product. They also verify that optional features are properly implemented.

Currently, products can be tested to comply with IEEE 802.11a/b/g standards. But 802.11n compliance can't be tested until the specification is finalized. Device testing for specification compliance requires equipment designed to emulate a golden node (client or AP), while executing test plans generated and verified by established testing labs like the Interoperability Lab at the University of New Hampshire.

The 802.11 specifications include many optional components that can be implemented differently by various manufacturers. Disparate implementations can affect cross-vendor interoperability, even if both products conform to approved standards. The draft 802.11n specification contains more optional components than any previous standard, making interoperability testing that much more important. As an example, the 802.11n draft specification currently includes 576 possible data-rate configurations, while the 802.11g specification had only 12.

In the wireless LAN industry, the Wi-Fi Alliance serves the valuable purpose of defining the elements of interoperability, conducting tests, and granting the Wi-Fi Certified logo to products that pass their tests. It's not only important to prove interoperability with products based on the same technology, but also to prove backward-compatibility with previous generations of Wi-Fi products.

As part of the design process, vendors of 802.11n-based products must rigorously test the interoperability and backward-compatibility of their products. Because the specification isn't yet approved and the Wi-Fi Alliance isn't yet certifying 802.11n products, vendors should verify that their products meet Wi-Fi certification for products as a/b/g devices.

In addition, as the industry prepares for the next generation of Wi-Fi certification, vendors are performing their own draft-802.11n interoperability tests. Confirming basic compatibility can be done without environmental impairments. That said, the throughput performance of different MIMO products will depend on the environment, and therefore it can only be tested using a channel emulator.

Interoperability and backward-compatibility testing requires test equipment that can measure results of the device under test communicating to other real products in an isolated RF environment.

As we have seen, a critical factor in 802.11n performance is its ability to leverage multipath. The maximum throughput performance of MIMO products can be measured in an appropriate setup without multipath, but a performance characterization (a more faithful representation of a real-world environment) must consider the multipath-effect. MIMO products should be tested in controlled impaired environments that mimic real-world situations. A test setup that directly connects a client with an access point will produce a very artificial environment, one not seen in real operating conditions.

To ensure repeatable results in a multipath environment, traditional Wi-Fi field tests should be replaced by methodologies that use a MIMO channel emulator. The channel emulator, common in cellular testing, will be essential for testing the performance of 802.11n products because it emulates multipath as experienced by real-world wireless devices.

A channel emulator is a laboratory tool that emulates "RF channels" by reproducing the multipath characteristics of real-world environments. Though an unlimited number of channels exists in the real world, the IEEE 802.11n task group has defined six mathematical models that represent typical Fi environments (Table 2).

Referred to as models A-F, these six channel models describe signal reflections, delays, fading, and other channel effects. They can be used as a baseline to benchmark product performance for various system parameters (e.g., the appropriate number of transmitters, receivers, antenna spacing, etc.).

A channel emulator provides a simulation of real-world environments. It typically does this by digitizing the RF signal from a wireless device (AP or client), impairing the digitized signal, and regenerating the impaired signal into an RF signal that can be read by an AP or client (Fig. 3).

A MIMO channel emulator can be connected as part of a broader test setup, but it's primarily a point-to-point device. This differs from the scalable testing platform that's typically used for the performance, conformance, and interoperability tests discussed earlier. A MIMO channel emulator is typically used to:

  • Test MIMO algorithms (baseband DSP algorithms) and debug errors
  • Optimize the performance of Wi-Fi devices in MIMO environments
  • Streamline QA processes for new MIMO products
  • Perform competitive performance benchmarks
  • Test interoperability between MIMO vendors

The channel emulator employs DSP technology to emulate signal reflections, delays, fading, and other channel effects. These effects are represented by statistical channel models defined by the IEEE 802.11n task group. The emulator provides a real-time emulation of the models to test how well the MIMO transmitter and receiver can exploit the multiple correlated paths in which the MIMO system will operate.

Overall, 802.11n technology holds tremendous promise for next-generation Wi-Fi products. The new standard will take advantage of MIMO and other technology advances to deliver greater wireless throughput and range—which is necessary to support emerging high-bandwidth voice, video, and data applications. The new specification will be more complex than existing 802.11 standards, which is important because testing 802.11n devices requires a more sophisticated testing methodology.

Comprehensive testing of 802.11n products will include specification conformance, basic functionality, interoperability, and performance characterization. Most of the conformance, functionality, and interoperability testing can be done without an impaired channel on the type of scalable system platform testers that exist today. Because of the complexities of MIMO environments, however, performance characterization requires the use of a channel emulator to simulate real-world channel impairments.

Channel emulation is critical to ensuring repeatable testing of MIMO technology because it submits RF signals to real-world multipath conditions. An effective channelemulation device can accurately create multipath behavior and measure its effect on system performance.

Vendors and design engineers looking to perform comprehensive tests on their next-generation products will need both a scalable system platform without multipath capability and a channel emulator that emulates multipath. These tools and a comprehensive test plan will enable the delivery of high-performance, interoperable MIMO products in the shortest possible timeframe.

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.