Testing Challenges and Solutions for Accurate Wi-Fi 6 Chipset and Device Characterization

Wi-Fi 6 chipsets and devices harness the latest wireless standard with improved capabilities and performance. Enabling metrics such as faster data throughput and greater network efficiency may be favorable for consumers, but it places considerable demands on Wi-Fi characterization and compliance testing. When faced with test equipment that is not up for the task, engineers often resort to testing compromises, threatening measurement accuracy.

Channel bandwidths up to 160 MHz require instrumentation with adequate video bandwidth (VBW), modulation techniques necessitate statistical depictions, and long data streams with MIMO architectures demand time gating, extended measurement times, and multi-channel time alignment. This article will describe the various challenges associated with Wi-Fi chipset characterization, as well as their corresponding solutions that enable accurate power measurements to reveal the true performance of the latest generation of Wi-Fi. 

Wi-Fi-capable devices have become indispensable forms of technology in the connected society we live in today. According to the Wi-Fi Alliance, more than 4 billion devices shipped in 2019 alone, with an estimated 13 billion devices currently in use today. Wi-Fi’s prevalence has grown so much that it has been deemed the most commonly used wireless communications technology, as well as the primary medium for global internet traffic. Looking at its worldwide integration from a monetary perspective, Wi-Fi has surged in economic value over the years and now stands at an estimated $2 trillion. With a history spanning a little more than two decades, Wi-Fi has managed to effectively alter the way the world communications.

In 1997, the first version of the 802.11 protocol was released to consumers, and ever since the Wi-Fi standard has continually improved and evolved to enable a broader set of applications. The newest generation is known as 802.11 ax but has been given the simplifier moniker of “Wi-Fi 6” due to the Wi-Fi Alliance’s new naming structure. Wi-Fi 6 promises higher data rates, increased capacity, improved power efficiency, and greater network performance in dense or congested environments. To achieve such metrics, the latest standard of Wi-Fi utilizes channel bandwidths of 80 MHz or 160 MHz, high order modulation schemes such as OFDM and 1024-QAM, and MIMO architectures, among others. As Wi-Fi 6 pushes the capabilities of earlier standards, it also places new pressures on the test and measurement industry, presenting various RF power measurement challenges and tremendous demands on test instrumentation.

Essential RF Power Measurements to Assess Wi-Fi Chipset and Device Performance

RF power measurements are fundamental when it comes to the characterization of today’s advanced Wi-Fi 6 chipsets and devices. With wide channel widths up to 160 MHz, the VBW capabilities of test instruments must respond accordingly to accurately characterize Wi-Fi 6 chipsets, even at their highest level of performance. VBW denotes the ability of a sensor to follow signal variations of envelope power measurements. To deliver exact results, a sensor must respond fast enough to keep pace with the modulating signal’s rate of change in amplitude, and VBW insufficiency will fail to truly track the waveform, resulting in erroneous readings for envelope power, peak envelope power (PEP), as well as average power.

For reference, envelope power is a continuous function that represents a signal’s amplitude change due to modulation or distortion; it is more frequently referred to as peak power. However, sometimes the term peak power may refer to the highest point of the envelope power, and this singular value is known as PEP. Additionally, average power can refer to the average over an entire waveform or pulse repetition interval (PRI). However, oftentimes engineers want to measure the average power within a specific signal pulse or packet, referred to as packet average power.

In Figure 1, a modulated signal exhibits fast changes in amplitude, represented by a VBW-capable sensor in blue. The green waveform is the result of a sensor with insufficient VBW, which provided an inaccurate depiction of the pulse. 

Although a vital parameter for test instruments, there is currently only one company that offers RF power meters and USB power sensors with sufficient VBW to effectively capture the peak power of Wi-Fi 6 signals. This leaves the majority of test instruments lacking the necessary VBW widths for adequate Wi-Fi 6 characterization. Consequently, engineers often resort to measuring the average power as a substitute for peak power readings, but this testing compromise may mask compression of signal peaks and the resultant signal distortion. An equally unsatisfactory approach is to make measurements with very expensive and complicated instrumentation, such as a vector signal analyzer or Wi-Fi test set.

Linear Operation – Crest Factor and CCDF

Each evolution of wireless communication ups its capabilities and service quality to consumers, requiring engineers to employ technologies that support high speed data transmission, mobility, and efficient use of available spectrum and network resources. For example, multi-carrier modulation techniques help realize the latest wireless and telecommunications standard, however these modulated waveforms often affect the peak-to-average-power ratio (PAPR), also known as crest factor. Crest factor is used to understand the severity of a waveform’s peaks by taking the ratio of its peak amplitude and average power.

In the case of Wi-Fi 6, the number of sub-carriers (or frequency tones) and the modulation of each influences the magnitude of a signal’s crest factor. As the number of sub-carriers increase, constructive interference results in a very large PEP value. Wi-Fi 6 can have 1,992 sub-carriers using up to 1024-QAM, and even though it may increase throughput across access points, it also produces signals with elevated crest factors. 

High crest factor waveforms can impact the linear operation of components along a transmission path within a Wi-Fi communications system, such as an amplifier. Therefore, engineers can uncover signal compression and maintain waveform fidelity by calculating this important metric. To ascertain an amplifier’s linear performance, engineers measure the crest factors of both the input and output signals of an applied modulated waveform with a peak power sensor. In Figure 2, the calculations on the left only differ by 0.2 dB, indicating linear operation. However, increasing the input power to the amplifier causes the output crest factor to decrease by a wider margin, as shown in the graph on the right. This gap in crest factor measurements indicates non-linearities in amplifier operation, which can reduce the dynamic range of a communications channel as well as lead to spectral regrowth.

Since crest factor is a single value (the maximum crest factor), it often does not paint the full picture of system performance for engineers and technicians. Providing additional yet related information, the complementary cumulative distribution function (CCDF) curve is used to show the rate of occurrence a signal spends at or above its average power level, or in other words, how frequently a specific crest factor will occur. Digitally modulated signals like those used for Wi-Fi transmissions are often hard to quantify due to their noise-like appearance in the time domain, and CCDF offers a statistical view of power levels to extract useful waveform data. 

In the CCDF plot depicted in Figure 3, the input channel in yellow (CH1) has a modulated signal with an average power of -11.9 dBm applied to the amplifier, and is represented in the corresponding CCDF curve along with the amplifier’s output power in blue (CH2). Users will need to set an application-dependent rate of crest factor occurrence that is of interest. Since 0.01% is a commonly chosen rate, Figure 3’s curve marker shows the crest factor of the input and output signals for 0.01% of the time, along with supplementary information in the table. Overall, the CCDF plot has verified linearity since the input and output crest factor curves are nearly identical.  

By increasing the input power to the amplifier, as seen in Figure 4, the output power’s crest factor measurement subsequently decreases by nearly 3 dB, as indicated by the leftward movement of the CH2 curve. The growing gap between channels signifies compression has occurred, pointing toward an amplifier with non-linear properties. Interestingly, the more common method of determining amplifier linearity, which assess changes in gain at various power levels, only revealed a 0.2 dB reduction in gain; amplifier gain is the ratio of average output power to average input power. This illustrates the importance of using CCDF plots to fully understand the real magnitude of signal compression. 

The high crest factor of Wi-fi 6 signals, which are fraught with time-domain randomness and inconsistencies, places increased demands not only on component design, but also on RF and microwave test equipment to perform critical statistical analysis that enables a comprehensive assessment of Wi-Fi chipset performance.

Time Gating, Time Duration, and Time Alignment

Oftentimes, it is preferred to zoom into specific portions of a signal instead of looking at the waveform in its entirety. For instance, engineers may want to focus on or exclude the leading preamble portion of a Wi-Fi signal and determine essential measurements like peak, average, and minimum power of the gated signal. Narrowing in on certain pulse portions can be achieved using gate qualifiers and delay options available using vendor-supplied or customer-developed power measurement software. 

Let us take, for example, OFDM Wi-Fi signals that often have a noise-like appearance in the time domain (see Figure 5). Random noise spikes and modulation dips that go above or below the user-defined gate threshold can falsely trigger readings to start or stop, altering the accuracy of measurement results. To minimize erroneous triggers, some software includes the ability to specify start and end qualifiers, which define a set amount of time a signal must satisfy the qualification requirements in order to begin or terminate power readings. Once a start qualifier is satisfied, users may want to take advantage of additional software parameters such as start or end delay, which changes the start or end times of a measurement by a preset duration. Both the start and end delays enable users to pick specific waveform portions, such as the preamble of a Wi-Fi signal, on which to take power measurements. Alternatively, markers may be used to define the start and end of each gate interval.

While time gating can aid in viewing specific portions of single Wi-Fi packets, waveform analysis is also necessary for the entire Wi-Fi data steam. Certain measurement software packages achieve this by taking numerous samples along the waveform, all of which are stored in a buffer. However, the buffer size is often limited by the sensor’s memory capacity, and therefore the observation window is reduced to restricted timespans, often less than 1 second. Using a methodology with constrained testing durations can fail to capture important Wi-Fi waveform anomalies, such as power droops, signal dropouts, and spacing drift. A more efficient technique is to discard non-relevant information outside the packet interval, storing only essential data for each packet, such as the minimum, average, and maximum power, as well as the start time and packet duration. By limiting the amount of accumulated data, samples can be collected and processed from a virtually unlimited number of consecutive packets, effectively eradicating bottlenecks caused by buffer size restrictions.

In addition to narrowing into portions of packets and employing extended measurement duration times, time alignment between transmitted packets across channels must also be considered when characterizing MIMO Wi-Fi architectures. MIMO utilizes multiple transmitters and receivers to transfer multiple data streams simultaneously, increasing data rates and spectral efficiency. To enable this capability, testing devices must deliver precise, time-aligned measurements for multi-channel signals. Within the test and measurement industry, some power meter-based characterization systems cannot address MIMO time alignment on their own, requiring either circuitry created by the customer or the purchase of additional test equipment, such as oscilloscopes, which can drastically ramp up costs.

However, alternative testing options exist where multiple sensors share a common time base on multiple synchronous or asynchronous channels. Users can set up, for example, multiple observations periods to determine if packets from different data streams align or overlap. Using this technique, distribution of the shared time base takes place through a simple cable connection between each sensor’s multi-function input-output ports, effectively providing time alignment capabilities as an inclusive feature without necessitating additional expenditures or supplementary hardware.

Revealing the True Performance of Wi-Fi Chipsets and Devices

Companies delivering leading RF and microwave test solutions, such as Boonton Electronics, can provide test equipment with the capabilities to properly characterize Wi-Fi chipsets and devices. For example, Boonton RF power meters and USB RF power sensors offer the widest VBW of 195 MHz, crest factor measurements and statistical analysis (CCDF), packet time gating, Real-Time Power Processing (RTPP) technology in conjunction with the Measurement Buffer Mode to captures data of almost infinitely long data streams, and Synchronized Independent Gate Mode for simultaneous, multi-channel packet power and time alignment measurements. Characterization of Wi-Fi chipsets is laden with various testing challenges, however, understanding its essential testing demands can help engineers and technicians eliminate compromises and unveil their Wi-Fi chipset’s true potential.