Những thách thức trong việc giảm độ trễ không dây
A new and much faster version of Wi-Fi is beginning to infiltrate the IoT market, reducing latency that has begun to creep up as more data is generated, processed, and moved wirelessly from one device to another.
An estimated 20 billion connected devices are currently in use. Over the next several years, devices will start to include faster wireless connectivity, enabling more rapid transfer of data, faster over-the-air updates, and significantly increased capability for streaming video, in-vehicle infotainment, and just about any data sent to a device that isn’t tethered to some sort of cable. Nearly all of those will require upgraded chips, chiplets, faster interconnects, and some type of advanced packaging.
“What usually happens in industries when the next generation of Wi-Fi is introduced into the market is the lead adopters of that technology tend to be the routers, along with all forms of PCs, such as laptops and tablets, as well as smartphones,” said Sivaram Trikutam, vice president of the Wi-Fi product line at Infineon. “Those three segments are typically the ones that tend to adopt a new Wi-Fi standard first. With Wi-Fi 7, we are at that stage now where only the top-of-the-line devices in a segment — the latest iPhone, the latest Samsung Galaxy phone, or the $3,000 laptop — have just started adopting 7. Most of the industry is still on Wi-Fi 6. For most of the home IoT devices, which typically are the ones that adopt the next generation of Wi-Fi last, such as robotic vacuum cleaners, door openers, and the solar panel on your roof, for all those kinds of devices, the Wi-Fi 6 transition is happening now. In the next five years or so we will see Wi-Fi 6 volumes growing. Maybe it will peak by the end of the decade, and then the Wi-Fi 7 devices will start increasing in volume around that time frame, so this is going to be the decade of Wi-Fi 6 in a broad way.”
As with all improvements in connectivity and faster data rates, chip designs will become increasingly complex. This is evident starting with the antennas, which will receive the data.
“People under-appreciate the effort needed to integrate antennas into devices,” said Matt Commens, principal product manager at Ansys. “An antenna wants to interact with its environment. Its purpose in life is to pick up and send signals from the outside. It wants to interact with the world. As the product designer, you can do analog-to-digital (A-to-D) conversion, D-to-A conversion, get your signal encoded, get it on to a transmission medium, send it to an antenna board, and it will look good.”
But that’s not enough. “One of the more challenging ones is ‘desense,’” Commens said. “What happens with desense is you’re designing your PCB, you’re looking at your signal. You have a port where the antenna is going to go, and you’ve got your vector network analyzer hooked up to that port, or you’ve got your signal analyzer hooked up to that port, and everything looks great. Then you hook up the antenna, and suddenly your signal-to-noise ratio drops, and you’re getting some sort of back interference from the antenna into your electronics on your PCB because, again, that’s what an antenna wants to do. It wants to send EM energy out. And so you’ve got a desense problem, which is intimately tied to the intent, creation, placement, and other key things associated with signal integrity. Worse, it can be unforeseen. You could be far into the design process before you detect that, when it’s more expensive to fix.”
Connectivity standards
It’s easy to get bogged down in acronyms with anything related to communications. But with Wi-Fi, there are four key standards for modern devices:
There are four Wi-Fi standards in use today:
- Wi-Fi 5 (IEEE 802.11ac), which was adopted in 2013.
- Wi-Fi 6 and Wi-Fi 6E, which is actually a version of Wi-Fi 6 that can additionally operate in the 6 GHz band. These were adopted in 2021 as IEEE 802.11ax.
- Wi-Fi 7 (802.11be), the newest one with a data rate of 40Gbps, which is more than 20 times that of Wi-Fi 5’s 1.3Gpbs.
WiFi 7 devices were already in the marketplace in late 2023, but the protocol’s official debut was at last January’s CES. It’s reasonable to assume that will inspire a lot of equipment upgrades. IDC, as quoted by the Wi-Fi Alliance’s Beacon Blog, predicted that last year two-thirds of shipments would be Wi-Fi 6 or Wi-Fi 6E. It projected that by 2025, more than 32% of all Wi-Fi 6 device shipments being Wi-Fi 6E.
The drop in latency in Wi-Fi 7 will both usher in new devices and make it easier to interact with AI-enabled ones. The difference is achieved by several technical changes, per the Wi-Fi Alliance’s overview listed below. (More in-depth descriptions may be found here.)
- 320 MHz super-wide channels, which are only available in 6 GHz, provide twice the throughput of Wi-Fi 6, enabling multi-gigabit Wi-Fi device speeds;
- Multi-Link Operation (MLO), which supports more efficient load balancing of traffic among links, resulting in increased throughput and enhanced reliability;
- 4K QAM, which achieves 20% higher transmission rates than Wi-Fi 6’s 1024 QAM for greater efficiency;
- 512 Compressed Block Ack, which improves efficiency and reduces overhead, and
- Multiple RUs to a single STA, which improves flexibility for spectrum resource scheduling to enhance spectrum efficiency.
For consumers and enterprises, all of this means the usual tradeoffs between taking advantage of the performance of new devices and trying to extend the useful life of older devices. For the Cellular Telecommunications Industry Association (CTIA), which represents the wireless industry, it means another round in the ongoing war over spectrum. Its most recent argument is that the design of Wi-Fi devices is now so advanced that no additional spectrum is required for optimal performance.
Standards and certification
FCC certification adds another constraint. “At the most extreme, everything’s designed and you’re ready to go to market, but then you fail the FCC standard,” said Ansys’ Commens. “At some point, you’re making a smart product. It’s going to radiate. There are very rigorous requirements about what energy can be within certain frequency bands. You can be blocked at the most expensive stage of the process to implement a fix.”
FCC certification is just the beginning of regulatory compliance. In addition to engineering standards bodies, designers face a worldwide alphabet soup of government agencies. “There are national bodies in countries from Afghanistan and Albania to Zambia and Zimbabwe. There are also various multinational regulatory groups in Africa, the Caribbean, and Europe,” said Brad Jolly, senior applications engineer at Keysight.
Additionally, there are battery standards. The International Electrotechnical Commission (IEC) has standards such as IEC 62133, which relates to the safety of lithium-ion batteries. Another IEC standard, IEC 62281 relates specifically to battery transportation, as does United Nations standard UN 38.3.
For cybersecurity, the U.S. Cyber Trust Mark, from the FCC, is one of the most influential IoT-specific initiatives, but there are others, including ETSI EN 303 645 and National Institute of Standards and Technology (NIST) standards 8259 and SP 800-213.
The number of regulations reflects both the technical complexity of wireless devices and the criticality of many IoT uses. “The rise in wireless communications, both in terms of the number of devices and the volumes of data, has made coexistence testing more important than ever,” said Jolly. “Coexistence testing ensures that devices can maintain functional wireless performance in the presence of other devices operating in the same physical area and portion of the wireless spectrum. Much coexistence testing is based on American National Standards Institute (ANSI) standard C63.27. The nature of coexistence for connected medical devices is exceptionally important, so the Association for the Advancement of Medical Instrumentation (AAMI) has created technical information report TIR69, Risk management of radio-frequency wireless coexistence for medical devices and systems, to guide medical device manufacturers in the coexistence testing of medical devices.”
Yet for all of that, there’s still a crucial element missing. “People do not think enough about what standards do not require and how persons with evil intent can exploit those omissions,” Jolly said. “For example, standard testing for electromagnetic compatibility (EMC) and electromagnetic interference (EMI) ensure that devices operate properly in their intended electromagnetic environments. Similarly, coexistence documents like ANSI C63.27 and AAMI TIR69 focus on assessing devices in their intended RF environments. All this emphasis on testing devices in their intended environments is certainly reasonable, but what about testing devices in unintended electromagnetic environments? For example, we are seeing cybercriminals using devices like deauthers [which demonstrate a vulnerability in the 2.4GHz WiFi protocol], jammers, and wireless hacking multi-tools to defeat security systems that rely on wireless connectivity, such as home camera doorbells and remote security cameras.”
Matter brings IoT ecosystems together
The great variety of IoT devices for the smart home and industrial automation markets has led to painful interoperability issues, frustrating both home consumers and professional IT admins. In response, the Connectivity Standards Alliance (CSA) an industry consortium, formerly known as the Zigbee Alliance, developed Matter, an open standard that allows such devices to seamlessly interconnect. Hundreds of companies have joined the Alliance, including Google, Apple, Amazon, Siemens, and Infineon.
“What Matter does at a very high level is bring all these various ecosystems together,” said Infineon’s Trikutam. “Whether it’s Apple’s ecosystem, Google’s, Amazon’s, or others, it provides compatibility across all these private walled gardens. You want the end consumer to be able to select whatever device they buy, from vacuum cleaner to lightbulb, based on the features and the price point of that product, and not have to worry about whether it will play with their Amazon Alexa speaker. Matter basically eliminates that as a care for the end customer. They can be confident their purchase will work with any of the ecosystems.”
The Matter standard was first published in 2022. Version 1.3 was published May 5, 2024, and there are plans for bi-annual updates. As Tobin Richardson, president and CEO of CSA said at its launch, “It is one huge step in a long process.”
Matter, which operates at the application layer of the OSI protocol stack, is compatible with transport protocols that operate at lower layers, such as Wi-Fi, Ethernet, and Thread. This means that while Matter may appear transparent to consumers, for network engineers there’s a lot going on below the surface.
As new devices are developed, and older ones dropped — such as happened with Nest and Google Home — designers and consumers alike must consider practical choices, said Ansys’s Commens. “There’s a push for the smart home, but do you really need a smart garage door opener? You don’t need to throw Wi-Fi at that because it would be overkill. There are other standards, like Matter, looking to address smart home needs that don’t necessarily need the full glory of Wi-Fi. It’s also a tradeoff for designers. If you’re designing a phone, do you want to put in one more chip in support of another standard?”
Designers wrestling with that question may be exactly what’s holding up universal Matter adoption. “The Matter protocol has the backing of huge companies and a clear plan to evolve the standard periodically, so it should have every opportunity to succeed,” said Jolly. “On the other hand, some information suggests not every manufacturer of smart home devices wants to deal with the cost and device architecture implications of Matter, so that could limit adoption.”
Connecting with light fidelity
Another communications technology that could also work with IoT devices and potentially overcome both latency and security issues is light fidelity, or LiFi, which communicates information using visible light, most often generated by an LED source.
“LiFi is part of a domain of photonics called ‘free space’ photonics/optics, which covers a lot of applications, including military satellite-to-satellite laser-based communications. LiFi for IoT is closer to home,” said Gilles Lamant, distinguished engineer at Cadence.
There was a great deal of buzz about LiFi in the last decade, when it was named one of Time Magazine’s “Best Inventions of the Year” for 2011. The technology was given a major boost with last year’s release of the IEEE 802.11bb standard that defined wireless light-based communications, especially as bandwidth congestion inspires a search for alternative transmission solutions. Visible Light Communication, as the LiFi industry group has dubbed the method, operates at the speed of light in a vacuum, while radio waves are slowed by traveling in air. In addition, since visible light occupies a wider part of the EM spectrum than the now congested radio waves reserved for communications, LiFi devices could potentially reduce the congestion problem, or at least not increase it.
Arguably the biggest advantage of LiFi, compared to Wi-Fi, is that it doesn’t pass through walls. That eliminates a potential attack vector by confining the signal within a room. This is a tradeoff between convenience and privacy, but it opens up markets such as military, medical, or other highly confidential applications.
A frequent objection to LiFi as a widespread solution is that, because LiFi is line-of-sight, signals could easily be blocked. As the LiFi group explains on their website, that claim is based on a misunderstanding, “Since light bounces off of surfaces, this means that LiFi is not strictly a line-of-sight technology. Of course, being in direct light is a definite advantage because the signal will be stronger, but the light will also bounce off walls and other objects and that reflection can also be used in data transmission.”
Increasing LiFi’s practical value, the standard was written to allow Wi-Fi and LiFi to work together on the same device, making them complementary rather than competitive, said David Plets, associate professor at imec-WAVES, Ghent University.
“LiFi is also a broadcast technology just like Wi-Fi,” Plets said. “It is subject to the same vulnerabilities. However, LiFi has the inherent benefit that light does not penetrate through opaque objects like walls, and that LiFi cells are typically smaller than Wi-Fi cells, requiring the eavesdropper to be very nearby. In the marketplace, LiFi might in the end complement Wi-Fi for specific use cases where Wi-Fi is less suited — for example, underwater communication, or in environments where RF communication is less desirable, such as hospitals, or for some high-rate new 6G applications like XR. For another example, it might be combined in hybrid networks, joining the ubiquitous connectivity offered by Wi-Fi with the high data rates offered by LiFi. As such, LiFi can offload some pressure on the Wi-Fi network. However, integration of both technologies is still for tomorrow. In the end, it will come down to finding a good match between LiFi’s technical performance and the right use cases.”
Related Reading
Wi-Fi 7 Moves Forward, Adding Yet Another Protocol
Wireless technology is getting faster and more reliable, but it’s also becoming more challenging to support all of the necessary protocols.
