Wi-Fi 7 might be stealing the spotlight, but that’s certainly not the only thing unfolding in wireless
With the certification of the Wi-Fi 7 just around the corner, the wireless industry is abuzz with excitement around what the new standard will enable. However, Wi-Fi 7’s arrival isn’t the only development that experts are keeping their eye on as 2023 comes to a close.
Below are three Wi-Fi trends that aren’t really about W-Fi 7.
Wi-Fi HaLow’s ‘big breakthrough’
Wi-Fi HaLow, or long range, low power Wi-Fi, is designed specifically for IoT applications. As the Wi-Fi Alliance explained it, it augments Wi-Fi by operating in spectrum below 1 gigahertz (GHz) to offer longer-range and lower-power connectivity.
“Wi-Fi HaLow enables the low power connectivity necessary for applications including sensor networks and wearables,” continued the Alliance. “Its range is longer than many other IoT technology options and it provides a more robust connection in challenging environments where the ability to penetrate walls or other barriers is an important consideration.”
While the standard was completed back in 2017, few real-life or laboratory measurements have been made for Wi-Fi HaLow due to the lack of equipment on the market. However, Spirent is seeing increasing interest in this technology and studies are starting to pick up, according to the company’s Principal Architect for Wi-Fi Steve Shearer.
“For instance, in agricultural settings, you may want to control your irrigation sprinklers a kilometer away and Wi-Fi HaLow allows you to do that,” he said, adding that Spirent’s testbeds are fully prepared to accommodate such use cases, even if right now most of the company’s instrumentation focus is on Wi-Fi 7. “HaLow [instrumentation] will pop up at some point,” he confirmed.
In addition to agriculture, Wi-Fi HaLow is also ideal for industrial control and in smart building applications, such as providing the long range and extended battery life needed for something like outdoor security cameras or sensors.
Quectel, too, noted rising awareness of this technology. “Quectel receives a lot of inquiries and sample demands from our customers for our module FGH100M HaLow module,” a company spokesperson shared with RCR Wireless News, adding that the company expects Wi-Fi HaLow to “experience a big breakthrough in 2024.” Quectel’s FGH100M HaLow module operates in 850–950 MHz bands with 1/2/4/8 MHz channel width and features 21 dBm maximum output power and 32.5 Mbps maximum transmission rate.
“Wi-Fi HaLow has a performance sweet spot,” Lazaros Kapsias, Quectel’s Wi-Fi department product manager, wrote in a blog post, explaining that it offers up to 78 Mbps capacity over short distances with 150Kbps achievable at up to 1km. “This makes it a real alternative to LoRaWAN, Zigbee, Bluetooth Low Energy and even narrowband-IoT (NB-IoT). Importantly, the data rate is sufficient for streaming video over the network, setting HaLow apart from LoRa, Sigfox and NB-IoT, plus of course standard Wi-Fi,” he continued.
Power consumption as a ‘critical metric’
For Broadcom, another trend worth following closely is the fact that power consumption is becoming “a critical metric” for those deploying Wi-Fi solutions. “In the past, we’ve always been building bigger, better, faster… but now we are seeing our customers push power consumption as critical to meet regulatory requirements,” said the company’s Associate Director of Product Marketing Kevin Narimatsu. As such, the chip maker is focused on providing lower-power modes in its products.
In addition to meeting tightening power efficiency requirements at the regulatory level, Andy Davidson, the senior director of technology planning at Qualcomm, pointed out that in order to support the growing user interest in multi-device experiences, power efficient Wi-Fi connections must also be supported.
“There’s an increasing expectation that your phone, tablet and laptop are all going to interact and you’re going to be able to use the same I/O [input/output] interfaces to control all of them,” Davidson continued. They would all be connected and the I/O would be able to move seamlessly to the device most suitable for what the user is trying to do and the compute would be where it makes the most sense.
“All of that requires connections between those devices, some of which do require very low latency, [especially] if your using the I/O from one device to control the screen of another,” he said, adding that as a result, multiple different Wi-Fi connections are needed — between the devices themselves and then from each devices back to the network. Without accounting for power consumption at the device level, Wi-Fi use cases like these — as well as things like untethered XR experiences — are not possible.
The integration of AI and Wi-Fi in the WAN
We’ve been hearing it all year: AI will touch everyone and everything in one way or another. And Wi-Fi is no different. Davidson confirmed that Qualcomm is researching how AI can improve Wi-Fi devices, but is also interested in how AI would impact the Wi-Fi connections themselves.
“In that area, there has been this constant push for higher throughput, better latency and then on to more deterministic latency. At some point, when Wi-Fi can meet the required latency. you can get into a more distributed compute,” explained Davidson. This, he said, is especially interesting for AI if a user is doing something like using one device as a display and another for the compute.
“AR is kind of the perfect example,” he continued. “The display is on your glasses, your user interface, the controls and input is coming from the glasses, too, but they are sent to a specialized compute device in order to process that information and send something back. That leads to very deterministic latency requirements as you’re moving our compute to a different place than the display. It is interesting to consider how this capability might be enabled in some AI applications where you might want to move the AI compute around to different places, like the edge or one hop away from the edge.”
For Broadcom, the future of AI and Wi-Fi means more integration in the Wireless Access Network (WAN). In October, the chip maker announced a collaboration with Comcast to co-develop an AI-powered access network with a new chipset that embeds AI and machine learning (ML) within the nodes, amps and modems that comprise the last few miles of Comcast’s network.
According to the pair, the new end-to-end chipsets will be the industry’s first to incorporate AI and ML capabilities into the WAN in such a way, and doing so will enable smarter network performance decisions, enhanced monitoring and issue detection, the transformation of network maintenance, improved customer assistance and the ability to monitor home IoT devices for connectivity disruptions. Trials are slated for early in 2024 and the companies expect to begin deploying the new chipset in live networks before the end of the same year.
“At the end of the day, if you’re only talking about Wi-Fi and you don’t have the capacity and the seamlessness all the way back to that cloud you’re really just talking about that last few meters rather than the whole data path,” summarized Christopher Szymanski, director of product marketing for Broadcom’s Wireless Communications Division. “So, I think as you start to see AI and Wi-Fi being integrated in the WAN, you’re going to see higher orders of QoS [Quality of Service] throughout the entire stack and more intelligent routing.”
