"Cognitive In The News"
July 17, 2023

Wi-Fi Sensing – The Next Big Evolution of Wi-Fi

Since the dawn of civilization, humans have relied on light, starting with fire, to see and comprehend the world. Our methods of illuminating our homes have evolved from igniting sticks in caves to using Wi-Fi. Light and Wi-Fi are related, both being on the electromagnetic (EM) spectrum, with light allowing us to see and Wi-Fi enabling digital communications. Today, Wi-Fi represents the second-largest volume opportunity in the EM spectrum after light. The average household had 22 Wi-Fi-enabled devices in 2022, accounting for shipments of nearly 3 billion Wi-Fi chips. In comparison, households typically use an average of 40 light bulbs, with around 2.5 billion shipped annually.

Homes were dark in terms of Wi-Fi energy before the introduction of Wi-Fi technology in 2000. However, homes are now saturated with an abundance of Wi-Fi energy, illuminating the spaces within. As this Wi-Fi energy is transmitted around the home, it reflects off objects and becomes attenuated as it moves through space and objects. In a static environment, this energy distribution is somewhat stationary, not changing during the transmission of a Wi-Fi signal. However, as a person moves around in the environment, the spectrum of energy distribution changes. These changes indicate there is motion within the home. In the home shown in the diagram, the client devices illuminate the space with Wi-Fi energy and the access point (AP) observes the changes in the Wi-Fi spectrum due to human motion.

What is Wi-Fi Sensing and How Does it Work?

Wi-Fi Sensing is the principle of measuring and classifying disturbances in the Wi-Fi spectrum over time due to a person’s movement. The sensing machine learning software to measure these disturbances is typically contained within the Wi-Fi APs. The Wi-Fi clients radiate the EM waves for sensing in the form of standard Wi-Fi communication data frames. The clients transmit signals back to the AP at the sounding rate, which can be adjusted to be slower or faster. There are limits to this sounding rate adjustability, however. If the rate is too slow, some of the motion will be missed, while a rate that is too fast can impact the Wi-Fi throughput performance as the sensing software within the AP consumes more CPU. To achieve good sensing performance, an optimal sounding rate is around 10 Hz or 100 ms. A significantly higher sounding rate may negligibly improve motion detection latency but at the expense of degrading Wi-Fi network data communication performance, making that an inadvisable trade-off.

Read the full article from the Microwave Journal here.