PARTNER CONTENT:  The Wi-Fi network has been widely used and become an important communication infrastructure in the current society. In fact, Wi-Fi can not only be used for communication, but also be used to perceive and measure the activity of a particular target in an environment, as it constantly sends wireless electromagnetic waves like radar. Researchers are actively researching the Wi-Fi Sensing technology to expand the use of Wi-Fi network devices.

Wi-Fi Sensing Technical Principles

Influenced by the environment and human activities, Wi-Fi signals have effects such as fading, shadow, and multipath during transmission. By measuring the linear transformation relationship between transmitted signals and received signals, the Channel State Information (CSI) that defines channel properties can be obtained. Based on signal processing, feature analysis, and deep learning technologies, Wi-Fi Sensing filters signal noise and extracts features from CSI, and then identifies the activity status, action types, and activity patterns of a person in the environment.

Wi-Fi Sensing Technical Features

Compared with traditional cameras and infrared technologies, Wi-Fi Sensing has the following technical features:

  • Easy deployment and low investment. Wi-Fi networks have been deployed in most public places and homes in society today, and the existing Wi-Fi routers and Wi-Fi CPEs can form a Wi-Fi Sensing network.
  • Passive, non-contact, non-invasive monitoring. The monitoring process is insensitive to the monitored object, and sensitive information such as voice, image, and video will not be collected to avoid leakage of personal privacy.
  • Non-line-of-sight monitoring. Compared with cameras and infrared sensors, Wi-Fi Sensing is not affected by light. The 2.4 GHz and 5 GHz electromagnetic waves have better diffraction characteristics, which enables it to overcome the problem that walls block signals in local environments. It can also work in the dark, weak light, along with privacy-sensitive places such as washrooms, which are not suitable for installing cameras.

Wi-Fi Sensing Applications

In recent years, mainstream Wi-Fi chip vendors have gradually launched chips that support real-time output of the CSI information to upper-layer applications. Wi-Fi router vendors at home and abroad have launched routers with the capability of sensing the CSI. At the same time, telecom operators have deployed cloud Wi-Fi Sensing services, which support real-time analysis of CSI flow data, statistics of activity patterns, and exception alarms. At the same time, the IEEE-802.11bf working group is actively promoting the standardization of Wi-Fi Sensing, developing the standard format and process of collecting and distributing the CSI in the Wi-Fi network.

It can be concluded that the implementation and application of the Wi-Fi Sensing technology has a certain foundation. With the rapid development of AI and big data processing technologies, we expect to see more products integrating the Wi-Fi Sensing technology in our lives in the Integrated Sensing And Communication (ISAC) era, which will bring convenience to people’s lives in terms of home security, intelligent elderly-care, and health monitoring.

Home security

With ubiquitous Wi-Fi signals in the home, the Wi-Fi Sensing system can accurately determine whether anyone in the monitored area is active without increasing extra hardware costs. When the homeowner is at work and there is no one at home, or when the homeowner is sleeping at night, the Wi-Fi Sensing intrusion detection function of the home router can be enabled to ensure the security of the home.

Intelligent elderly-care

In terms of intelligent elderly-care, Wi-Fi Sensing has two major typical application scenarios: fall detection and activity pattern analysis. Wi-Fi Sensing fall detection can timely detect falls of the elderly, inform their children and community assistants to take rescue measures in order to prevent tragedy. Based on the activity analysis capability of Wi-Fi Sensing, data of the elderly’s daily activity frequency, trajectory, amplitude, and pattern can be collected, and a behavioral model for the elderly can be established. When behavior suddenly changes, exception warnings will be given, and family members will be notified in advance to eliminate health hazards.

Health monitoring

The millimeter wave Wi-Fi device using the 802.11ad protocol has excellent range resolution, direction resolution, and Doppler speed measurement precision. By using the professional Wi-Fi Sensing algorithm for this type of device, the heartbeat frequency and breathing frequency of a person can be analyzed, and whether the person suffers from asphyxiation during sleep can be monitored. If the health data has an exception, the Wi-Fi network can be used to report the result in real time to gain valuable rescue time.

Conventional medical devices often require contact with the human body, and require medical professionals to perform operation and analysis, which results in high costs. In comparison, the Wi-Fi Sensing health monitoring device can implement all-time, passive and non-contact detection, and is characterized by convenient, cheap, real-time, and high accuracy, which can be used as a beneficial supplement to the professional medical devices.

ZTE has designed and trained a deep learning algorithm model that can precisely senses human activities and identify falls. Together with Wi-Fi chip vendors, ZTE is verifying the CSI sampling performance of Wi-Fi baseband chips and building CSI sample libraries related to several types of actions. Taking into account market and customer needs and specific application scenarios, ZTE is poised to combine these algorithms and models into commercial Wi-Fi router devices or cloud servers based on the 802.11bf WLAN sensing standard, and bring more commercial and social value to customers.