Q: Please briefly introduce yourself, Samsung Research America, and the kind of work that goes on there. What projects are you working on?

I am Yuming Zhu, a member of the Standard and Mobility Innovation (SMI) Lab at Samsung Research America (SRA). I lead a small team within the lab that develops intelligent wireless connectivity solutions using Wi-Fi and ultra-wideband (UWB) in collaboration with the Mobile eXperience (MX) Business at Samsung Electronics.

We apply machine learning (ML), AI, and advanced signal processing technologies to improve users’ wireless connectivity experience. We also develop core technologies and use cases that utilize wireless signals to sense human activities and the environment and determine user or device location.

Q: Please tell us about the importance of your research field or technology.

The development of wireless connectivity technologies has dramatically changed our lives in the past few decades. Currently, on average, people spend more than three hours on mobile devices daily, and more and more of our devices are becoming connected (most of them wirelessly). According to a study, there will be 125 billion connected devices by 2030, averaging 15 devices for each consumer.

When the growing number of devices and data traffic loads encounter the increasing demand for real-time applications with ultralow latency, such as videoconferences, online gaming, and AR/VR events, it creates a significant challenge to ensure good user experience, especially in congested radio environments. The first area we worked on was enhancing user connectivity, particularly the Wi-Fi experience. We pioneered using ML/AI techniques to detect the underlying network requirement based only on data traffic patterns so that real-time data could go through the network smoothly, like a car in the high-priority lane on the highway.

With the ubiquitous availability of wireless signals (Wi-Fi, UWB, Bluetooth, mmWave), another hot trend in academia and the industry in recent years is to utilize these signals for sensing. One distinct advantage of using wireless signals for sensing is that they do not raise privacy concerns compared to other technologies, such as cameras. By monitoring the changes in wireless signals, we could detect motion in a room. It is also possible to detect human vital signs (respiratory rate, etc.) and human activities (gestures, falls, etc.) without attaching any device to users. Recognizing these potentials, the Institute of Electrical and Electronics Engineers (IEEE) 802.11 working group is passing an amendment (802.11bf) to facilitate Wi-Fi sensing functions. We have researched the use of Wi-Fi and UWB for sensing for several years, and it would be exciting to see a quicker and broader adaptation of these sensing technologies in phones and smart homes in the coming years.

Wi-Fi and UWB technologies can enable distance measurements among devices. To find its location, the user’s device can simply communicate with or listen to a few “anchors” whose locations are already known. They are effective candidates for providing localization or navigation services for indoor locations with weak or nonexistent GPS signals. The beauty of wireless signals is their effectiveness even in rooms that are pitch dark or filled with smoke or dust, as in some emergencies where other sensors (camera, LiDAR) could easily fail. We have developed algorithms and prototypes for high-accuracy localization using Wi-Fi and UWB.

Q: Could you tell us about the main achievements and the most rewarding moments in your research or memorable episodes?

We have been extremely fortunate to work on these exciting, rapidly developing technologies and bring our research from the lab to commercialization, contributing to making Samsung’s products successful. The first memorable moment for me was in 2020 when we created a UWB post-processing solution based on ML and statistical filters, which were adopted in the release of the world’s first UWB-enabled Android phone (Galaxy Note 20). I was proud of our team members’ extreme dedication and close collaboration with our colleagues in headquarters, which helped us overcome the tremendous difficulty posed by frequent site closures due to COVID-19, the quickly evolving hardware design, and the tight timeline.

Another memorable moment was in 2023 when we played a part in three ML-enabled Wi-Fi features commercialized in the Galaxy phones despite our small team. Among the three features, we contributed to the key Wi-Fi 7 feature multi-link operation (MLO) in Galaxy S24 Ultra (the world’s first smartphone with Wi-Fi 7 certification). We frequently discussed refining innovative solutions and collaborating with colleagues in headquarters and other R&D centers around the world to resolve issues promptly. For that, I am beyond grateful for our talented and hardworking team members.

Q: What is your vision for the future, and what goal would you want to achieve at Samsung Research?

As quoted in Tao Te Ching, “The highest good is like water, nourishing all things and not contending with them.” Good connectivity technologies should also be like water or air; everybody benefits from them and is not hindered. We are plugging away with the mission to make wireless connectivity smoother, more efficient, and smarter. The technologies that connect us to the world need to understand users: who we are, where we are, what we are doing, and how we are doing it. Samsung is well positioned to bring about this “highest good.”