PnPSelect: Plug-and-play IoT Device Selection Using Ultra-wideband Signals
Zhaoxin Chang, Fusang Zhang, Jie Xiong, Ziyu Li, Badii Jouaber, Daqing Zhang · 2025
Read PnPSelect: Plug-and-play IoT Device Selection Using Ultra-wideband Signals by Zhaoxin Chang, Fusang Zhang, Jie Xiong, Ziyu Li, Badii Jouaber, Daqing Zhang (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2511-03534.txt.
Introduction
Zhaoxin Chang and colleagues propose PnPSelect, a way to pick a smart-home device simply by pointing at it. Crowded homes make scrolling through device lists slow and error prone. PnPSelect uses ultra-wideband radio on ordinary commercial gear, with no extra hardware bolted onto each device, keeping deployment cheap and truly plug-and-play.
Pointing direction is inferred from readings at a single anchor, a trick that avoids a full anchor array. A light localization step lets users register where devices sit with little effort. Together, those pieces turn selection into a natural gesture. The authors test the design on real hardware and argue it can scale from a few lamps to a house full of connected gear.
Catalog record
| CREATOR | Zhaoxin Chang, Fusang Zhang, Jie Xiong, Ziyu Li, Badii Jouaber, Daqing Zhang |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 61 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2511-03534
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
PnPSelect: Plug-and-play IoT Device Selection Using Ultra-wideband Signals CCS: Human-centered computing Ubiquitous and mobile computing systems and tools Thanks: This work is supported by the European Union through the Horizon EIC pathfinder challenge project SUSTAIN (No. 101071179) and the Innovative Medicines Initiative 2 Joint Undertaking project IDEA-FAST (No. 853981). Zhaoxin Chang, Fusang Zhang, Jie Xiong, Ziyu Li, Badii Jouaber, Daqing Zhang Abstract . In recent years, the number of Internet of Things (IoT) devices in smart homes has rapidly increased. A key challenge affecting user experience is how to enable users to efficiently and intuitively select the devices they wish to control. This paper proposes PnPSelect, a plug-and-play IoT device selection solution utilizing Ultra-wideband (UWB) technology on commercial devices. Unlike previous works, PnPSelect does not require the installation of dedicated hardware on each IoT device, thereby reducing deployment costs and complexities, and achieving true plug-and-play functionality. To enable intuitive device selection, we introduce a pointing direction estimation method that utilizes UWB readings from a single anchor to infer the user’s pointing direction. Additionally, we propose a lightweight device localization method that allows users to register new IoT devices by simply pointing at them from two distinct positions, eliminating the need for manual measurements. We implement PnPSelect on commercial smartphones and smartwatches and conduct extensive evaluations in both controlled laboratory settings and real-world environments. Our results demonstrate high accuracy, robustness, and adaptability, making PnPSelect a practical and scalable solution for next-generation smart home interactions. 1. Introduction With the rapid advancement of IoT technology, an increasing number of devices can now be connected to the Internet. A prominent example is smart home appliances, such as lights, air conditioners, and TVs, which enable remote control and automation. This trend has revolutionized how we interact with our surroundings, enhancing convenience, comfort, and efficiency in daily life. Unlike traditional devices that are typically controlled individually (e.g., through remote controls and switches), IoT devices within the same environment can be managed through a unified user interface. Currently, in many smart home environments, smartphones or voice assistants serve as primary interfaces for controlling IoT devices. For instance, users often select devices from a list displayed on a smartphone. However, this method can be time-consuming and inconvenient, particularly as the number of smart devices increases and similar devices (e.g., multiple lamps in the same room) become more common. On the other hand, voice assistant-based solutions offer convenience but face challenges such as limited working range, disruption in quiet environments, and susceptibility to noise interference. Recent research has explored alternative interaction methods that utilize common household devices, such as smartphones, smartwatches, and earphones. These methods allow users to select an IoT device by looking at or pointing toward it, offering a more intuitive and user-friendly experience. Several solutions ( 8 ; 5 ; 22 ; 28 ) leverage smartphone cameras for IoT device selection, using computer vision (CV) algorithms to recognize objects in the device’s field of view (FoV). However, camera-based methods may struggle in low-light conditions and are limited to recognizing devices present in predefined training datasets. Moreover, distinguishing between devices of the same type (e.g., two identical lamps in different locations) remains challenging for deep-learning-based CV algorithms ( 31 ) . The latest research took a different route by leveraging information from wireless signals for IoT device selection. The key insight behind wireless signal-based solutions is that, by installing dedicated wireless modules on each IoT device, a wireless channel can be established between the user device and IoT devices. The device being pointed at or looked at can then be determined by extracting signal features such as Time-of-Flight (ToF) and Angle-of-Arrival (AoA). For example, ToF estimation allows the system to measure distance variations between the user device and IoT devices during a pointing gesture. The device that exhibits the most significant distance change is likely the one being pointed at (Figure 2a ). Similarly, AoA measurements help determine the direction of IoT devices relative to the orientation of the user device, allowing the system to identify the selected device based on which AoA value is closest to zero (Figure 2b ). Unlike camera-based approaches, wireless signal-based solutions maintain high accuracy regardless of lighting conditions. Additionally, similar devices can be distinguished more easily because each IoT device maintains a unique wireless channel with the user device. Moreover, the computational complexity of signal processing is lower than that of image processing. Various wireless technologies, including UWB ( 1 ) , acoustic signals ( 38 ) , and Bluetooth ( 47 ) , have been explored for this purpose. While promising, a significant limitation of existing wireless signal-based solutions is the requirement that each IoT device be equipped with a dedicated wireless module, such as a UWB transceiver or a Bluetooth module. However, most smart home appliances do not currently include UWB transceivers, and Bluetooth-based solutions require compatibility with specific Bluetooth protocols (e.g., Bluetooth 5.1), which limits applicability. Therefore, the requirement for each IoT device to be equipped with dedicated hardware presents a major barrier to achieving a true plug-and-play solution for real-world deployments. Figure 1. The envisioned IoT device selection application scenario. In this paper, we propose PnPSelect, an IoT device selection solution that addresses this limitation, bringing wireless signal-based systems one step closer to real-world deployment. Specifically, we ask a question: Can the selection of a large number of IoT devices be achieved with a single dedicated wireless device? Figure 1 illustrates the envisioned application scenario. The user can simply point at the IoT device they wish to control using their device (e.g., a smartphone). By analyzing the connection between the user device and a single wireless device, the pointing direction of the user can be estimated for selecting a specific IoT device. Compared to existing solutions, this approach significantly reduces deployment costs and complexity. To realize this vision, the first step is to determine the most suitable wireless signal for device selection. While various wireless technologies (Wi-Fi, Bluetooth, 4G/5G, and UWB) are available on commodity devices, most are primarily designed for communication purposes. In contrast, UWB technology is inherently designed for precise ranging, and using it for device selection does not interfere with its original function. Notably, UWB modules are now integrated into a wide range of consumer devices, including those from Apple, Google, Huawei, Samsung, and Xiaomi. Therefore, we choose to utilize UWB signals available on commercial devices for IoT device selection. Achieving IoT device selection with a single UWB device in the environment presents several challenges. The first one is estimating the pointing direction of the user device. Since IoT devices are not equipped with UWB transceivers, the user device lacks a direct connection to them and cannot rely on conventional ToF or AoA measurements. Instead, we leverage a key observation: when a user performs a pointing gesture, UWB signals can be used to track the movement of their device. By analyzing the trajectory of this movement, we can infer the pointing direction. Building on this insight, we develop an algorithm that estimates the pointing direction using UWB readings from a single anchor. The second challenge is mapping the estimated pointing direction to specific IoT devices, which requires knowledge of their locations relative to the