Sensing Without Borders: A Sensing-Centric Handover Scheme for Continuous WiFi Sensing

WiFi has emerged as a promising sensing medium for ubiquitous perception networks due to its widespread deployment, low cost, and device-free sensing capability. While significant progress has been made, most existing systems operate in a standalone manner, with sensing capability tightly coupled to single or fixed multi-device deployments, which inevitably hinders sensing continuity across large areas. In this paper, inspired by pioneering handover mechanisms in communication systems, we propose ScHO , a sensing-centric handover scheme that enables continuous WiFi sensing capability across large areas. Specifically, we first design a sensing-centric handover framework tailored for sensing tasks, e.g., wide-ranging fall detection, enabling uninterrupted sensing under centralized network control. By jointly considering user proximity and the sensing signal-to-noise ratio (SSNR), a theoretical model of sensing-centric handover boundary in multi-link sensing scenarios is formulated. We reveal that when multiple sensing links are associated with the same access point, the handover boundaries follow an Apollonian circle determined by the device deployment. To further enhance system stability and reduce handover delay, we design a novel sensing-oriented handover mechanism to mitigate frequent ping-pong handovers caused by SSNR fluctuations in overlapping coverage regions of adjacent nodes. Extensive experiments demonstrate that: (i) ScHO supports reliable device transitions with an average handover latency of 1.18 s and an overall failure rate of 25.6%; and (ii) even in a small-scale three-link sensing network, ScHO improves sensing quality by 1.67× compared to the single-node approach, while achieving approximately 1.8× higher resource efficiency than the multi-node method.