Beam-Fi: Integrated Sensing and Communication via MU-MIMO upon Commodity Wi-Fi

While next generation Wi-Fi standard endeavors towards an integrated sensing and communication (ISAC) design, the novel multi-user multiple-in-multiple-out (MU-MIMO) technology pervasively adopted by new Wi-Fi standards has not been fully leveraged for ISAC sake. To this end, we present Beam-Fi as the first ISAC system directly exploiting MU-MIMO upon commodity Wi-Fi, distinct from recently proposed ISAC systems totally neglecting MU-MIMO. Essentially, Beam-Fi leverages MU-MIMO to generate beam patterns that simultaneously serve communication and sensing, bearing in mind the signal quality on both sides. To better schedule beamforming for ISAC sake, we first design a user discovery scheme to detect potential sensing users in addition to existing communication users, then we propose a user grouping strategy to pair sensing users with communication users demanding similar beam patterns. We further develop an optimization-driven beamforming design to ensure both communication and sensing quality. Finally, we implement a prototype of Beam-Fi on commodity Wi-Fi devices and demonstrate its superior efficacy through three typical applications: gesture detection, activity recognition, and respiration monitoring.