Wireless technologies such as Wi-Fi, GNSS, and cellular communications have become essential social infrastructure. However, current identification systems remain fragmented: humans are identified through biometrics, devices through device IDs, and spaces through positioning information from GNSS. These mechanisms are independently managed across domains, making unified entity recognition difficult in future societies where humans, IoT devices, and robots coexist and collaborate.

This presentation introduces a new perspective that redefines wireless signals as infrastructure for entity identification rather than merely communication media. Specifically, we present our research on radio frequency fingerprinting (RFF) technologies utilizing physical-layer characteristics embedded in wireless signals such as Wi-Fi and GNSS. Wireless signals contain features originating from transmitter hardware characteristics, oscillator imperfections, frequency offsets, and propagation characteristics. By combining these features with AI-based analysis, our research aims to realize identification technologies that recognize devices and entities without relying on identifiers such as MAC addresses or IDs.

We also introduce challenges including identification of unseen devices, recognition under randomized MAC environments, and cross-modal identification across Wi-Fi and GNSS. This research aims to realize practical non-biometric identification technologies for future societies involving electronically augmented humans, robots, and autonomous Physical AI systems.