SpectralDNA
Primary Investigator:
Research Independant
Stephen Glombicki, Jacob Jannotta, Christopher Brantley
Abstract
SpectralDNA investigates whether an individual can be reliably distinguished using only the ambient radio frequency emissions produced by the devices they carry and use. We perform passive, non intrusive captures with software defined radio and extract observable features. These features include protocol fields exposed in broadcast or discovery traffic as well as signal and timing characteristics that persist across short capture windows. We then fuse the heterogeneous observations into a single composite profile per subject and evaluate how consistently that profile can be re linked across sessions and environments, even when some protocols employ rotation or randomization.
Our findings reveal that the combination of signals emitted by a single person's ecosystem of devices constitutes a multi-layered identifier effectively forming a unique "Spectral DNA".