Subnoise Signal Detection and Communication

Radio frequency spectrum is one of the scarcest commodities in existence, with progressively increasing value. As a physical foundation of an untethered society, it now carries the majority of social, defense, and commercial interactions. All of these must reside within narrow, strictly regulated spectral windows allocated for cellular, military, navigation, and broadcast services. Band localization minimizes interference but also mandates that the entire cellular traffic be confined in less than one percent of the physical radio-frequency range. To defy this restriction and emit freely in any band, the signal power must be small to avoid interference with existing traffic. By spreading the signal over a sufficiently wide spectral range, the emission in any band can be maintained below naturally occurring noise. Unfortunately, the reception of a spectrally broadened, subnoise data channel poses a fundamental challenge: a fast, bursty waveform must be detected, separated from noise and reconstructed at rates exceeding gigahertz. Here, we show that a 20-MHz-wide signal can be spread by 300-fold, detected and reconstructed by a physical Fourier transform even when it is much weaker than the received noise. Rather than quantizing the 6-GHz-wide signal and computing its correlation with the decoding waveform, the signal was physically detected and reconstructed by coherently coupled frequency combs. By eliminating high-speed electronics from the receiver, it is now possible to access the entire radio-frequency range that extends beyond 100 GHz. We anticipate that new, band-unrestricted wireless services will emerge to maximize throughput, mitigate interference, and achieve a high level of physical security.