Parametric Frequency Conversions for Broadband Spread Spectrum Signal Processing in RF Front-Ends

This paper demonstrates a correlation-based RF front-end approach that leverages pump-modulated parametric frequency conversion to enable continuous-time, waveform-selective signal processing. By treating the pump as a time-varying reference key, only input waveform components that remain correlated with the key are translated into a fixed, narrowband idler resonance where they accumulate coherently; uncorrelated components fail to build up and are suppressed. The accumulated energy is then reconstructed at RF using the same pump, enabling envelope and spectrum-preserving recovery of wideband modulated signals. A hardware prototype operating at 1.5 GHz demonstrates recovery of BPSK waveforms with bandwidths up to five times larger than the physical resonance bandwidth, overcoming the bandwidth-settling limitations of conventional front-end networks. These results establish correlation-driven parametric mixing as a viable approach for broadband spread-spectrum signal processing in RF front ends, with potential applications in despreading, and in-band self-interference suppression.