Skip to Main Content
Recovering even a small amount of information from a broadband radio frequency (RF) environment using conventional analog-to-digital converter (ADC) technology is computationally complex and presents significant challenges. For sparse or compressible RF environments, an alternate approach to conventional sampling is analog-to-information (A2I) to enable sub-Nyquist rate sampling based on compressive sensing (CS) principles. This paper presents the Nyquist Folding Receiver (NYFR), an efficient A2I architecture that folds the broadband RF input prior to digitization by a narrowband ADC. The folding is achieved by undersampling the RF spectrum with a stream of short pulses that have a phase modulated sampling period. The undersampled signals then fold down into a low pass interpolation filter. The pulse sample time modulation induces a corresponding phase modulation on the received signals that is scaled by an integer modulation index that varies with the Nyquist zone (i.e., fold number), allowing the signals to be separated based on the measured modulation index. Unlike many schemes motivated by CS that randomize the RF prior to digitization, the NYFR substantially preserves signal structure. This enables information recovery with very low computational complexity algorithms in addition to traditional CS reconstruction techniques. The paper includes a comparison of seven other A2I architectures with the NYFR.