A Lower-Correlation Random Demodulator Architecture Using Random Sampling Period Zeroing Integration Method

Compressed Sampling (CS) overcomes the oversam-pling trap caused by using Shannon's sampling theorem in sparse ultra-wideband signals. The Random Demodulator (RD) is the infrastructure for compressive sampling, which uses a random sequence to induce an aliasing spectrum, then passes through a filter (equivalent integrator) and sampled at a low rate. Numer-ous scholarly works have been refined and optimized utilizing this framework. This work also proposes an optimized, lower-correlation RD framework named Random Sampling Period Ze-roing Integrator (RSPZI). It retains the aliasing sequences $P_{c}$ and period reset integrator of RD while using a randomly triggered holders (Matrix operations are accumulations of pseudo-random intervals) of Analog-to-Digital Converter (ADC) to smear the tones and increase the reconstruction probability. This work gives a possible solution for constructing an RSPZI and explains why the RSPZI can smear the tones, allowing us to recover the full frequency band from the baseband samples. Furthermore, we mathematically demonstrate that RSPZI has a lower correlation between the sparse and compressed matrices compared to RD. Therefore, the probability of exact reconstruction of RSPZI is higher. Experimental results support these conclusions.