Sequential time interleaved random equivalent sampling for repetitive signal

Area of Science:

• Electrical Engineering
• Signal Processing

Background:

• Compressed sensing (CS) offers advantages for sparse signal spectrum sensing.
• Random equivalent sampling (RES) enhances efficiency but requires multiple runs for signal reconstruction.
• Existing CS-based RES methods are limited by considering only one sample per acquisition, increasing sampling time.

Purpose of the Study:

• To develop a more efficient CS-based sequential random equivalent sampling method.
• To reduce the number of acquisition runs and overall sampling time in sparse signal spectrum sensing.
• To improve the efficiency of non-uniform sampling signal reconstruction.

Main Methods:

• A novel sampling sequence is acquired in each RES run.
• A block measurement matrix is constructed using the Whittaker-Shannon interpolation formula.
• Block matrices are combined into an equivalent measurement matrix for all sampling sequences.
• Implementation utilizes a multi-core analog-to-digital converter (ADC) with time-interleaved cores.

Main Results:

• A prototype of the CS-based sequential random equivalent sampling method was developed.
• The system achieved an equivalent sampling rate of 40 GHz while physically sampling at 1 GHz.
• The proposed method demonstrated high efficiency for sparse signal spectrum sensing.

Conclusions:

• The developed CS-based sequential random equivalent sampling method significantly enhances efficiency.
• This approach effectively reduces sampling time and acquisition runs compared to traditional CS-based RES.
• The method is suitable for capturing high-rate analog waveforms efficiently, particularly for sparse signals.