Multiband sparse signal reconstruction through direct one-bit sampling.

Area of Science:

• Signal Processing
• Wireless Communications
• Information Theory

Background:

• Compressive Sensing (CS) reduces sampling rates but often requires high quantization bits.
• One-bit CS minimizes quantization bits but poses reconstruction challenges.
• Multiband sparse signals are prevalent in wireless communication, particularly cognitive radio.

Purpose of the Study:

• To propose a novel one-bit CS system for acquiring multiband sparse signals.
• To introduce a hardware-efficient system with simplified implementation.
• To enhance signal reconstruction algorithms for improved accuracy.

Main Methods:

• Development of the direct one-bit sampler (DOS) system, utilizing a single comparator at the Nyquist rate.
• Equivalence of the DOS reconstruction stage to a multicoset sampler.
• Proposal of an enhanced binary iterative hard thresholding (BIHT) algorithm for multiple measurement vectors in one-bit CS.

Main Results:

• The DOS system demonstrates hardware simplicity and efficiency.
• The enhanced BIHT algorithm effectively handles multiple measurement vectors.
• Accurate recovery of active subband positions and approximate estimation of their power for multiband sparse signals.

Conclusions:

• The proposed DOS system, combined with enhanced BIHT, offers an effective solution for one-bit CS of multiband sparse signals.
• This approach simplifies hardware implementation while maintaining high reconstruction accuracy.
• The findings have significant implications for efficient spectrum sensing in cognitive radio and other wireless applications.