High-speed real-time periodic weak pulse signal detection with simplified phase-weighted stacking

Abstract

Phase-weighted stacking (PWS) is an efficient noise reduction technique widely used in exploration seismology. It uses the coherence of the instantaneous phase to enhance signals by reducing incoherent noise. However, the high computational cost makes it difficult to apply PWS for the real-time detection of weak signals with high repetition frequency. This paper proposes a novel simplified PWS method with low computational complexity. The complex plane is divided into four quadrants, and the instantaneous phases in the same quadrant are simplified to the same phase. Based on the proposed method combined with a one-bit analog-to-digital converter, a novel field-programmable gate array-based high-speed real-time periodic weak pulse signal detection technique is presented. A prototype is implemented to verify the proposed technique. For weak pulse signals with an input signal-to-noise ratio (SNR) of −21 dB to −15 dB for 2500 cycles, the results obtained show that the simplified PWS algorithm can improve the SNR of the coherently integrated signals by about 10–15 dB with a latency of about 2 µs.