Abstract: As a novel distributed optical fiber sensing technology, Phase-sensitive Optical Time-Domain Reflectometer (Φ-OTDR) offers advantages such as high sensitivity, high spatial resolution, long sensing distance, and wide dynamic range. It is widely applied in scenarios including oil and gas pipeline safety monitoring, perimeter security, large-scale structural health monitoring, railway infrastructure maintenance, and microseismic wave detection. Compared with conventional uniform sampling, non-uniform sampling strategies can overcome the inherent trade-off between sensing distance and response frequency, thereby enhancing the system's high-frequency detection capability. In Φ-OTDR vibration sensing, the performance of the moving differential algorithm depends critically on the selection of the differential interval. Experiments using arithmetic progression incremental sampling revealed that recognition results periodically exhibited extremely low signal-to-noise ratios for specific differential intervals. Theoretical derivation elucidated the causes of this phenomenon. Consequently, a random permutation arithmetic progression interval sampling method was proposed. Numerical calculations and vibration sensing identification results indicate that this method maintains convenient sampling interval settings while avoiding excessively low signal-to-noise ratios and the resulting missed detections.