A Linear-Time-Complexity Dynamic Histogram Method for Distributed Pulse Deinterleaving

Jie XUE

doi:10.1016/j.dsp.2026.106503

Return

A Linear-Time-Complexity Dynamic Histogram Method for Distributed Pulse Deinterleaving

2026-09-19

PRE

DOI：10.1016/j.dsp.2026.106503

Original

Original request for help

Save

Abstract

En 中文

• In terms of deinterleaving accuracy, a neighborhood search and adaptive bin merging mechanism is introduced to fundamentally eliminate TDOA counting distortion caused by measurement errors. A harmonic peak secondary processing is devised to remove residual TDOAs and suppress false alarms under pulse loss conditions. An unpaired pulse classification mechanism is further proposed to recover unlabeled pulses and improve deinterleaving accuracy. • In terms of operational efficiency, the pulse pairing strategy is optimized by exploiting the monotonicity of TOA sequences, reducing the pairing complexity from O ( N2 ) to O ( N ) . Single-pass histogram statistics with local incremental updates replaces global recursive reconstruction, lowering the overall time and space complexity to O ( N ) . • Theoretical analysis and results from 1000 Monte Carlo simulations demonstrate that ADEH achieves 99.9% deinterleaving accuracy with zero false alarm and missed detection rates under ideal conditions. The average runtime is only 1.4% of that of conventional TDOA histogram methods, and scales linearly with the number of pulses. Under pulse loss conditions, the algorithm maintains zero false alarms and zero missed detections, exhibiting strong robustness against complex PRI modulations and non-ideal reconnaissance environments.

Keywords:

Pulse Deinterleaving

Electronic Reconnaissance

Distributed System

TDOA (Time Difference of Arrival)

Pulse Pairing