Scholarlink

To effectively monitor unauthorized unmanned aerial vehicles (UAVs), this paper proposes a passive localization system based on distributed radio frequency sensors. First, a Multi-Stage Search Cross Ambiguity Function (MSCAF) algorithm is introduced to reduce the mean absolute error of Time Difference of Arrival (TDOA) to 1.14 ns through coarse and fine search mechanisms combined with frequency offset compensation, thereby enhancing the accuracy of time delay estimation. Second, an improved version of the Beluga Whale Optimization (IBWO) algorithm is employed to address the nonlinear localization problem by integrating Halton sequence initialization, adaptive Levy flight, and a hybrid particle swarm optimization strategy. Simulation results demonstrate that the IBWO algorithm achieves a positioning error of 1.39 m at a signal-to-noise ratio of 20 dB, outperforming the Chan-Taylor algorithm, which yields a positioning error of 2.29 m.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A UAV TDOA Localization Method Based on Distributed RF Sensors

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Junhao Liu,
  • Junyu Wei,
  • Shaojing Su,
  • Junhao Ba,
  • Zongqing Zhao,
  • Qing Liu

摘要

To effectively monitor unauthorized unmanned aerial vehicles (UAVs), this paper proposes a passive localization system based on distributed radio frequency sensors. First, a Multi-Stage Search Cross Ambiguity Function (MSCAF) algorithm is introduced to reduce the mean absolute error of Time Difference of Arrival (TDOA) to 1.14 ns through coarse and fine search mechanisms combined with frequency offset compensation, thereby enhancing the accuracy of time delay estimation. Second, an improved version of the Beluga Whale Optimization (IBWO) algorithm is employed to address the nonlinear localization problem by integrating Halton sequence initialization, adaptive Levy flight, and a hybrid particle swarm optimization strategy. Simulation results demonstrate that the IBWO algorithm achieves a positioning error of 1.39 m at a signal-to-noise ratio of 20 dB, outperforming the Chan-Taylor algorithm, which yields a positioning error of 2.29 m.