A multi-resolution estimation method for interference spectrum separation from a one-dimensional broadband acoustic intensity is presented in this letter. The key of the proposed method is the recursion reconstruction and singular value decomposition of a $2\times L$ Hankel matrix, which is constructed by a broadband intensity data sequence of length L + 1. Numerical simulation results show that this method can be used to achieve the zero-phase-shift estimation of each interference spectrum without any prior environmental information.

中文翻译：

---

频域中每对模式干扰频谱的多分辨率估计

这封信提出了一种用于从一维宽带声强中分离干扰频谱的多分辨率估计方法。该方法的关键是递归重构和奇异值分解。$2\times \mathrm{大号}$汉克尔矩阵由长度为L  + 1的宽带强度数据序列构成。数值仿真结果表明，该方法可用于实现每个干扰频谱的零相移估计，而无需任何先验环境信息。