中文摘要：

研究了面向声阵列网络的白定位问题，提出了三种基于移动信标和波达方向（DOA）的最大似然白定位机制：基本最大似然估计（BML）、具有声信号传播延时补偿功能的最大似然估计（PCML）和基于二次计算的最大似然估计（RML）．研究了声信号传播延时效应，并采用PCML和RML对此进行了补偿．详细分析了信标移动速度、DOA误差和信标广播间隔等关键因素对以上三种节点自定位方法的精度和计算复杂度等方面的影响．仿真和实验数据表明，这三种方法具有不同的计算复杂度，可实现不同精度的位置和声阵列朝向信息估计．

英文摘要：

Self-localization issue for acoustic array network is studied, and three maximum likelihood self-localization mechanisms are proposed based on mobile beacon and DOA （direction of arrival）： BML （Basic Maximum Likelihood esti- mation）, PCML （propagation compensated ML） and RML （recalculation ML）. The impact of propagation delay is studied, and PCML and RML are adopted to compensate the impact. Factors impacting the self-localization precision and the com- putational complexity of our proposed self-localization mechanisms are discussed in details, such as velocity of the beacon, DOA error and beacon interval. Simulation and experiment results show that location and orientation information of acoustic array node are well estimated through the three mechanisms with different positioning accuracies and computational com- plexities.