一种基于重要抽样的麦克风阵列时延估计

Time delay estimation of microphone array based on important sampling

麦克风阵列信号处理已经广泛引用于语音处理、通信等领域。时延的精确估计决定了麦克风阵列声源定位系统的性能。将重要抽样的方法应用到麦克风阵列的时延估计中,可以保证时延估计精确性的前提之下,有效地回避了网格搜索和迭代计算,使结果不依赖于初始值,并且降低了计算的复杂度。仿真实验结果表明,重要抽样时延估计方法在计算量降低一个数量级时,仍保持着与ML网格搜索法相近的性能,更适合麦克风阵列时延的实时估计。

Microphone array signal processing has been widely used in the fields of speech processing and communication.The performance of microphone array sound source localization system is determined by the accurate estimation of time delay. If the important sampling method is applied to the time delay estimation of microphone array,the grid search and iterative calculation can be effectively avoided in the precondition of guaranteeing the accuracy of time delay estimation,so that the results do not depend on the initial value,and the computational complexity is reduced. The simulation results show that the time delay estimation method based on the important sampling keeps the performance similar to ML grid search method while the calculated amount is reduced by an order of magnitude,and is more suitable for the real-time estimation for the time delay of the microphone array.