中文摘要：

本文关注的是多路信号之间时延差异的联合估计问题。不同于传统的自适应时延估计算法，本文以合成信号作为自适应时延估计的参考信号，给出了基于信号合成的联合自适应时延估计算法。同时本文推导和仿真了该算法时延估计的均值、学习曲线及方差特性。数据仿真结果与性能分析一致，均显示本文提出的基于合成的多路信号自适应时延估计为渐进无偏的时延估计。在不明显增加计算量的条件下，当算法收敛时，联合时延估计算法的方差显著低于传统的两路信号之间自适应时延估计算法方差。

英文摘要：

This paper focuses on the joint adaptive time delay problem of multiple received signals. Different to the classi- cal adaptive time delay estimation algorithms, the combined signal performs the newer common reference signal in this pa- per. Base on this, using the combined signal as the reference signal for joint estimation of multiple signals in time delay estimate, a new joint explicit adaptation of time delay algorithm is derived. The joint explicit adaptive delay estimate and its respective performance surface, learning characteristics and delay variance are also given in the paper. Numeral simulation results are consistent with the performance analysis. The numeral simulation results show that the proposed joint explicit a- daptive process provides an asymptotic unbiased delay estimate. The proposed algorithm efficiently reduces the time delay estimate variance with only increasing a little computational cost compare to the conventional adaptive time delay estimation algorithm, which adaptive estimates time delay between two signals.