可适应稀疏度变化的非均匀范数约束水声信道估计算法

对于具有典型时频双重扩展特性的水声信道，利用其稀疏分布特性在估计算法中引入范数约束可提高信道估计性能。但当水声信道多径稀疏度变化时，经典的l₀或l₁范数约束由于缺乏对不同稀疏模式的适应性,将导致性能下降。通过引入非均匀范数约束自适应算法并对其进行收敛性分析，针对水声信道稀疏度变化利用该算法通过非均匀范数的形式提高适应性。不同接收深度水声信道的仿真及海上实验结果表明,该算法相对经典的l₀或l₁范数约束算法有较明显的性能改善。

Non-uniform Norm Constraint Estimation Algorithm for Underwater Acoustic Channels at the Presence of Varying Sparsity

For the typical underwater acoustic channels with time-frequency double extension characteristics, the channel estimation performance can be improved by introducing a norm constraint into the channel estimation algorithm based on the sparse distribution feature of the channels. However, at the presence of varying multipath structure caused by change of depth or velocity gradient, the classic l₀ or l₁ norm constraint methods are subject to performance degradation due to lack of adaptability to sparsity. A previously derived non-uniform norm constraint LMS (NNCLMS) algorithm is introduced, and then a convergence analysis is made on it. In the form of non-uniform norm, the NNCLMS algorithm is used to accommodate the different sparsities caused by different multipath structures. Numerical simulation and sea experimental results show that the estimation performance of the proposed method is superior to that of the classic l₀ or l₁ norm constraint algorithm.