一种改进的稀疏近似最小方差DOA估计算法研究

Direction-of-arrival estimation of a modified sparse asymptotic minimum variance approach

针对稀疏信号的超分辨方位估计问题,提出一种可变因子的稀疏近似最小方差算法(α-Sparse Asymptotic Minimum Variance,简记为SAMV-α)。该算法利用一个折衷参数进行最大似然估计值和稀疏性能的折衷处理,在迭代过程中改变稀疏近似最小方差算法(Sparse Asymptotic Minimum Variance,SAMV)的指数因子,得到强稀疏性能和超低旁瓣的方位谱图,实现邻近目标的超分辨方位估计和相干处理性能,且无需预估角度和信源数目等先验信息,并且折衷参数的取值为0到1之间,取值区间明确,避免了稀疏信号处理算法中正则因子选取困难的弊端。计算机仿真表明SAMV-α算法方位估计性能明显优于波束扫描类算法和子空间类算法,与同类型稀疏信号处理类算法相比仍具有较高的方位估计精度,同时对于邻近声源分辨能力,SAMV-α算法较SAMV-1算法性能提高约3dB。海上试验数据处理给出了分辨率更高的方位时间历程(Bering-Time Recording,BTR)图,有效验证了SAMV-α算法的性能。

A Direction-of-Arrival （DOA） estimation algorithm named the power factor variable Sparse Asymptotic Minimum Variance （SAMV-α） is proposed in this paper. The super-resolution DOA estimation, ultra low side lobe and coherent processing performance of SAMV-α algorithm are able to be obtained after altering the power factor of the algorithm in each iteration by means of a compromise parameter which is used to compromise the maximum likelihood estimation and the sparse performance of directional spectrum. Moreover, there are no DOA pre-estimates of the incident signals and the number of sources to be required in this algorithm. And the value of compromise parameter is limited between 0 and 1 which is more explicit than the regular parameter applied in sparse signal processing algorithms. Computer simulations indicate that DOA estimation performance of SAMV-α algorithm surpasses the kind of beam scanning algorithms and subspace algorithms. Comparing with the same type of sparse signal processing algorithms, SAMV-α algorithm also reaches a better bering estimation accuracy of incident signals. Meanwhile, the performance of SAMV-α algorithm is improved by about 3 dB compared with SAMV-1 algorithm in terms of the resolving power of adjacent sound sources. The performance of DOA estimation of SAMV-α algorithm is also verified by the sea experiment results, which can provide a clearer Bering-Time Recording map.