【摘要】 分析了基于四阶累积量进行阵列扩展的MUSIC_LIKE算法,研究了阵列扩展的基本原理.并通过分析MU SIC_LIKE算法对均匀线阵的阵列扩展原理,摒弃了原MUSIC_LIKE算法在均匀线阵DOA估计中的大量数据冗余,提出一种新的阵列扩展方式,该算法可以将M2×M2的四阶累积量矩阵转化为(2M-1)×(2M-1)的矩阵,有效地降低了累积量矩阵的运算量.通过和原有算法的扩展原理比较,可以看出,新的扩展阵列在虚拟阵元上没有加权,因而对阵列的扩展也更加合理.计算机仿真的结果表明,该算法与MUSIC_LIKE算法有相同的阵列扩展能力,并且对算法运算量的降低效果明显.更多还原

【Abstract】 <Abstrcat>Analysis is done on a MUSIC_LIKE algorithm increasing the virtual array elements based on the forth order cumulant and the principle of array extension. The massive redundant data can be eliminated by analyzing the array extension principle of MUSIC_LIKE algorithm to uniform linear array. A new algorithm is presented that can convert the fourth cumulant matrix of M~2×M~2 into the (2M-1)×(2M-1) matrix. The method efficiently reduces the computation amount for a cumulant matrix. Comparing the principle of the orginal algorithm shows the new extended array has no weight on the virtual elements and is more reasonable.Computer simulations show that the extension ability of this method is the same as MUSIC_LIKE and that the method reduces computation effort.更多还原