一种快速自适应最优小波包基搜索算法  

Fast Adaptive Search Algorithm for Optimal Wavelet Packet Basis

作　　者：李姣军[1] 李恒[1] 苏理云[2] 李刚[1] 

机构地区：[1]重庆理工大学电子信息与自动化学院,重庆400050 [2]重庆理工大学数理学院,重庆400050

出　　处：《现代电子技术》2011年第11期72-75,共4页Modern Electronics Technique

基　　金：重庆市自然科学基金项目(CSTC2010BB2310;CSTC2009BB2420);重庆市教委科技研究项目(KJ100810)

摘　　要：在小波包调制系统中,小波包基的选取对于系统性能有着显著的影响。针对目前BBS算法与BFA算法的缺点,分析了小波包基的构成原则与选取方法,提出了一种基于根节点逐级向下按二进分支路分解,利用度量函数决定父节点与子节点取舍的最优基搜索算法,并解决了约束条件下的最优基输出问题。最后将该算法在4径Rayleigh信道模型下进行了测试,仿真结果表明：在相同条件下,该算法比BBS算法快15%~20%,比BFA算法快10~15倍,且输出的最优基比全分解结构的ICI与ISI联合功率降低约20%。In wavelet packet modulation system,wavelet packet based selection has a significant impact on the system performance.To overcome the defects of best basis selection（BBS） algorithm and brutal force algorithm（BFA）,the composition and selection principles of the wavelet packet basis are analyzed,and a novel search algorithm for wavelet packet basis is proposed.This algorithm decomposes the whole wavelet packet library from top to bottom by a binary branch,uses metric function to choose the parent node or child nodes,and solves the competition between same-level nodes.The algorithm was tested in 4-path Rayleigh fading channel model.The simulation results indicate that under the same conditions,the proposed algorithm is 15%～20% faster than BBS algorithm,10～15 times faster than the BFA algorithm,and its optimal basis structure is 20% less than ICI and ISI integrated power of full decomposition structure.

关 键 词：小波包调制 最优小波包基 度量函数 BBS算法 BFA算法 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]