论文部分内容阅读
讨论了多维 DCT和多维 DWT的并行行列算法和并行多项式变换算法 ,并用 Log P模型对算法进行了分析。在仔细分析一维和二维离散小波变换与小波包变换计算结构的基础上 ,提出了它们的并行算法。算法只需相对较少的通信时间 ;适合大规模并行机 ( MPP)和工作站或微机机群系统 ;方法也适合信号处理中的各种塔式分解过程。用 Fortran语言和 PVM编制了算法的程序。在机群系统和大规模并行机上的实验表明 ,算法是正确的且具有较高的加速比。
The parallel parallel computing algorithm and parallel polynomial computing algorithm of multidimensional DCT and multidimensional DWT are discussed, and the algorithm is analyzed by Log P model. On the basis of careful analysis of one-dimensional and two-dimensional discrete wavelet transform and wavelet packet transform computing structure, their parallel algorithms are proposed. The algorithm requires relatively little communication time; it is suitable for large-scale parallel machines (MPPs) and workstations or microcomputer clusters; and the method is also suitable for various tower decomposition processes in signal processing. The algorithm was compiled using Fortran language and PVM. Experiments on cluster systems and large-scale parallel machines show that the algorithm is correct and has a high speedup.