摘要： 提出了一种新的GML数据流压缩算法。研究了GML结构与数据动态分离并动态压缩，采用多线程技术进行并行传输，再在接收端动态解压并动态合并的过程。实验结果表明，该方法对大幅提高GML数据传输效率具有实际意义。

关键词: GML, 数据流, 压缩, 多线程

Abstract: This paper proposes a kind of new GML dataflow and compressed algorithm, studies GML structure and data separate dynamically and compress dynamically, multi-thread technology is adopted in order to parallel transmit, and then receive the data and solve the course that depresses dynamically and amalgamates dynamically. The experimental result indicates that this method has actual meanings in improving GML data transmission efficiency by a wide margin.

Key words: GML, Dataflow, Compression, Multi-thread