大型数据的编码传输技术研究

Research on Encoding Transmission of Big Data File

提出利用信息学编码理论中的线性纠错码LECC来改善当前计算机网络中进行大型数据传输时存在速度慢、可靠性低等问题。线性纠错码方法对数据进行分块冗余编码,在有损信道(如Internet)上传输编码块。接收端只要接收到足够数量的编码包,就可解码出初始数据信息,无需反馈信道,减少包应答及丢失包重传的时间。实验结果表明,LECC编码传输平均只要接收到比源数据包多4%的编码包即可完成解码。对于大型文件传输,编解码及冗余包的传输所增加的负载比传统差错控制小,有效地提高了信道的可靠性及传输效率。

This paper uses linear error correction code（LECC） of the informatics encoding theoretics to solve the problems of slowness and low reliability on transmission of big data file. LECC algorithm partitions the data into blocks and encodes it redundantly, then the encoded packages are transmitted in the erasure channel （Internet for example）. If enough encoded packages are received, the receiver can decode the original data, regardless of which package is lost. It saves the time for package acknowledging and retransmitting. In the experiment, the receiver can finish decoding as long as it gets 4 percent encoded packages more on average in LECC transmission, the cost of encoding, decoding and the increased packages transmitting is less than the ARQ error control＇s in transmission of big data file. LECC algorithm improves the reliability of channel and the efficiency of data transmission.