面向水声网络可靠传输的FDR编解码算法

FDR coding and decoding algorithm for reliable transmission in underwater acoustic network

通过分析RLT码,针对RLT码译码算法中存在的缺陷,提出一种过滤式降维(FDR)算法,消除了传统译码算法在收到一定数量编码分组才开始解码的等待时间,实现了边接收边尝试解码的快速译码方式。此外,通过编码分组之间的异或运算,有效增加了度为1的编码分组的产生概率,不再仅依赖于从发送端获取度为1的编码分组,在降低传输时延的同时通过增加度为1的编码分组出现的概率从而提高译码成功率。在此基础上,提出一种与FDR译码算法相结合的优化度分布函数,适当提高度为2、度为3、度为4的编码分组的比例,从而大大增加一次降维即可得到度为1的编码分组的概率,加快了译码速度。NS3仿真结果表明,FDR算法的解码成功率普遍高于RLT码。

By analyzing the shortcomings of RLT coding and decoding algorithm,a filtering dimension reduction(FDR)algorithm was proposed,which eliminated the waiting time of the traditional decoding algorithm and achieves fast decoding.In addition,XOR operation between encoded packages effectively increased the number of one-degree encoded packages,and improved decoding probability while reducing transmission delay.An optimized degree distribution function for FDR decoding algorithm was proposed,which increased the proportion of two-degree,three-degree and four-degree encoded packages,further increased the probability of one-degree packet,so that speeded up the decoding progress.Simulation results with NS3 show that the decoding success probability of FDR algorithms higher than RLT algorithm.