RIA: TWO NEW CODING IDEAS

这个项目正在探索两个简单而强大的编码思想。第一种涉及从外码到内码的分段功率转移，并且可以在格子系统和传统的卷积编码系统上使用。这一新技术不仅改善了级联编码系统在其通常的操作机制下的性能，即非常低的译码误码率，而且还将它们的使用扩展到对未压缩图像或语音的较不严格的要求，其中译码复杂性和短延迟是重要的。基于卫星的移动电话或数据服务可以立即从这些技术中受益。第二种思想是一种在里德-所罗门码字集合的成员之间共享奇偶校验的方法。其效果是减少了所需的冗余量，即提高了码率。对这种技术的最大需求来自这样的系统，其中符号大小以某种方式由信道规定，例如，具有256进制P.P.M.的8比特里德-所罗门码。在光子计数光学通道上。即使符号大小不受限制，即可以使用更大的里德-索罗门码，对于相同的性能增益，在新方法的每信息比特的运算中测量的编码器和解码器的复杂度似乎都较低。

This project is exploring two simple and powerful coding ideas. The first involves staged power transfer from an outer to an inner code and can find use on both trellis and conventional convolutionally coded systems. This new technique not only improves the performance of concatenated coding systems in their usual operating regime, i.e., very low decoded error rates, but also extends their use to the less stringent requirements of uncompressed images or speech where decoder complexity and short delays are important. Satellite based mobile telephone or data service could benefit immediately from these techniques. The second idea is a method of sharing parity checks among the members of a set of Reed-Solomon codewords. The effect is a decrease in the amount of reduncancy required, i.e., an increase in the rate of the code. The greatest need for this technique comes from systems where the symbol size is in some way dictated by the channel, e.g., an 8 bit Reed Solomon code with 256-ary p.p.m. on a photon counting optical channel. Even if the symbol size is not constrained, i.e., larger Reed Solomon codes could be used, both the encoder and decoder complexity measured in operations per information bit of the new method appear to be lower for the same performance gain.