Caching has emerged as an efficient technique to reduce delivery latency and network congestion. The focus of this paper is on content delivery in the caching-based wireless systems. In view of the fact that in such systems, users store fractional of popular contents during off-peak hours, we propose a modified version of the Luby Transform (MLT) encoding for the delivery phase to take the advantage of channel coding. The novelty of our proposed MLT encoding lies in employing the user’s partial information to create an appropriate degree distribution. To evaluate our proposed scheme, this study considers a realistic channel model inspired by the Gilbert-Elliott channel including the erasure mode and a binary symmetric channel model. The effectiveness of our MLT encoding is analyzed by the average degree and the number of encoding symbol. A comprehensive simulation evaluation is performed which shows that MLT encoding outperforms conventional LT coding in terms of the required number of encoding symbols and demonstrates a better performance on the recovered symbols. In particular, the MLT encoding, compared to the conventional LT encoding, reduces the amount of required symbols to recover the original file for at least 20%, when the user stored more than half a file.

缓存已经成为一种有效的技术，可以减少传递延迟和网络拥塞。本文的重点是基于缓存的无线系统中的内容交付。鉴于在这样的系统中，用户会在非高峰时段存储一部分受欢迎的内容，因此我们建议在交付阶段使用Luby变换（MLT）编码的修改版本，以利用信道编码的优势。我们提出的MLT编码的新颖性在于利用用户的部分信息来创建适当的度数分布。为了评估我们提出的方案，本研究考虑了一个由Gilbert-Elliott信道启发的现实信道模型，其中包括擦除模式和二进制对称信道模型。我们的MLT编码的有效性通过平均程度和编码符号数来分析。进行了全面的仿真评估，该评估表明，就所需的编码符号数量而言，MLT编码的性能优于传统的LT编码，并且在恢复的符号上表现出更好的性能。特别地，与传统的LT编码相比，当用户存储了一半以上的文件时，与传统的LT编码相比，MLT编码减少了恢复原始文件所需的符号量至少20％。