| Summary: | 碩士 === 臺灣大學 === 電信工程學研究所 === 98 === In a random gift channel, in addition to the distorted channel signals, the destination
can also receive some information regarding the sent message, which we refer
as gifts. The information can be randomly selected from the source bits or encoding
symbols, and we call those two types of gifts as source bit gift(SBG) and encoding
symbol gift(ESG). Among this thesis, first, we prove the channel capacity of random
gift channel with SBG and ESG, compare the performance of those two types of gift,
and give the simulation result of turbo code and LDPC code to see whether they
can perfectly fit the gift without changing the code structure. Next, since systematic
codes will decrease the effect of gifts, we also provide the capacity of random
gift channel restrict on systematic code to estimate the effect. Finally, we give some
applications that can use the concept of random gift channel and analysis the performance,
including hybrid-ARQ, soft handover and finding the maximum likelihood
bound for turbo code and LDPC code.
|
|---|
