Research Of Cooperative HARQ Technologies Based On Cognitive Radio Ad-hoc Network

With the rapid growth of business in wireless communication network, the existing spectrumresources become more and more scarce. Cognitive radio technology can use idle spectrumeffectively and will be applid wildly in wireless communication of the future. Ad-hoc network alsohas many advantages, which is a kind of distributed, multi-hop and self-organized mobile network.Cognitive radio Ad-hoc network, which combines cognitive radio technology and Ad-hoc network,is a distributed cognitive network. It not only uses spectrum effectively, but also is suitable fortemporary and emergency communication.This thesis takes cooperative HARQ technology in cognitive radio Ad-hoc network as theresearch object, and discusses key issues, such as retransmission protocols, channel coding anddecoding, HARQ combining decoding. Firstly, this thesis introduces the principle of cooperativediversity and types of HARQ, analyzes relay strategies and selection of relay. Also it makes asummary of HARQ retransmission protocol, which ensures the system work.Then, the coding/decoding algorithm of LDPC is introduced due to its strong correcting ability,and Log-BP algorithm is completed in c program. For the calculation of log-likelihood ratio (llr) inLog-BP, two methods which based on Guass channel and Euclidean distance are analyzed,afterwards a posterior probability calculation model and improved algorithm are proposed. Thesimulation results show that the improved algorithm has better performance and lower bit error rate.Particularly, construction of complementary punctured LDPC codes which applied in PartialIncremental Redundancy HARQ is studied. A pair of complementary punctured LDPC codes withrate2/3are constructed, which has similar performance with special rate2/3code in multipathchannel, and the result confirms the validity of punctured method.Finally, to improve decoding performance and applicability of HARQ combining together,through analyzing maximal ratio combining and bit-level combining, a algorithm with weighted llris proposed, and complexity are stydied. The proposed method combines the weighted bit llr ofeach received signal according to the channel impulse power in frequency domain, and can achieveperformance near maximal ratio combining. Furthermore, it is feasible for variable HARQ systems,while maximal ratio combining can only be used in Chase HARQ.