Near-Optimal Rate Adaptation Based on Spinal Codes for Incremental Redundancy Truncated HARQ

This paper proposes a rate adaptation strategy based on Spinal codes, a novel rateless codes, in truncated incremental redundancy hybrid ARQ (HARQ) transmission. The information transmission between Satellite or deep space probes and ground receivers is considered. Too many retransmissions, which are caused by error frame, and long distance lead to considerable delay, up to several seconds. The delay is unacceptable. Especially when rigid delay requirements are imposed on communication system, retransmissions will impair communication quality significantly. Besides, channel state is time-varing because of atmospheric scintillation and rain attenuation. Only once transmission with lots of redundancy has a severe effect on spectral efficiency although it can decrease delay dramatically. How to maximize spectral efficiency under constraint on the frame error rate and transmission delay is key of this paper. We construct spectral efficiency optimal problem and get near-optimal rate distribution during transmission attempts. Simulations are implemented for proposed scheme at most twice transmission attempts with accurate and inaccurate SNR respectively, and the results show a significant improvement in spectral efficiency compared with conventional ACM-HARQ scheme.