Adaptive Sparse VectorCoding for Short Packet URLLC
Sparse Vector Coding(SVC)is a promising transmission technology for short packet ultra-reliable and low-latency communications(URLLC).However,the current SVC based on combination sparse transformation does not have ideal transmission performance when encoding efficiency is high.In order to improve this situation,an adaptive sparse transformation-based SVC(A1D-SVC)scheme is proposed by utilizing the inherent correspondence between binary and decimal conversions of index bit streams.The proposed AD-SVC scheme selects a sparse transformation model by counting the difference between the current and the next index bit streams,and combines the idea of segmented sparse transformation with construction transfer when a small values of index bit stream appears to construct the sparse vectors.The proposed a-daptive sparse transformation limits the length of sparse vectors within a certain range.Simula-tion results show that when the encoding efficiency is in a constraint improved,the proposed a-daptive sparse transformation can construct shorter sparse vectors,which enables AD-SVC to a-chieve better block error rate(BLER)performance and transmission delay performance than ex-isting SVC schemes.
Ultra-reliable and low-latency communicationsshort packet communicationsSparse vector codingSparse transformation
NETL
NSTL
万方数据
