Rate-Splitting Sparse Vector Code for Next-Generation URLLC Systems

In this letter, we propose a rate splitting sparse vector coding (RS-SVC) scheme for the next generation ultra-reliable and low-latency communication (xURLLC). The key idea is to segregate the users' data into common and private parts. A part of users' information is taken from each user to form the common data (CD) and the rest of the information are considered as private data (PD). Subsequently, the CD is sparse mapped into a sparse vector and each user's PD is symbol modulated and embedded into the non-zero positions of the sparse vector forming the RS-SVC. This RS-SVC is pseudo randomly spread over a code book and transmitted via a wireless channel. Upon reception, the RS-SVC undergoes a sparse demapping process to pinpoint the non-zero positions representing the CD, concurrently demodulating the embedded symbols will extract the PD. Our numerical analyses and Monte-Carlo simulations substantiate the RS-SVC scheme's superiority over traditional SVC and its enhanced variant, demonstrating significant improvements in block error rate, user fairness and spectral efficiency.