Multi-User Sparse Vector Coding for eXtreme Ultra-Reliable Low-Latency Communication in Beyond 5G

Abstract

Short A short packet transmission scheme, such as Sparse Vector Coding (SVC), is a primary candidate for achieving ultra-low latency and high-reliability communication (URLLC). This paper proposes a spectral-efficient multi-user SVC (MU-SVC) scheme for achieving next-generation URLLC or eXtreme URLLC (xURLLC) in beyond 5G (B5G) communications. The key idea is to transmit multiple user information within a single sparse vector where the users are segregated into far users (FU) and near users (NU) depending on the distance from the base station. The classification into FU and NU paves way to optimize resource allocation, user fairness, manage interference, ensure reliable communication and quality of service requirements. Firstly, the FU binary data is converted into a sparse vector and secondly, the NU data is modulated and embedded into the non-zero positions of the sparse vector to form an MU-SVC. On transmission, the FU data is obtained through sparse demapping, while the NU adopts symbol detection techniques like the maximum likelihood detector. A new performance metric, called position error rate (PoER), is introduced to study the performance of the FU since it is based on the correct identification of the non-zero positions. Theoretical analyses of PoER and symbol error rate (SER) were carried out for FU and NU, respectively and the results are also validated through Monte-Carlo simulations. Further, the bit error rate, complexity, spectral and latency analyses are performed for MU-SVC and compared with the SVC and enhanced SVC schemes. The simulation results demonstrate an improved spectral efficiency and low latency with high reliability for the proposed MU-SVC scheme, thus, achieving xURLLC with reduced complexity in the multi-user scenario for B5G.