System-Level Study of Data Duplication Enhancements for 5G Downlink URLLC

摘要

Data duplication is studied as a fundamental enabler for ultra-reliable low-latency communication (URLLC) in fifth-generation cellular systems. It entails the simultaneous usage of multiple radio links delivering redundant data between a terminal and the network to boost the transmission reliability. However, the improved reliability comes at a cost of reduced spectral efficiency, since the transmission of multiple instances of the data message on different links occupies more radio resources as compared to sending only one instance using a single link. It is therefore crucial to improve the performance of data duplication schemes, with the aim of reducing the radio resource consumption without degrading the reliability gain provided by this transmission paradigm. In this paper, we propose several methods to increase the downlink URLLC capacity supported by data duplication in fifth-generation cellular networks based on the New Radio standard. A single-user analytical model is derived to evaluate a combination of the proposed enhancements. The most promising solution, namely selective data duplication upon failure which entails a massive reduction of the overall number of duplicate transmissions, is finally evaluated by means of extensive multi-user system-level simulation campaigns. The simulation results with background mobile broadband traffic show that, in the investigated scenario, the proposed solution with 4 Mbps offered URLLC traffic outperforms the baseline approach for data duplication with 1 Mbps offered URLLC traffic, thus increasing the amount of URLLC user equipments that can be effectively sustained by the network.