End-to-end URLLC slicing based on packet duplication in 5G optical transport networks

Abstract

With the stringent requirements of high reliability and low latency, ultra-reliable low latency communications (URLLCs) are one of three typical services in the fifth generation of mobile network (5G) paradigm. The packet duplication (PD) technique has been proposed as a promising solution to satisfy the extreme requirements of URLLC services. However, the benefit of PD comes at the cost of doubling resource usage. This paper investigates how to efficiently perform end-to-end URLLC slicing based on PD in 5G optical transport networks. Two heuristic algorithms are proposed, including PD-based slicing with wavelength isolation (PDS-WI) and PD-based slicing with link isolation (PDS-LI). The target of this paper is to maximize the number of accommodated URLLC slices by optimizing the allocation of heterogenous resources. The time complexity of these two algorithms is analyzed in detail. With dynamic slice requests, simulations are performed to evaluate and compare their performance in terms of slice blocking probability and resource usage. Meanwhile, two different network scenarios are considered in simulation. Simulation results indicate that PDS-WI outperforms PDS-LI by accommodating more URLLC slices.