URLLC-eMBB hierarchical network slicing for Internet of Vehicles: An AoI-sensitive approach

Abstract

Network slicing has been considered as a promising candidate to meet the heterogeneous Quality of Service (QoS) requirements of various vehicular applications. In this paper, network slicing is utilized on the Road Side Unit (RSU) side for the heterogeneous data freshness requirements in Internet of Vehicles (IoV). The services in enhanced Mobile BroadBand (eMBB) slice are usually bandwidth-consuming, thus the performance can be characterized by throughput. Unlike eMBB slice, the services in Ultra-Reliable Low-Latency Communications (URLLC) slice are usually time-critical, calling for fresh content delivery within short delay, in this case, the trade-off can be made between Age of Information (AoI) and delay in URLLC slice. The resource limitation and coupling effect at both the inter-slice and the intra-slice level have brought challenges to the resource allocation of vehicular network slice. To this end, this paper models the inter- and intra-slice resource allocation problem as a Stackelberg game, in which the RSU acts as the leader and the two slices act as followers. To minimize the weighted sum of AoI and delay in URLLC slice while ensuring the throughput of eMBB slice, this paper proposes a multi-win Stackelberg game resource allocation (MWRA) algorithm. The proposed algorithm can achieve Stackelberg equilibrium (SE) and the SE point is taken as the optimal resource allocation strategy of inter- and intra-slice. Simulation shows the trade-off of inter- and intra-slice resource allocation, and further reveals the interaction of performance between different slices. Numerical results show that compared with the traditional hard slicing, the proposed algorithm can improve the AoI and delay performance of URLLC slice by about 30% under the premise of ensuring the throughput of eMBB slicing.