Skip to Main content Skip to Navigation

Multi-connectivity and resource allocation for slices in 5G networks

Abstract : Future mobile networks envision unprecedented innovation opportunities and disruptive use cases. As a matter of fact, the 5G and beyond networks' pledge to deliver mission-critical applications mandates a versatile, scalable, efficient, and cost-effective network capable of accommodating its resource allocation to meet the services' heterogeneous requirements. To face these challenges, network slicing has emerged as one of the fundamental concepts proposed to raise the 5G mobile networks' efficiency and provide the required plasticity. The idea is to provide resources for different vertical industries by building multiple end-to-end logical networks over a shared virtualized infrastructure. Each network slice is customized to deliver a specific service and adapts its architecture and radio access technologies.Precisely, applications such as industrial automation or vehicular communications pose stringent latency and reliability requirements on cellular networks. Given that the current mobile network cannot meet these requirements, ultra-reliable low-latency communications (URLLC) embodies a vital research topic that has gathered substantial momentum from academia and industrial alliances. To reach URLLC requirements, employing multi-connectivity (MC), i.e., exploiting multiple radio links as communication paths at once, is a promising approach.Therefore, the objective of the present manuscript is to investigate dynamic scheduling techniques, exploiting redundant coverage of users, guaranteed in numerous 5G radio access network scenarios. We first review the evolution of mobile networks and discuss various considerations for network slicing architecture and its impact on resource allocation design. Then, we use tools from queuing theory to model a system in which a set of URLLC users are connected simultaneously to two base stations having the same bandwidth; we refer to this scenario as the homogenous case. We introduce suitable scheduling policies and evaluate their respective performances by assessing their reliability. Next, we extend the homogenous case's results to a more general setting where the physical interfaces manage different bandwidths, referred to as the heterogeneous case. Finally, we merge the above elements to validate the choice of resource allocation schemes considering the deployed architecture.
Complete list of metadata
Contributor : ABES STAR :  Contact
Submitted on : Wednesday, June 29, 2022 - 4:23:12 PM
Last modification on : Monday, July 4, 2022 - 8:37:53 AM


Version validated by the jury (STAR)


  • HAL Id : tel-03709139, version 1


Abdellatif Chagdali. Multi-connectivity and resource allocation for slices in 5G networks. Signal and Image processing. Université Paris-Saclay, 2022. English. ⟨NNT : 2022UPAST052⟩. ⟨tel-03709139⟩



Record views


Files downloads