Resource Allocation with Admission Control for GBR and Delay QoS in 5G Network Slices

Buyakar, Tulja Vamshi Kiran and Agarwal, Harsh and Tamma, Bheemarjuna Reddy and et al, . (2020) Resource Allocation with Admission Control for GBR and Delay QoS in 5G Network Slices. In: 12th International Conference on COMmunication Systems & NETworkS, 7 - 11 January 2020, Bengaluru, India. (Submitted)

Full text not available from this repository. (Request a copy)

Abstract

Network slicing is an integral part of 5G whichsupports next-generation wireless applications over a sharednetwork infrastructure. It paves the way to leverage the full po-tential of 5G by increasing the efficiencies through differentiationand faster time-to-market. In this work, we propose a MobileVirtual Network Operator (MVNO) Slice Resource AllocationArchitecture (MSRAA) for supporting different network slices inthe 5G data plane. MSRAA supports QoS parameters, includingGuaranteed Bit Rate (GBR) and Maximum Delay Budget. Usinglong short-term memory (LSTM) neural networks, we predictnetwork slices bandwidth requirements for efficiently allocatingthe resources. To reduce revenue loss to the network operatorsdue to forecasting errors, in the proposed Bandwidth AdmissionControl (BAC) algorithm reallocates resources from lower prior-ity slices (ex: guaranteed service users) to higher priority slices(ex: best-effort users). Using Mondrain Random Forests in ourDelay Admission Control (DAC) algorithm, we predict the end-to-end delay and admit flows into slices that can satisfy delayrequirements.We implement MSRAA on our developed 5G Core testbedand evaluate User Service Request (USR) acceptances and doa complete cost-benefit analysis of our architecture. We showthat for eMBB-GBR and eMBB-Non-GBR slices, our algorithmis showing a significant reduction in costs and an increase inprofits.

[error in script]

IITH Creators: