Machine Learning Based Resource Orchestration for 5G Network Slices
Authors:
- 1 December 2019
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Abstract
5G will serve heterogeneous demands in terms of data-rate, reliability, latency, and efficiency. Mobile operators shall be able to serve all of these requirements using shared network infrastructure's resources. To this end, we propose in this paper a framework for resource orchestration for 5G network slices implementing four Quality of Service pillars. Starting from traffic classification, demands are marked so that they are best served by dedicated logical virtual networks called Network Slices (NSs). To optimally serve multiple NSs over the same physical network, we then implement a new dynamic slicing approach of network resources exploiting Machine Learning (ML). Indeed, as demands change dynamically, a mere recursive optimization leading to progressive convergence towards an optimum slice is not sufficient. Consequently, we need an initial well-informed slicing decision of physical resources from a total available resource pool. Moreover, we formalize both admission control and slice scheduler modules as Knapsack problems. Using our 5G experimental prototype based on OpenAirInterface (OAI), we generate a realistic dataset for evaluating ML based approaches as well as two baselines solutions (i.e. static slicing and uninformed random slicing-decisions). Simulation results show that using regression trees as an ML based approach for both classification and prediction, outperform other alternative solutions in terms of prediction accuracy and throughput.Keywords
This publication has 12 references indexed in Scilit:
- Activity Recognition Method for Home-Based Elderly Care Service Based on Random Forest and Activity SimilarityIEEE Access, 2019
- Feature-Selection Based Data Prioritization in Mobile Traffic Prediction Using Machine Learning2018
- Throughput-Aware RRHs Clustering in Cloud Radio Access Networks2018
- On the configuration of radio resource management in a sliced RAN2018
- Optimization of the implementation of network slicing in 5G RAN2018
- What can machine learning do? Workforce implicationsScience, 2017
- Mobile traffic forecasting for maximizing 5G network slicing resource utilization2017
- Mobile Network Traffic Prediction Using MLP, MLPWD, and SVM2016
- Modeling Oscillation Behavior of Network Traffic by Nested Hidden Markov Model with Variable State-DurationIEEE Transactions on Parallel and Distributed Systems, 2012
- Network Traffic Prediction and Result Analysis Based on Seasonal ARIMA and Correlation Coefficient2010