NET2 – Virtualisation, Cloud, and Convergence
Wednesday, 17 June 2020, 12:15-14:30 CEST, Recommended re-viewing, https://www.youtube.com/playlist?list=PLjQu6nB1DfNCUc5Cjz1QsQk6SBw8OTl0E
Wednesday, 17 June 2020, 12:15-16:00 CEST, Non-Live interaction (Chat), link sent only to Registered people
Hadi Razzaghi Kouchaksaraei (Paderborn University, Germany); Ashwin Prasad Shivarpatna Venkatesh (University of Paderborn, Germany); Amay Churi (Universität Paderborn, Germany); Marvin Illian and Holger Karl (Paderborn University, Germany)
To reduce the cost and enhance the flexibility of network functions, Virtual Network Functions (VNFs) are deployed on commercially-of-the-shelf (COTS) resources instead of specialized hardware. This, however, downgrades the performance of network functions. To mitigate this issue, leveraging acceleration hardware such as GPU and FPGA has been suggested. This is because accelerators provide better parallelization and pipelining for compute- and network-intensive VNFs. However, studies show that using accelerators are not beneficial for all cases. One of the reason is that accelerators are expensive resources, possibly increasing the service cost. Consequently, the use of accelerators needs to be decided for each situation individually. To address that, in this study, we have extended Pishahang (i.e., a MANO framework) to support dynamic usage of accelerators in the NFV environment. Using the extended Pishahang, then, we analysed the management overhead due to such dynamic usage of accelerators.
Venkatarami Reddy Chintapalli (IIT Hyderabad, India); Koteswararao Kondepu (Sculoa Superiore Sant’Anna, Italy); Andrea Sgambelluri (Scuola Superiore Sant’Anna Pisa, Italy); Antony Franklin A (Indian Institute of Technology Hyderabad, India); Bheemarjuna Reddy Tamma (IIT Hyderabad, India); Piero Castoldi and Luca Valcarenghi (Scuola Superiore Sant’Anna, Italy)
In the Zero-touch network and service management (ZSM) architecture devised by ETSI making predictions on the observed data is among the functions provided by the Analytics block of the control loop cycle. Prediction performance depends on several parameters, such as the utilized computational resources, the leveraged prediction techniques, the deployment location of the prediction tools with respect to the data.
This paper proposes a Hybrid Forecast Framework (HFF) running both at the network edge and in the cloud to provide forecasting with the performance required by the control loop cycle. Forecasting at the edge might shorten the control loop cycle if resources shall be made available locally where data are collected. However, in general, edge computational resources are less abundant than the cloud ones, thus causing longer time to perform the prediction. On the opposite, forecasting in the cloud might require more time for the data to reach the utilized tools but more computational resources could be exploited. The HFF is based on utilizing traditional time series analysis prediction algorithms to minimize the utilized resources and energy at the edge while it exploits AI/ML tools to make predictions in the cloud.
Results shows that for short lead time (i.e., the time, in the future, at which the status of the considered parameter is predicted) edge-based prediction exploiting time series analysis provide better accuracy, requires less resources, energy, and time than cloud-based prediction. However, if the lead time is long, cloud-based prediction exploiting Artificial Intelligence/Machine Learning (AI/ML) provides better accuracy. Thus, if the lead time is long, it is preferable because the long lead time compensates for the higher time for prediction due, mainly, to data transfer.
Asad Ali, Ying-Dar Lin and Chi-Yu Li (National Chiao Tung University, Taiwan); Yuan-Cheng Lai (Information Management, NTUST, Taiwan)
Multi-access Edge Computing (MEC) is a key technology for supporting low latency applications close to the end user. Users can access application servers in MEC instead of routing to the internet by passing through a core cellular network. Few security challenges arise as the traffic does not traverse through the core network, and these can be solved by providing authentication services in the MEC. However, authentication and application mobility issues arise in the case of multiple MECs where a user is mobile and needs continuous service from application servers, without needing to establish a new session and providing authentication information repeatedly to every new MEC it connects with. We propose two solutions, namely TC3A (Token-based Cookie transfer & 3rd-party Authentication) and TS3A (Token-based State transfer & 3rd-party Authentication) for resolution of authentication and application mobility issues while achieving low latency. Experiments were conducted on a testbed which emulated the handover of a user between two MECs. The results show that TC3A and TS3A provide authentication to users at reduced latency by 4.6% and 25%, respectively, as compared to simple login method, and most importantly application service continuity.
Nikos Psaromanolakis, Athina Ropodi and Pavlos Fragkogiannis (Incelligent PC, Greece); Kostas Tsagkaris (Incelligent, Greece); Luiz Anet Neto, Anas El Ankouri and Minqi Wang (Orange Labs, France); Gael Simon (Orange, France); Philippe Chanclou (Orange Labs, France)
Software Defined Networking (SDN) has come up as a promising technology for making the communication networks more flexible and dynamically (re)configurable. With the advent of fifth generation (5G) of mobile networks, the aforementioned flexibility became more urgent and critical than ever. SDN offers the decoupling of the network’s control plane from the data plane. As a result, the infrastructure layer can be abstracted from the application and network services and therefore the network control can be done completely through a centralized component/ orchestrator. In this work, a detailed practical deployment of an SDN controller for a converged Fiber-Wireless (Fi-Wi) 5G network is presented, followed by the SDN architecture and functional components. Lastly, an experimental trial using an SDN controller directly connected to Fi-Wi lab equipment is described, in which the physical layer parameters are managed and monitored successfully, proving the applicability of SDN in a Fi-Wi network.