Nikita Jalodia, a PhD Researcher with the Emerging Networks Lab research unit in TSSG at Waterford Institute of Technology, recently had her work published in the IEEE Conference on Network Function Virtualization and Software Defined Networks held in Dallas, Texas, USA from 12-14 November, 2019. The publication is titled ‘Deep Reinforcement Learning for Topology-Aware Resource Prediction in NFV Environments’, and is an output of her ongoing research supervised by Dr. Alan Davy, and funded by SFI funded CONNECT Research Centre for Future Networks and Communications.
Notably, Nikita was among the three recipients to be awarded a travel grant of $600 sponsored by the IEEE Communications Society to travel and present at the conference in USA. The grant decision is made based on merit, awarded to high quality papers and demos from among the pool of applicants.
At the conference, she presented her ongoing work during the Fast Track Session, and discussed about the application of Deep Reinforcement Learning and Graph Neural Networks in the Network Function Virtualization application area.
Pictured is Nikita presenting her work at the IEEE Conference in Dallas
Network Functions Virtualization (NFV) and Software Defined Networks (SDN) are an accepted evolution in all areas of network concepts and technologies. They are expected to further radically and dramatically transform not only telecommunication networks, enterprise and data centre networks, but accelerate the introduction of smart cities/homes/cars/businesses and green infrastructures. Currently, NFV and SDN are in the transition phase from development into first trials, evaluation and early deployments too.
The IEEE NFV-SDN conference, further onto its sixth year in 2020, is an accelerator of the continuous exchange on the latest ideas, developments and results between all ecosystem partners in the academia and industry area, in the field of NFV and SDN. The conference sessions actively aim to foster discussions on new approaches as well as dedicated work on missing aspects for improvements of NFV and SDN enabling architectures, algorithms, frameworks and operation of virtualized network functions and infrastructures.
Abstract of Nikita’s Publication and Presentation in the Conference:
Network Function Virtualization (NFV) has emerged as a key paradigm in network softwarization, enabling virtualization in future generation networks. Once deployed, the Virtual Network Functions (VNFs) in an NFV application’s Service Function Chain (SFC) experience dynamic fluctuations in network traffic and requests, which necessitates dynamic scaling of resource instances. Dynamic resource management is a critical challenge in virtualised environments, specifically while balancing the trade-off between efficiency and reliability. Since provisioning of virtual infrastructures is time-consuming, this negates the Quality of Service (QoS) requirements and reliability criterion in latency-critical applications such as autonomous driving. This calls for predictive scaling decisions to balance the provisioning time sink, with a methodology that preserves the topological dependencies between the nodes in an SFC for effective resource forecasting. To address this, we propose the model for an Asynchronous Deep Reinforcement Learning (DRL) enhanced Graph Neural Networks (GNN) for topology-aware VNF resource prediction in dynamic NFV environments.