research | Mariam (Marium) Kiran, Ph.D.

Links for AI/ML Research for Network Infrastructure (Wired, wireless and more)
Selected press releases
Awards
Patents
Research projects
Bio

Useful Links: AI and ML for Networks

Data repositories

CAIDA: Great resource for Internet Data.
- TCP and ML analysis for good and bad flows. Cubic, Reno and Hamilton analysis. Panorama datasets.
- Studying how TCP conflicts in same link, Exploring BBR v1, v2 and v3. Poseidon Datasets.
Sample Wireshark captures, open data set.
MAWI data: for traffic anomalies: http://www.fukuda-lab.org/mawilab/documentation.html
Internet Traffic data: https://ant.isi.edu/datasets/index.html
Security related data released by cybersecurity lab at LANL: https://csr.lanl.gov/data/cyber1/
Netflix data set for movie recommendations: https://www.kaggle.com/netflix-inc/netflix-prize-data

Selected Press

January 18, 2024: ORNL’s Quantum Initiatives Flourish with Over $10M in New Funding Awards Link
January 16, 2024: Three recent funding awards bolster ORNL’s quantum leadership Link
Nov 28, 2023: SWARM Project Release Link
Nov 2021: Career Goals for female researchers, how does one choose between academia and industry?” Link
Sept 10th 2021, HPC Wire: Leverage AI and ML for Networks for DOE Workflows Link
Jul 21 2021, SSURF: At ESnet, Mariam Kiran Wins Rising Star Award for Network Engineering Link
Feb. 26, 2021, Using AI to Direct Traffic: Building Self Learning Networks on Chameleon Link
Nov 6, 2020, Machine Learning to Add Another Dimension to ESnet's Toolbox for Predicting Data Patterns Link
Oct 19th 2020, Published Blog: One Network Dashboard to Rule them All = NetGraf
Sept 23rd 2020, Discussing AI for Networking = AI Control Plane
2017: Women in STEM -“Thinking networks” Mariam Kiran [Link]
Dec 2nd 2015, University of Bradford lecturer swaps lab coat for legislationLink
2010, Issues with Economic and Social systems modelling Mariam Kiran University of Sheffield Future Research Directions in Agent Based Modelling June 2010.Link

Abstracts, Videos & Talks

July 2023: Studying TCP Fairness in Transfers over High Throughput Links[Link]
SC23 demo abstracts [Link]
SC21 Network Research Exhibition: Publishable Abstract HECATE: Towards Self-driving Networks[Link]
HECATE:Data driven, machine learning dynamic path optimization [Youtube link]
Basics2Breakthroughs: AI for better networking [Youtube link]
Investigating sFlow for Monitoring Networks [Youtube link]
EVIAN: tool to enable intents to be deployed on network links, EPOC - Engagement Performance Operations Center [Youtube link]
At NERSC lunch time talks: Time-series Analysis of ESnet Network Traffic: Statistical and Deep Learning models, NERSC Data Seminars, [Youtube link].
iNDIRA Demo (SC16), [Youtube Talk].
Teaching big data processing with open source software, [Talk].

Awards and Recognition

Best Paper Award: MLN 2021 Best Paper AwardLink
ACM N2Women Rising Star in Communications 2021
Faculty Award: New Manager Training HaaS Berkeley 2021
Runner up for R&D 100 Award for ESnet Portal 2020
US DOE ASCR Early Career Award 2017
BradStack featured on Bradford News 2016
UK Royal Society Award for Researcher in Residence at Westminster (UK Parliament) 2015
Fully funded PhD Award 2007-2011
Fretwell-Downing Prize for Best MSc Dissertation in Engineering, University of Sheffield, 2007

Patents filed

Provisional Patent Application based on technology: Data driven, machine learning augmented dynamic path optimization, 2021
Provisional Patent Application based on technology: Reinforcement Learning for Quantum Network Control, 2021

Research Projects

SWARM(PI, Ewa Deelman USC). Funded by DOE
Self-driving 5G Network. (PI,Kiran). Funded by DOE
Intelligent Automation Wireless Network. (PI,Kiran). Funded by LBNL LDRD
POSEIDON (PI, Ewa Deelman USC). Funded by DOE
AI for Quantum Transducer (LDRD, PI Mekena Metcalf)
Uncertainty Quantification for Hyper parameter tuning (HYPPO, PI Julianne Mueller)
DAPHNE DAPHNE HOME (PI,Kiran)
PANORAMA Panorama Website
Agent-based Models FLAME Framework
Intent-based Network ESnet R&D Pages
Deep Learning based control for Lasers and Accelerators (LBNL-funded), Collaborating with Qiang Du
Deep reinforcement learning for Smart Build Control (Flexlab), Collaborating with Samir Touzani

Older Blog Posts

Professional Services

Programme Committee for CCGrid, Supercomputing, IPDPS (since 2019)
INDIS Committee by Scinet (since 2021)
Fellowship co-chair for N2Women for improving women researchers at Networking Conferences (2018-2020)
Associate Editor, IEEE Networking Letters
Elected member of COMSES Network for OpenABM (agent-based modelling) at Arizona State University.
Developer for FLAME (Agent-based modelling framework) used World-wide
Programme Committee duties at SciPy2018-2019, SC17, ICML workshop 2018
Reviewer for proposals submitted to NSF and EU
Instigated the ACM-Women professional chapter for UK

Bio

Dr. Kiran is Computer Networking and AI Scientist and currently lead Oak Ridge National Lab’s Quantum Networking Group. Her research explores machine learning and decentralized optimization for wide area networks, wireless, and Cloud infrastructures, needed for building ‘self-driving networks’ with themes for improving network performance, flexibility and high-speed distributed data transfers for science workflows. She received her Ph.D. and MSc (Eng) in Computer Science from the University of Sheffield, UK, in 2011, 2008, respectively. Before joining the lab in 2016, Kiran has worked with Universities of Sheffield, Leeds, Oxford, UCL, UC Berkeley, with industrial collaborations across the world - European, Asia Pacific and US. She also helped develop FLAME, an open-source agent-based platform, currently used worldwide for complexity research.

Her notable awards include the Royal Society Scientist in Westminster in 2015, the 2017 U.S. DOE Early Career Award, and ACM’s N2Women Rising Stars in Networking Award in 2021. She is an active member of ACM and a Senior Member of IEEE communities. </p>