Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

Attendees

Sridhar Rao

Rohit Singh Rathaur 

Girish

Karthik

renukananda td

Hemashree R

Vishnu Ram

Beth Cohen

Ildiko

Deekshitha J P


Sl. No.TopicPresenterNotes
1

Thoth - ITU

(a) Problems and Datasets in ITU.

(b) Host and Competitor


Challenge: 

https://aiforgood.itu.int/about/aiml-in-5g-challenge/

Wiki Page to capture the collaboration: Collaboration - ITU 

  1. Take up relevant problems (NFV) and start creating models - 2 Problems.
    1. Rohit:
      1. Log Analysis - Anomaly detection problem shared by China Unicom. (Dikshita and Rakshita)
      2. Failure detection - KDDI.
  2. Initiate dialogue with TSC-Anuket, LFN and EUAG to be one of the Hosts for next round of challenge.
    1. Tentative dates: Kickstart (February).
    2. EUAG: Can consider Opendataset and define a novel problem.
  3. Compete in the challenge.
  4. Join as Mentor.

Github link for the Problems/Solutions: https://github.com/ITU-AI-ML-in-5G-Challenge 

Data from Real-World: Turkcell and China Unicom.

2

Kubernetes - AI/ML

(a) Problems

(b) Data Sources.


Refer below the Problems.

Logs: System components, Pods and Deployment.

System Components: Fluentd. Refer to presentation by Zebrium.

3Kubeflow - An Introduction
Alternative: LF Acumos.
4Developers Workshop sessions

Please remember to propose sessions for the upcoming Developer's Conference. 

LFN Developer & Testing Forum, Jan 10-13, 2022


More Links:

  1. This contains real network data from Turkcell - https://github.com/Turkcell/ITU-AIMLin5GChallenge-2021/blob/main/RLF_Prediction_ITU_AIML_Challenge_Data/TurkcellExampleProject.ipynb
  2. This contains dataset from NIST: https://github.com/usnistgov/PS-002-WALDO#authors check also:  https://github.com/usnistgov/PS-002-WALDO/ 
  1. GNN challenge dataset from Barcelona: https://bnn.upc.edu/challenge/gnnet2021
  2. Xilinx: https://challenge.aiforgood.itu.int/match/matchitem/34 uses DeepSig RadioML 2018 dataset.
  3. ZTE: https://wiki.lfaidata.foundation/display/ADLIK/2021+Combinatorial+Optimization+Challenge%3A+Delivery+route+optimization has the dataset
  4. UPF : https://www.upf.edu/web/wnrg/2021-edition uses https://zenodo.org/record/4059189
  5. Brazil: https://github.com/knowledgedefinednetworking/RouteNet-challenge/tree/master/data/sample_data/validation


Categories of Problems

Example Reference Works

Performance Metric Prediction

1.End-to-End Latency Prediction of Microservices Workflow on Kubernetes

2.Edge Intelligence in Softwarized 6G: Deep Learning-enabled Network Traffic Predictions

Resource Scheduling

1.Speculative Container Scheduling for Deep Learning Applications in a Kubernetes Cluster

2.A Proposal of Kubernetes Scheduler Using Machine-Learning on CPU/GPU Cluster

3.Ananke: A framework for Cloud-Native Applications smart orchestration

ML Platform on K8S*

1.Multi-Tenant Machine Learning Platform Based on Kubernetes

2.EdgeAI: Edge-native distributed platform for artificial intelligence

Resource Consumption Monitoring (Detection/Prediction)

1.A Machine Learning Model for Detection of Docker-based APP Overbooking on Kubernetes

2.Architecture for Enabling Edge Inference via Model Transfer from Cloud Domain in a Kubernetes Environment (Energy)

3.AI-Based Resource Management in Beyond 5G Cloud Native Environment

4.Intelligent and Autonomous Management in Cloud-Native Future Networks

Testing

Artificial Intelligence assisted Canary Testing of Cloud Native RAN in a mobile telecom system

TE and Others

“End-to-End Management of All Optical Disaggregated Network and Applications with Cloud Native Environment”

Intelligent O-RAN for beyond 5G and 6G wireless networks