1. OpenDaylight Based Machine Learning for Networks
Contributions from the ODL MILA WorkGroup
YuLing Chen, Cisco
Prem Sankar, Ericsson
Sept. 2016

2. Contributed by…

3. Why we need machine learning in ODL?

4. Why we need Machine Learning in SDN
Software Defined Network needs to be intelligent.
To be aware of the runtime status of the network.
To make the right decisions to adjust the policies for the traffic control.
To dynamically change the policies according to the analytics results.

5. Make ODL the Brain of Network
GBP
Topo
TSDR
SFC
VPN
service
AI/ML
OF/Netconf/BGP

6. Use Cases of a smart and intelligent SDN controller
Traffic Control and Routing Optimization
Congestion Control
Traffic Pattern Prediction
Routing Optimization
Resource optimization
Networking resource allocation optimization
Cloud resource management optimization
Security and Anomaly Detection
DDoS Attack detection
Troubleshooting and Self-healing

7. Example Use Case – Traffic congestion prediction with automated control

8. Prediction using Weka leveraging data collected in TSDR
#ODSummit

9. The Goals of ODL MILA Group

10. Goals of ODL MILA Workgroup
To adapt SDN architecture to machine learning requirements.
To provide an application framework for Machine Learning in ODL
To add the necessary components for Intelligence(Knowledge) Plane
To integrate with ODL native network data collection and traffic control services.
To facilitate machine learning application development on ODL.
Integrate with third party machine learning algorithms
Provide abstract and generic northbound interfaces for Machine Learning applications
Hide the details of Advanced Analytics and Machine Learning complexities.

11. How to realize a smart and intelligent SDN Controller
Network status awareness
Rely on time series data collected from the network
Traffic Control Policy Change decision making
Based on the advanced analytics and machine learning.
Dynamic change of Control policies
Automatically change the traffic control policies based on the analytics results.
Advanced Analytics & Machine Learning
Automated Traffic Control
Time Series Data Collection

12. Time Series Data Repository

13. TSDR Capabilities and Architecture Framework Roadmap
Control Flow
Data Flow

14. TSDR Integrated Architecture in ODL
TSDR Data Services including Data Collection, Data Storage, Data Query, Data Purging, and Data Aggregation are MD-SAL services.
Data Collection service receives time series data published on MD-SAL from MD-SAL southbound plugins.
Data Collection service communicates with Data Storage service to store the data into TSDR.
TSDR data services access TSDR Data Stores such as HBase Data Store through generic TSDR Data Persistence Layer.

15. ODL MILA Framework Architecture

16. ODL MILA framework in the ODL ecosystem
Enable AI/ML on both historical and real-time data paths.
Many use cases would require both offline and online ML on the time series data.
External events could be additional input for accurate machine learning results.
Feed back the results to SDN control path for automatic traffic steering and policy placement.
Well-defined interface among the components towards future standardization of advanced analytics in SDN.

17. ODL MILA framework PoC Architecture
PoC of both historical offline machine learning and real-time online machine learning
Collect the time series data
Persist into scalable data storage
Publish to high performance data bus
Integrate with external machine learning libraries
Spark MLlib
DeepLearning4J
Collect OpenFlow Stats and apply machine learning algorithms
k-means clustering

18. ODL MILA framework PoC Result
Snapshots of flow size categorization are captured using k-means clustering algorithms.
Different colors show the different category/cluster that each flow falls into based on the flow size.

19. Demo

20. ODL MILA Workgroup related info and contacts
Contacts
YuLing Chen
Prem Sankar

21. Thank You