Stratosphere Project

Protect NGOs from malicious behaviors in the network with Machine learning

Sebastián García PhD.

sebastian.garcia@agents.fel.cvut.cz

Live Slides        bit.ly/stf-cisco

NGOs are at risk

Nepal Earthquake 2015.

People In Need NGO

NGOs are highly political targets.

Attacked by powerful actors

  • No resources. Not their goal.

  • Strong concerns about their privacy.

  • Concerns about Trust.

Problems for NGOs

Stratosphere Project

https://stratosphereips.org/

Free

Software

Machine

Learning

Behavioral

IPS

Protecting

NGOs

Stratosphere Technical Pillars

Less is More

Disassociate

Verify

Analyze the behavior of group of flows.

Representation of behaviour from detection.

With real and labeled datasets and networks

Behaviors

  • Your behavior is usually the same when connecting with the same service.
     

  • Group flows going to a specific service by ignoring the source port. We have our connection.

    • ​10.0.2.2-60.60.60.1-80-tcp
       

  • The connection, composed of several flows, now shows a behavior.

  • Using a service, you go from one state to the next state.

  • Each flow has its own state.

  • We model the states using four features.

    1. Size of the flow.

    2. Duration of the flow.

    3. Periodicity of the flow.

    4. Time between flows.

From Connections to States

States as Letters

The Behavior of a Connection

10.0.2.111-217.23.10.139-80-tcp     55*V0v00v*E*v*v*v*v*E*v

1 flow -> 4 features -> 1 letter + 1 symbol

Stratosphere Behavior Demo

 

Stratosphere Testing Framework: Create and Analyze Behaviors

https://mcfp.felk.cvut.cz/publicDatasets/CTU-Malware-Capture-Botnet-112-1/

About the Behaviors

  • Malware mostly generate the same behavior.

  • Changing the behavior is costly for the attacker.

  • These behaviors do not expire quickly.

  • Infections go unnoticed for hours. There is time.

From the letters create a Markov Chains behavioral model

Detection with Markov Chains

  • Train Markov Models with known Behaviors: Malware and Normal.

Detection with Markov Chains

  • Compare the unknown traffic of a network to each trained Markov Model.

Trained M1

Unknown Connection

87,a,a,b,B,i*i*i*i*i (?)

Trained M2

Trained M3

Stratosphere Detection Demo

 

  • Stratosphere Testing Framework

  • Stratosphere Linux IPS

Results

  • How to measure?

    • Packets/Flows/Connections/IPs?

    • Per minute? Per hour? Per day?

    • Who is putting the labels?

    • In Stratosphere it also depends on the models used.
       

  • Stratosphere Malware and Normal Dataset

    • https://stratosphereips.org/category/dataset.html 

Results

  • In our datasets

    • 96% TPR. Our own botnet traffic connections that are detected.

  • Real Traffic

    • ~0.0002% FPR (30 FP in 132,000 connections/5min)

  • Novel Success cases: Linux Botnet, DDoS, etc.

  • Errors? For sure.

Stratosphere Data Analysis

  • Cloud-based Detection service for NGOs.

  • Add new algorithms continually.

  • Update the models.

  • Verify the detections if necessary.

 

  • NGOs can send the Flows or only the letters! Privacy matters.

NGOs believing in us

  • NGO helping 22 countries. Ukraine, Syria, Africa, etc.

  • Manages .cz domain and protects +2,000 Internet Networks with his own Router/Firewall.

  • ICT help for policy makers in 20 African Countries

Our wish to Grow needs Funds

  • Support more Researchers.

  • Security Analysts for NGOs traffic.

  • Developers of GUI.

  • Infrastructure for Stratosphere SDA.

  • We want to protect more NGOs!

Thanks!

Sebastian Garcia

sebastian.garcia@agents.fel.cvut.cz

@stratosphereips

https://stratosphereips.org