DeepDig

A framework for deception-enhanced IDS training and evaluation.

Network Traffic Generation

The platform can be used to generate attack and benign traffic and evaluation data.

Supported attack types are described below. Additional attacks can be customized in the framework.

#	Attack Type	Description	Software
1	CVE-2014-0160	Information leak	OpenSSL
2	CVE-2012-1823	System remote hijack	PHP
3	CVE-2011-3368	Port scanning	Apache
4–10	CVE-2014-6271	System hijack (7 variants)	Bash
11	CVE-2014-6271	Remote Password file read	Bash
12	CVE-2014-6271	Remote root directory read	Bash
13	CVE-2014-0224	Session hijack and info leak	OpenSSL
14	CVE-2010-0740	DoS via NULL pointer deref	OpenSSL
15	CVE-2010-1452	DoS via request lacking path	Apache
16	CVE-2016-7054	DoS via heap buffer overflow	OpenSSL
17–22	CVE-2017-5941	System hijack (6 variants)	Node.js

Attack generation

Note: the IP address of the target servers can be modified in each attack script

Note: the IP addresses will have to point to a server with vulnerabilities listed above which has is not provided to execute attack workload:

Sample docker images with the vulnerabilities mentioned above can be found at https://hub.docker.com/r/hmlio/vaas-cve-2014-0160/ and https://github.com/Medicean/VulApps/tree/master/n/nodejs/1

cd trafficgen/attackgenerator
./run.sh

Benign traffic generation

Pre-requisite: the benign traffic generators need a wordpress deployment. It also requires the Buddypress and Woocommerce plugins for creating application profiles.

Please install firefox web browser if it is not available on your machine.

Install selenium for Python:

pip install selenium 

Run traffic generator:

cd trafficgen/benignGenerator/DataGen/wordpressautonew
./general.sh

ML Module & Experiments

Change dir to ml submodule: cd ml

Download and decompress the datasets from the link below:

https://drive.google.com/drive/folders/1Skrzw62SC5X8qAWcHVL9k8BpMTI4TNYG?usp=sharing

Create data directory:

mkdir data

For the OML experiments:

tar -C data -xvzf datafiles\_oml.tar.gz


### Online Metric Learning (OML)
Build container:

docker build -t oml -f Dockerfile.oml .

Run interactive shell into container:

docker run -it --rm -v data:/workspace/datafiles_oml oml bash

docker ps

get the id of the container

copy the data to the container by running this command on the host machine.

sudo docker cp data/datafiles_oml/ [yourcontainerid]:/workspace

To run experiments:
Inside the container prompt

cd /workspace/code/oml

./runincalltest_single.sh > output

cat output | grep Acc

cat output | grep FPR

cat output | grep TPR

cat output | grep F2

./runincalltesthuman\_single.sh > outputh

cat outputh | grep Acc

cat outputh | grep FPR

cat outputh | grep TPR

cat outputh | grep F2

For the SVM experiments: 

cat data_svm_split.tgz_* | tar -C data -xz

### SVM
Build container:

docker build -t svm -f Dockerfile.svm .

Run interactive shell into container:

docker run -it --rm -v data:/workspace/datafiles oml bash

To run experiments:

cd /workspace/code/svm ./scriptme_16.sh python parseResultsFile.py

Check `output` folder