NTPP : Network Transmission Point Process

A probabilistic deep machinery that models the traffic characteristics of hosts on a network and effectively forecasts the network traffic patterns, such as load spikes.

File Distribution

|- data/
|- ntpp /
	|- models/
		|- data.py - Declare the Data Class
		|- model.py  - Declare the model
		|- plotter.py - Function to plot the Loss vs Epoch 
		|- predicter.py - Main function to train and validate the data
		|- scorer.py - Functions to find the discriminator and generative loss
	|- utils.py - Some useful functions 

How To Use

1. Prepare data : Use tshark or wireshark to convert .pcap to .csv

2. Download:

   git clone https://github.com/vedic-partap/NTPP.git
   cd NTPP

3. Install Requirements: pip install -r requirements.txt

4. Extract Data: python ntpp/test.py

5. Train Model: python -m ntpp.models.predicter

Output

Parameter

Arguments	Decription
events	Event File containing the vents for each host
times	File containing the time of the events for each host
save_dir	Root dir for saving models
int_count	Number of intervals
test_size	Train Test split. e.g. 0.2 means 20% Test 80% Train
time_step	Time Step
batch_size	Size of the batch
element_size	Element Size
h	Hidden layer Size
nl	Number of RNN Steps
seed	SEED
mode	What do you want ? train
epochs	Number of epochs
workers	Number of workers
learning_rate	Learning rate for the optimiser
metric	Metric used in discriminator loss
is_cuda	use GPU or not
optim	Optimiser

e.g. python -m ntpp.models.predicter --epochs 500 --batch_size 11 --optim Adam

Version of Python

Use Python 3

Links of Dataset

1. MACCDC : https://www.netresec.com/?page=MACCDC
2. ISTS: https://www.netresec.com/?page=ISTS
3. WRCCDC: https://archive.wrccdc.org/pcaps/
4. ISACDC: http://www.westpoint.edu/crc/SitePages/DataSets.aspx
5. WorldCup98: http://ita.ee.lbl.gov/html/contrib/WorldCup.html
6. UNIDC: http://pages.cs.wisc.edu/~tbenson/IMC10_Data.html
7. snu/bbittorrent: https://crawdad.org/~crawdad/snu/bittorrent/20110125/tcpdump/

In case of any doubt or if you want to contribute, contact vedicpartap1999@gmail.com