Privacy Preserving Graphical Security Model

This software enables one to evaluate the security of a network system in a privacy preserving mannaer.

Prerequesites

• OS: Ubuntu 18.04 above
• HE libraries: TFHE and SEAL homomorphic encryption (HE) have to be installed.

1. Installating and Testing

Install the two HE libraries.
To use PPGSM, install two homomorphic encrpytion libraries: TFHE and SEAL.

1.1. Installing TFHE

Get the TFHE installation file available at: https://github.com/tfhe/tfhe/tarball/master
(Note that the TFHE github main version would have a problem on installing TFHE)
Unzip the tar.gz file into a directory and go to the directory where README.md and LICENSE are.
Then, make directory named 'build':

mkdir build   

Go to the 'build' directory:

cd build   

Then, make installation file in the build directory:

cmake ../src   

Finally, using commands below, install TFHE:

make    
sudo make install   

1.2. Installing SEAL

Get SEAL installation file: git clone https://github.com/microsoft/SEAL
Go to the SEAL native directory:

cd SEAL/native   

Make 'build' directory, change directory and install:

mkdir build    
cd build    
make installation file: cmake ../src

Finally, install SEAL:

make    
sudo make install       

2. Running testcode
   Install PP-GSM module:

   git clone https://github.com/PPGSM/PPGSM

Move to PPGSM directory:

cd PPGSM/PPGSM

Compile test code:

cmake .   

Run compiled binary:

./testPPGSM sampleData/graphinfo1 sampleData/graphinfo2 sampleData/graphinfo3   

Utilities

A. User-side utilites (functions can found in utility/graph_client.h)
creating real/dummy node/edges
creating graph

B. Server-side utilites (functions can found in structure/graph.h)
probing encrypted graph
simple metrics: mincut, minlength, calculating degree vector
graph pruning operation
complex operations: probability of attack succeess, return on investment
centrality metrics: clossness/harmonic/etweenness centrality
heurisitc dummy node/edge addition