Improved FHE Bootstrapping and Its Applications in Privacy Preserving Inference of Discretized Neural Networks

For optimal performance, we also employ an approximate gadget decomposition and provide improved parameter sets as in LMKCDEY (see binfhecontext.cpp).

Requirements

A C++ compiler, the NTL libraries.

Run the code

1. Configure, build and compile the project.

mkdir build
cd build
cmake -DWITH_NTL=ON .. 
make 

2. Run the boolean-xzdnew and DiNN program in build/bin/examples/binfhe.

Experimental Result(12th Gen Intel(R) Core(TM) i7-12700H @2.70 GHz and 16 GB RAM, running Ubuntu 20.04.6 LTS):

We recommend using the following CMake command-line configuration for best performance.

cmake -DWITH_NTL=ON  -DNATIVE_SIZE=32 -DWITH_NATIVEOPT=ON -DCMAKE_C_COMPILER=clang-12 -DCMAKE_CXX_COMPILER=clang++-12 -DWITH_OPENMP=OFF -DCMAKE_C_FLAGS="-pthread" -DCMAKE_CXX_FLAGS="-pthread" ..