/SmartZone

Primary LanguageC

SmartZone

SmartZone is a secure memory management method based on TrustZone and includes the lightweight neural network framework Tinylib. This application is based on Darknet DNN framework and needs to be run with OP-TEE, an open source framework for Arm TrustZone.

Description

This is the code repository for our paper "Memory-Efficient and Secure DNN Inference on TrustZone-enabled Consumer IoT Devices". (paper id: 1570937997)

Environment

For execution

  • Development board: Raspberry pi 3B+ (ARM Cortex A53 * 4, 1GB RAM)
  • REE OS: Linux-for-arm 4.14
  • TEE OS: OP-TEE 3.8.0
  • Simulator: QEMU v8

For compilation

  • Host OS for compiling OP-TEE: Ubuntu-22.04, 64bit, x86_64
  • GCC version：11.3.0
  • toolchain: aarch64-linux-gnu-gcc 10.2.1

Setup

(1) Set up OP-TEE

  1. Follow step1 ~ step5 in "Get and build the solution" to build the OP-TEE solution. https://optee.readthedocs.io/en/latest/building/gits/build.html#get-and-build-the-solution

  2. For real boards: If you are using boards, keep follow step6 ~ step7 in the above link to flash the devices. This step is device-specific.

    For simulation: If you have chosen QEMU-v7/v8, run the below command to start QEMU console.

make run
(qemu)c
  1. Follow step8 ~ step9 to test whether OP-TEE works or not. Run:
tee-supplicant -d
xtest

Note: you may face OP-TEE related problem/errors during setup, please also free feel to raise issues in their pages.

(2) Build Tinylib

  1. clone codes and datasets
git clone https://github.com/nkicsl/SmartZone.git

Let $PATH_OPTEE$ be the path of OPTEE, $PATH_Tinylib$ be the path of Tinylib, and $PATH_datasets$ be the path of dataset.

  1. copy Tinylib to example dir
mkdir $PATH_OPTEE$/optee_examples/Tinylib
cp -a $PATH_Tinylib$/. $PATH_OPTEE$/optee_examples/Tinylib/
  1. copy datasets to root dir
cp -a $PATH_datasets$/. $PATH_OPTEE$/out-br/target/root/
  1. rebuild the OP-TEE

For simulation, to run make run again.

For real boards, to run make flash to flash the OP-TEE with darknetz to your device.

Inference

By simply typing the following command, you can do inference using a pre-trained model.

tinylib classifier predict cfg/imagenet1k.data cfg/mobilenet_v1.cfg models/mobilenet_v1.weights data/cat.jpg -mov_size 262144

You can adjust the parameter -mov_size to change the size of shared memory.

File description

  • Compiled file/: Compiled binary files
  • Tinylib/: N-Tinylib,S-Tinylib,Tinylibm
  • dataset/: the experimental data of our paper