Joint Protection Scheme for Deep Neural Network Hardware Accelerators and Models
- 5 June 2023
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
- Vol. 42 (12) , 4518-4527
Abstract
Deep neural networks (DNNs) are utilized in numerous image processing, object detection, and video analysis tasks and need to be implemented using hardware accelerators to achieve practical speed. Logic locking is one of the most popular methods for preventing chip counterfeiting. Nevertheless, existing logic-locking schemes need to sacrifice the number of input patterns leading to wrong output under incorrect keys to resist the powerful satisfiability (SAT)-attack. Furthermore, the DNN model inference is fault tolerant. Hence, using a wrong key for those SAT-resistant logic-locking schemes may not affect the accuracy of DNNs. This makes the previous SAT-resistant logic-locking scheme ineffective on protecting DNN accelerators. Besides, to prevent DNN models from being illegally used, the models need to be obfuscated by the designers before they are provided to end-users. Previous obfuscation methods either require a long time to retrain the model or leak information about the model. This article proposes a joint protection scheme for DNN hardware accelerators and models. The DNN accelerator is modified using a hardware key (Hkey) and a model key (Mkey). Different from previous logic locking, the Hkey, which is used to protect the accelerator, does not affect the output when it is wrong. As a result, the SAT attack can be effectively resisted. On the other hand, a wrong Hkey leads to substantial increase in memory accesses, inference time, and energy consumption and makes the accelerator unusable. A correct Mkey can recover the DNN model that is obfuscated by the proposed method. Compared to previous model obfuscation schemes, our proposed method avoids model retraining and does not leak model information.Keywords
All Related Versions
Funding Information
- U.S. Air Force Research Laboratory (FA8650-20-C-1719)
This publication has 28 references indexed in Scilit:
- Provably-Secure Logic Locking2017
- Removal Attacks on Logic Locking and Camouflaging TechniquesIEEE Transactions on Emerging Topics in Computing, 2017
- Eyeriss: An Energy-Efficient Reconfigurable Accelerator for Deep Convolutional Neural NetworksIEEE Journal of Solid-State Circuits, 2016
- Deep Residual Learning for Image Recognition2016
- SARLock: SAT attack resistant logic locking2016
- Evaluating the security of logic encryption algorithms2015
- Improving logic obfuscation via logic cone analysis2015
- A novel hardware logic encryption technique for thwarting illegal overproduction and Hardware Trojans2014
- Fault Analysis-Based Logic EncryptionIEEE Transactions on Computers, 2013
- Security analysis of logic obfuscation2012