Physical security threats appear at the chip level, where an attacker can measure or physically influence a cryptographic operation performed by the chip's circuit. The side-channel analysis exploits additional sources of information (called physical observations), including timing, power consumption, or electromagnetic emissions (EM), among others. Malicious data modifications can be caused by fault injection, which can be performed using optical, electromagnetic, power, and clock glitches. These attacks pose a serious threat to modern cryptographic implementations.

The project aims to implement the first secure RSA crypto library resistant to side-channel and fault injection attacks. We consider classical side-channel attacks like correlation power analysis and more powerful techniques like template attacks. 

The project consists of five parts:

1. Analysis of existing crypto libraries and choosing the most suitable implementation for implementing protections. The base implementation should be efficient and reasonably easy to extend. 
2. Literature Review and choice of countermeasures to implement: identifying relevant attacks and choosing the best countermeasures to stop them. The student will need to figure out how to combine different countermeasures efficiently. In this stage, new countermeasures can also be designed. 
3. Implementation - the chosen design needs to be implemented for an embedded target based on ARM-Cortex M4.
4. Performance Evaluation: the new protected implementation should be compared against the unprotected original implementation as well as other RSA implementations from commonly used crypto-librarires. 
5. Optional step: if there is time, a side-channel evaluation/testing can be performed.

The project approach is similar to what was done for Curve25119 in:
https://github.com/sca-secure-library-sca25519/sca25519

The division of the tasks within the projects is approximately: 60% programming and 40% literature study.