The aim of this thesis is to develop a tool for automatic detection of undocumented JavaCard API located on a smartcard with the JavaCard platform with a help of side-channel analysis. The list of installed packages (for other than public ones [2]) with their classes and methods cannot be directly obtained from smartcard due to (intentionally) missing support for such an operation. However, the instruction-level side-channel leakage may enable to detect and reconstruct undocumented API.  

The work will extend the analysis done by Deshmukh in 2018 [1] with power side-channel analysis to obtain more detailed insight into card's behavior during AIDs, classes and methods probing.

The theoretical part of the thesis will precisely describe the JavaCard package management system, structure of cap packages and existing research covering reverse-engineering of code on the JavaCard Virtual Machine level.

The practical part will develop an automatic testing toolchain, that will cover the following functionality:
1. An experiment to detect undocumented packages with unknown AID utilizing side-channel leakage during LOAD operation [3].
2. Generate code/applet cap to instantiate specific class using its token, its method(s) and possible method signature using known fingerprints from the public JavaCard API.
3. Examine results for at least five real smartcards.

The resulting source code will be available under permissive open-source licence (MIT, Apache2) in IS and corresponding GitHub repository.   

References:
[1] Sujeet Daulat Deshmukh, Tool for detection of JavaCard libraries on smartcards https://is.muni.cz/th/k06n3/
GlobalPlatform specification: https://globalplatform.org
[2] jcAIDScan: https://github.com/petrs/jcAIDScanOracle, JavaCard platform: https://www.oracle.com/technetwork/java/embedded/javacard/overview/index.html
[3] M.Paljak, GlobalPlatformPro: https://github.com/martinpaljak/GlobalPlatformPro