Master Thesis projects
Are you a student with a great idea for a master thesis project? I am always looking for new insights into cryptography, digital currencies like bitcoin, random number generation, computer security etc. Be sure to contact me so we can arrange something.
Below I keep a list of interesting ideas and possible openings. Be sure to also check the list of thesis ideas on the Information Coding Group homepage.
An optimized scrypt miner
Also check my separate page on digital currencies and bitcoin.
Digital currencies such as bitcoin is a hot, new trend that has many promises in store for the future.
Since the cryptographic algorithm that forms the basis of bitcoin is easy to implement in hardware, custom-built processors have completely disrupted the nature of bitcoin mining. Lately, alternative currencies have appeared that try to avoid this by using the heavily memory-dependent algorithm scrypt.
Examples of scrypt-based currencies include litecoin, dogecoin and many more. These currencies can be mined with much more affordable GPU:s without having to compete with the very expensive specialized hardware. This master thesis project has the aim of producing a heavily optimized scrypt mining software for GPU:s as well as analyzing differences in how the software can be implemented on different platforms such as NVidia and AMD.
This master thesis project would be supervised by myself and Ingemar Ragnemalm.
Random number generation based on quantum phenomena
Reliable random number generation is a cornerstone in modern cryptography. There are many examples where otherwise good cryptographic implementations have failed miserably due to bad random numbers. A very common method of generating (pseudo-)random numbers today is by linear feedback shift registers implemented in software. While they are easy to implement and generally have very high generation rate, they are linear and periodic which in some circumstances can weaken the system.
Quantum mechanics can provide randomness which is random at the fundamental level. Therefore, it is possible that a random number generator based on quantum-mechanical phenomena can give better results than a software implementation. But what is "better"? How can we be sure that the output is actually random?
We recently acquired a hardware random number generator that is based on quantum phenomena. The Quantis by ID Quantique (USB version) is a plug-and-play solution that works by sending photons through a semi-transparent mirror.
Your tasks are to:
- Determine state-of-the-art methods for testing random number generators
- Investigate the hardware random number generators.
- Compare the quantum device with other modern hardware and software random number generators.
This master thesis project would be supervised by myself and Jan-Åke Larsson.