# Is my high school cryptography/steganography science fair project practical?

I'm a 12th grade student interested in a science project in cryptography and steganography for my school's science fair. I chose this path of computer science partly because my coding skills are weak (even though I've taken computer classes for 3 years) and partly because I wanted a more math-oriented field. After doing my research, I found that for my science project, I wanted to test how to implement both cryptography and steganography techniques to maximize data security in a message. My plan to execute this is skewered by the fact that my topic is very advanced and usually requires years of complex math. So far, this has been my plan:

Q: which cryptographic algorithm and which steganography technique put together will result in the most efficient, secure data?

I will be testing 3 cryptographic algorithms: DES, AES, and RSA I will be testing 3 steganography techniques: Adding bits to a file, and Least Significant Bit (LSB), and RGB color encryption

The purpose is to see which two combination of methods offer the most security, and at the moment there are 3x3=9 combinations of tactics I can test to see which is strongest.

So far example, using the crypto library in Java, I would be able to implement one of the cryptographic algorithms into my code for encryption. Then, using a steganography software tool, I would then hide the file using one of the steganography techniques. I was thinking the security of the combined techniques could be measured by the encryption time (the higher it is, the less efficient and secure), but my main concerns are:

1. Is this a practical project? Can this be executed like a true science fair project and
2. In what way other than encryption time should I measure the 'effectiveness' of the two techniques?

I have to start collecting data very soon, and I feel as though I'm underestimating how complex this experiment is. But I would really like feedback!

• One practical issue you'll run into with your plan is that, while DES and AES are both block ciphers, RSA is a different kind of encryption altogether, and is not usually used on its own. Also, you will probably not be able to demonstrate any measurable difference between AES and DES on your own, although breaking DES is getting easier every day as computers get faster. – Ilmari Karonen Dec 4 '16 at 20:13
• Also, while this question is a bit different than most that we get here, in my opinion it does fall within this site's scope, and while it's somewhat broad and subjective, it's not too broad to be answerable. Of course, that won't guarantee that you'll necessarily receive good answers (or even any at all), but at least there's a chance that you will. – Ilmari Karonen Dec 4 '16 at 20:19
• I downvoted this question as: It doesn't provide as why DES,AES,RSA was chosen (and i'm fairly sure those were chosen without any research); It doesn't provide why it chosen it's stenography choices and most important: it doesn't define how it will measure security, which in my opinion makes it impossible to answer – axapaxa Dec 4 '16 at 22:14
• I personally don't see this as a practical project. The difference of DES and AES is well known and "just" RSA is not going to be used for encryption of larger data either. The strength of the steganography itself is not objective. – Maarten Bodewes Dec 5 '16 at 1:21
• @axapaxa Thank you for your commentary. I should have been clearer, but admittedly I don't have a solid grasp of how either works. Since I now know my project idea is too broad, should I focus on a more specific goal? For instance, instead of striving to enhance the security of cryptography and steganography in general, should I focus on making an aspect of their properties more secure? – Ingrid Dec 5 '16 at 15:56

which cryptographic algorithm and which steganography technique put together will result in the most efficient, secure data?

AES and DES are already a combination of separate algorithms involving multiple steps each. They are already "put together".

Steganography won't add additional security because of Kerckhoffsâ€™ principles.

In what way other than encryption time should I measure the 'effectiveness' of the two techniques?

Computation time is not a good indicator of security. If AES and DES are used together, the security will be equal to AES' security by itself. This is because DES' seed can be guessed faster than brute force - it has been broken.

Is this a practical project? Can this be executed like a true science fair project

It is not a practical project because neither AES nor DES can be broken using any machine you own, therefore output (no matter how much you massage or analyse it) will always appear as random bits. This means there will be no way to measure the security provided by combining methods.

• "If AES and DES are used together, the security will be equal to AES' security by itself. " This is not true. – Aleph Dec 7 '16 at 11:26