Ritwik Deshpande - Week #15 - The Language of Computers and the Internet

(Image Source: Wikipedia)

I’d like to deviate away from the usual topics surrounding the themes of language and memory for humans and focus on how these themes are implemented by computers. So, how exactly do computers communicate with each other, and how does their language work? As you might already know, your computer sends information via light waves with a particular frequency to your home’s router, which then goes underground through a wire and connects to the rest of the internet through your ISP (e.g. Xfinity).

However, there’s a lot more complexities to this than are apparent at first glance. If you truly think about it at a fundamental level, nothing is hidden by default when packets of data travel through the internet, and any attacker can listen in on these light waves and eavesdrop on your internet activity. Enter TLS. This is the technology that enables you to communicate over the internet with authenticity and privacy. You utilize this technology daily when you go to https://google.com (the “s” signalling secure, i.e TLS encryption is in use).

The way TLS works is truly fascinating and involves a lot of mathematical complexities. For the purpose of the blog post, I will try to simplify these and instead focus on the concepts themselves. First, some vocabulary. Symmetric encryption refers to encryption in which one “key” (a string of characters) is used to encrypt and decrypt the same information. Asymmetric encryption, on the other hand, has two functionalities, the second of which is important to this blog post: one “private” key is used to encrypt the information, and another “public” key is used to decrypt the information. The catch is that when you encrypt your information with a private key, a person with the public key who sees your encrypted message can confirm with the algorithm that it was truly you that signed it, and not a pretender.

TLS starts with asymmetric encryption, then transitions to symmetric encryption because it is more efficient. The reason symmetric encryption isn’t used from the start is because, intuitively speaking, the two parties communicating over the internet would not be able to agree on a symmetric encryption key without the possibility of an attacker listening in. This is where the Diffie-Hellman Key Exchange comes in. In DHKE, each party securely generates their own keys independently and also agree upon a common value that is transmitted publicly. Then, an algorithm is applied to “merge” these values in a way that the secret keys cannot be deciphered from the merged value. Then, the merged values are swapped and each individual’s secret key is added, resulting in both parties having a new secret key that is private to both parties yet transmitted over the public internet. The beauty of DHKE lies in the fact that a secure private key is generated over a publicly accessible channel! Although it might be hard to understand, a perfect visual that expresses DHKE is mixing paint, which I’ve attached at the top.