Diffie-Hellman & The Mathematics Behind It : Session 2

This session moved from primitives into protocols, exploring how two parties can agree on a shared secret over a completely insecure, publicly observable channel. Grounded in modular arithmetic and the discrete logarithm problem, the session walked through the full Diffie-Hellman exchange step by step and surfaced its key limitation the lack of built-in authentication which set up the hands-on implementation session later in the series.

This session went deep into the mathematics required to understand why Diffie-Hellman (DH) works. The group first reviewed modular arithmetic  the "clock arithmetic" where numbers wrap around after reaching a fixed modulus p. Modular exponentiation (g^x mod p) was introduced as the core operation, and participants worked through small numerical examples by hand to build intuition.

The discrete logarithm problem (DLP) was then introduced: given g, p, and the result of g^x mod p, it is computationally very difficult to determine x, even though computing g^x mod p itself (the forward direction) is fast.

This asymmetry easy in one direction, hard to reverse is what makes DH secure, similar in spirit to how hash functions are easy to compute but hard to invert.

The full Diffie-Hellman protocol was walked through step by step:

1. Both parties (Alice and Bob) publicly agree on a large prime p and a generator g.
2. 
   
3. Alice privately chooses a secret integer a and computes
4. A = g^a mod p, sending A to Bob.
5. 
   
6. Bob privately chooses a secret integer b and computes
7. B = g^b mod p, sending B to Alice.
8. 
   
9. Alice computes the shared secret as B^a mod p;
10. 
    
11. Bob computes it as A^b mod p. Both arrive at the same value: g^(ab) mod p.
12. 
    

Crucially, the session emphasized that an eavesdropper who intercepts p, g, A, and B still cannot feasibly compute the shared secret without solving the discrete logarithm problem to recover a or b.

The group also discussed the well-known limitation of basic DH: it provides no authentication. Without an additional mechanism to verify identities, the protocol is vulnerable to a man-in-the-middle (MITM) attack, where an attacker performs separate key exchanges with each party while impersonating the other.

This limitation directly motivated the implementation session later in the series, where authentication of exchanged keys was addressed in practice.

Parameter selection was also discussed why small or poorly chosen primes make DH vulnerable to attacks like Logjam, and why current best practice recommends using well-vetted standard groups (e.g., RFC 3526 groups) or moving to elliptic-curve-based Diffie-Hellman (ECDH), previewed as a topic connecting to the later ECC session.

This was a math-first session establishing exactly how two parties can agree on a shared secret over a completely insecure, publicly observable channel one of the most important breakthroughs in the history of cryptography while also surfacing its key limitation (lack of authentication) as a lead-in to later sessions.