RSA encryption is the backbone of secure internet communication. It is a type of public-key cryptography that protects emails, digital transactions, and sensitive data flowing across networks. The algorithm is named after its creators: Ronald L. Rivest, Adi Shamir, and Leonard M. Adleman. They developed the system while teaching at the Massachusetts Institute of Technology.
The security of RSA relies on a simple mathematical hurdle that is easy to create but nearly impossible to reverse. A user starts by choosing two large prime numbers, p and q. These numbers are so big that multiplying them to get n is trivial, but factoring n back into p and q is computationally prohibitive.
Think of it this way: multiplying two huge numbers takes a split second. Trying to figure out which two numbers were multiplied to get that result? That could take billions of years with current technology.
As of 2015, U.S. government security standards require the modulus n to be at least 2,048 bits. In decimal terms, this means p and q are around 308 digits each, creating a 617-digit number for n. The largest numbers we have successfully factored are only half that size. The difficulty of factoring roughly doubles for every additional three digits in the modulus. This exponential growth in complexity ensures that 617-digit keys remain safe from brute-force attacks for decades.
The Mechanics of Key Generation
Once p and q are selected, the system generates a public and a private key. The user picks an integer e less than n. This number must be relatively prime to the product of (p – 1) and (q – 1). In plain English, e shares no common factors with that product other than 1.
This specific choice guarantees the existence of another number d. When you multiply e by d and divide by the least common multiple of (p – 1) and (q – 1), the remainder is 1.
Finding d is easy if you know p and q. The Euclidean algorithm calculates it in moments. But if you do not know p and q, finding d from e is just as hard as factoring n. That difficulty is the entire basis of RSA’s cryptosecurity.
The security of RSA isn’t magic; it’s math. Specifically, it’s the asymmetry between easy multiplication and impossible factorization.
Public Keys vs. Private Keys
The labels d and e describe functions, not strict roles. The keys are interchangeable. This flexibility allows RSA to serve two distinct purposes: secrecy and authentication.
To create a secrecy channel, user A publishes e and n in a public directory. User A keeps d secret. When someone wants to send a private message to A, they convert the message into numbers smaller than n and encrypt it using e and n. Only A can decrypt it because only A holds d. The assumption—and strong evidence—holds that no one else can decrypt the message without factoring n.
Authentication works differently. User A publishes d and n and keeps e secret. User B wants to verify they are talking to A. B sends a challenge message to A. If A returns the encrypted version using e, B can decrypt it using the public d. If the decrypted message matches the challenge, B knows the sender possesses e. Since only A has e, the identity is verified.
Digital Signatures and Hashing
Digital signing adds a layer of complexity. It requires a cryptosecure hashing function. This is a public function that compresses any message into a smaller digest. Every bit of the digest depends on every bit of the original message. Change one bit in the original message, and roughly half the bits in the digest will flip.
A cryptosecure hash makes it computationally infeasible to find a message that produces a specific preassigned digest. It also makes it hard to find a different message with the same digest as a known one.
To sign a message, A encrypts the digest with their secret e and appends it to the message. The message itself does not need to be secret. Anyone can use A’s public key d to decrypt the signature and recover the digest. They then compute the digest of the received message independently. If the two digests match, the signature is valid. Only A could have encrypted the digest because only A knows e.
The Hybrid Reality
There is a cost to this security. Asymmetric encryption involves significantly more computation than symmetric (single-key) encryption. This high processing demand cuts throughput drastically. For about 20 years, single-key algorithms have achieved throughputs 1,000 to 10,000 times higher than two-key algorithms for comparable security levels.
Because RSA is slow, it is rarely used to encrypt large amounts of data directly. Instead, it powers hybrid systems. RSA handles the heavy lifting of authentication and digital signatures. It also securely exchanges a randomly generated session key. Once that key is shared, the system switches to a single-key algorithm for the high-speed transmission of the actual data. When the session ends, the key is discarded.
This combination gives us the best of both worlds. We get the secure key exchange of public-key cryptography with the speed of symmetric encryption for the bulk of our data.
The math holds up. The keys stay safe. But as computing power climbs, the definition of “large” shifts. What is unbreakable today might be solvable tomorrow. We keep building larger primes. We keep hoping the gap between easy multiplication and hard factorization stays wide enough to protect our digital lives.
