Cyclic Redundancy Check (CRC) Explained

Why a Stronger Method Was Needed

Parity bits and basic checksums (covered in the previous note) can miss certain patterns of multiple bit errors. CRC (Cyclic Redundancy Check) is a mathematically stronger technique, widely used in Ethernet and many other networking and storage systems, that catches a far broader range of real-world error patterns.

The Core Idea, Simplified

CRC treats the data being sent as one very large binary number, and divides it by a fixed, agreed-upon "generator" value using a specific kind of division (not ordinary arithmetic division). The remainder from that division becomes the CRC value, which is attached to the data before sending.

Data + generator polynomial  --(special division)-->  remainder = CRC value

Sender attaches this CRC value to the data before transmitting.

How the Receiver Verifies It

The receiver performs that exact same division on the received data (including the attached CRC value) using the identical generator value. If the remainder comes out to exactly zero, the data is considered intact; any other result means an error occurred somewhere during transmission.

Why CRC Catches Far More Error Patterns

The specific mathematical properties of the division process used mean CRC reliably detects all single-bit errors, all double-bit errors, and any burst of consecutive errors up to the length of the CRC value itself - a dramatically stronger guarantee than a simple parity bit or checksum provides.

Where CRC Shows Up in Real Systems

Ethernet frames (covered in a dedicated note in this section) include a CRC value in their trailer specifically for this reason - it's the mechanism that lets a network interface reliably detect when a frame was corrupted during transmission across the physical medium.