CRCsjekk

CRCsjekk, often written as CRC-sjekk, refers to the process of using a cyclic redundancy check (CRC) to verify the integrity of data in digital systems. A CRC is a short, fixed-size value derived from the data through polynomial division. The resulting checksum is stored or transmitted with the data, and the receiver recomputes the CRC to see if the data have been altered in transit or storage.

The strength of a CRC depends on the chosen polynomial and the size of the CRC (for example, CRC-16 or CRC-32). Different polynomials offer varying capabilities for detecting common error patterns, such as single-bit errors, two-bit errors, and burst errors of limited length. CRCs are widely used in network protocols (Ethernet, PPP), storage formats (ZIP, TAR), and various data buses and interfaces because of their speed and simplicity.

Implementation typically involves choosing an initial value, applying bitwise or table-driven calculations as data is processed, and then optionally applying a final XOR or data reflection. In practice, CRCs can be computed in hardware or software and can be optimized for streaming data, enabling fast error detection with minimal overhead.

Limitations include that CRCs are designed for error detection rather than data authentication; they do not protect against deliberate tampering or spoofing. They also do not guarantee detection of all possible errors, though they detect a high proportion of random errors and burst errors. For security-critical scenarios, CRCs are used alongside cryptographic checksums or digital signatures.