blocklengths

Blocklength, in information theory, coding theory, and cryptography, refers to the fixed number of units in a single processed block. The term is used to describe the size of a block that a system encodes, encrypts, or transmits in one operation. Different contexts measure blocklength in symbols, bits, or bytes, depending on the domain.

In error-correcting codes, a block code encodes k source symbols into n coded symbols; n is the blocklength. The rate R = k/n and the minimum distance d determine the code's efficiency and error-correcting capability (t = floor((d−1)/2)). Longer blocklengths can improve performance on many channels but increase latency and resource use and can complicate implementation for real-time or streaming data.

In cryptography, a block cipher processes data in fixed-size blocks, called the block length. AES uses 128-bit blocks, DES uses 64-bit blocks. When messages are not a multiple of the block length, padding is typically applied. The block length, together with the mode of operation (for example CBC, CTR), influences security properties such as resistance to certain attacks and the achievable throughput.

Blocklength also interacts with hardware and storage constraints, where block sizes align with memory pages or disk sectors, and with throughput and latency requirements in communications systems. In design practice, blocklength is chosen to balance error resilience, latency, implementation complexity, and compatibility with data formats and standards.