Binary Fields
A field is a set equipped with addition and multiplication operations, which satisfy the commutative, associative and distributive laws, and where everything nonzero has an inverse. A finite field is moreover finite.
The finite fields you're probably most used to look like , where is a large prime. Arithmetic in that set is carried out modulo . Elements of the field can be represented, concretely, as -bit strings. Some of those strings—i.e., the ones representing integers in the high range —will be "off-limits".
Here, we work in characteristic 2. This means that we construct our fields differently.
Construction
The simplest binary field is , the field with two elements.
By taking algebraic extensions of this field, we can get further binary fields. We pick a bit-length. For now, we'll stick with . We fix once and for all a global irreducible polynomial of degree . The corresponding quotient ring is a field, namely .
Concretely, we can represent elements as polynomials of degree less than ; say, . In other words, the list of monomials is an -basis of . Thus elements of get represented in practice as -bit words (or say as pairs of 64-bit words).
Algebraic Properties
We fix a binary field, say . The multiplicative group of units of , written , is the set of nonzero elements of , with multiplication as its group operation. The multiplicative group of units of a finite field is cyclic, and so in particular is. This means that we can fix an element such that the powers , for , exactly exhaust .
In , the Frobenius endomorphism sends . The Frobenius endomorphism is -linear and injective; in finite fields, it's also surjective, and so invertible.