Gilbert-Varshamov bound

From Wikipedia, the free encyclopedia

The Gilbert-Varshamov bound is a bound on the parameters of a (not necessarily linear) code. It is occasionally known as the Gilbert-Shannon-Varshamov bound (or the GSV bound), but the Gilbert-Varshamov bound is by far the most popular name.

[edit] Statement of the bound

Let

Aq(n,d)

denote the maximum possible size of a q-ary code C with length n and minimum Hamming weight d (a q-ary code is a code over the field \mathbb{F}_q of q elements).

Then:


\begin{matrix}
\\
A_q(n,d) \geq \frac{q^n}{\sum_{j=0}^{d-1} \binom{n}{j}(q-1)^j} \\
\\
\end{matrix}

For the q a prime power, one can improve the bound to A_q(n,d)\ge q^k where k is the greatest integer for which A_q(n,d) \geq \frac{q^n}{\sum_{j=0}^{d-2} \binom{n-1}{j}(q-1)^j}

[edit] Proof

Let C be a code of length n and minimum Hamming distance d having maximal size:

| C | = Aq(n,d).

Then for all x\in\mathbb{F}_q^n there exists at least one codeword c_x \in C such that the Hamming distance d(x,cx) between x and cx satisfies

d(x,c_x)\leq d-1

since otherwise we could add x to the code whilst maintaining the code's minimum Hamming distance d - a contradiction on the maximality of | C | .

Hence the whole of \mathbb{F}_q^n is contained in the union of all balls of radius d − 1 having their centre at some c \in C :

\mathbb{F}_q^n =\cup_{c \in C} B(c,d-1) .

Now each ball has size


\begin{matrix}
\sum_{j=0}^{d-1} \binom{n}{j}(q-1)^j \\
\end{matrix}

since we may allow (or choose) up to d − 1 of the n components of a codeword to deviate (from the value of the corresponding component of the ball's centre) to one of (q − 1) possible other values (recall: the code is q-ary: it takes values in \mathbb{F}_q^n). Hence we deduce


\begin{matrix}
|\mathbb{F}_q^n| & =& |\cup_{c \in C} B(c,d-1)| \\
\\
& \leq& \cup_{c \in C} |B(c,d-1)| \\
\\
& = & |C|\sum_{j=0}^{d-1} \binom{n}{j}(q-1)^j \\
\\
\end{matrix}

That is:


\begin{matrix}
A_q(n,d) & \geq & \frac{q^n}{\sum_{j=0}^{d-1} \binom{n}{j}(q-1)^j} \\
\end{matrix}

(using the fact: |\mathbb{F}_q^n|=q^n).

[edit] See also