Gilbert–Varshamov bound

In coding theory, the Gilbert–Varshamov bound (due to Edgar Gilbert[1] and independently Rom Varshamov[2]) is a limit 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 name "Gilbert–Varshamov bound" is by far the most popular. Varshamov proved this bound by using the probabilistic method for linear codes. For more about that proof, see Gilbert–Varshamov bound for linear codes.

Statement of the bound

Let

${\displaystyle A_{q}(n,d)}$

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

Then:

${\displaystyle A_{q}(n,d)\geqslant {\frac {q^{n}}{\sum _{j=0}^{d-1}{\binom {n}{j}}(q-1)^{j}}}.}$

Proof

Let ${\displaystyle C}$ be a code of length ${\displaystyle n}$ and minimum Hamming distance ${\displaystyle d}$ having maximal size:

${\displaystyle |C|=A_{q}(n,d).}$

Then for all ${\displaystyle x\in \mathbb {F} _{q}^{n}}$ , there exists at least one codeword ${\displaystyle c_{x}\in C}$ such that the Hamming distance ${\displaystyle d(x,c_{x})}$ between ${\displaystyle x}$ and ${\displaystyle c_{x}}$ satisfies

${\displaystyle d(x,c_{x})\leqslant d-1}$

since otherwise we could add x to the code whilst maintaining the code's minimum Hamming distance ${\displaystyle d}$ – a contradiction on the maximality of ${\displaystyle |C|}$.

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

${\displaystyle \mathbb {F} _{q}^{n}=\bigcup _{c\in C}B(c,d-1).}$

Now each ball has size

${\displaystyle \sum _{j=0}^{d-1}{\binom {n}{j}}(q-1)^{j}}$

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

${\displaystyle q^{n}=\left|\mathbb {F} _{q}^{n}\right|=\left|\bigcup _{c\in C}B(c,d-1)\right|\leqslant \sum _{c\in C}|B(c,d-1)|=|C|\sum _{j=0}^{d-1}{\binom {n}{j}}(q-1)^{j}}$

That is:

${\displaystyle A_{q}(n,d)=|C|\geqslant {\frac {q^{n}}{\sum _{j=0}^{d-1}{\binom {n}{j}}(q-1)^{j}}}.}$

An improvement in the prime power case

For q a prime power, one can improve the bound to ${\displaystyle A_{q}(n,d)\geq q^{k}}$ where k is the greatest integer for which

${\displaystyle q^{k}<{\frac {q^{n}}{\sum _{j=0}^{d-2}{\binom {n-1}{j}}(q-1)^{j}}}.}$

See also

• Singleton bound
• Hamming bound
• Johnson bound
• Plotkin bound
• Griesmer bound
• Grey–Rankin bound
• Gilbert–Varshamov bound for linear codes
• Elias-Bassalygo bound

References

1. Gilbert, E. N. (1952), "A comparison of signalling alphabets", Bell System Technical Journal, 31: 504–522, doi:10.1002/j.1538-7305.1952.tb01393.x.
2. Varshamov, R. R. (1957), "Estimate of the number of signals in error correcting codes", Dokl. Akad. Nauk SSSR, 117: 739–741.