Redirected from Chinese Remainder Theorem
The original form of the theorem, contained in a book by the Chinese mathematician Ch'in Chiu-Shao[?] published in 1247, is a statement about simultaneous congruences (see modular arithmetic). Suppose n1, ..., nk are positive integers which are pairwise coprime (meaning gcd(ni, nj) = 1 whenever i ≠ j). Then, for any given integers a1, ..., ak, there exists an integer x solving the system of simultaneous congruences
A solution x can be found as follows. For each i, the integers ni and n/ni are coprime, and using the extended Euclidean algorithm we can find integers r and s such that r ni + s n/ni = 1. If we set ei = s n/ni, then we have
For example, consider the problem of finding an integer x such that
Note that some systems of the form (1) can be solved even if the numbers ni are not pairwise coprime. The precise criterion is as follows: a solution x exists if and only if ai ≡ aj (mod gcd(ni, nj)) for all i and j. All solutions x are congruent modulo the least common multiple of the ni.
For a principal ideal domain R the Chinese remainder theorem takes the following form: If u1, ..., uk are elements of R which are pairwise coprime, und u denotes the product u1...uk, then the ring R/uR and the product ring R/u1R x ... x R/ukR are isomorphic via the isomorphism
f : R/uR --> R/u1R x ... x R/ukR x mod uR |-> ( (x mod u1R), ..., (x mod ukR) )
The inverse isomorphism can be constructed as follows. For each i, the elements ui and u/ui are coprime, and therefore there exist elements r and s in R with r ui + s u/ui = 1. Set ei = s u/ui. Then the map
g : R/u1R x ... x R/ukR --> R/uR ( (a1 mod u1R), ..., (ak mod ukR) ) |-> ∑i=1..k ai ei mod uR
One of the most general forms of the Chinese remainder theorem can be formulated for rings and (two-sided) ideals. If R is a ring and I1, ..., Ik are ideals of R which are pairwise coprime (meaning that Ii + Ij = R whenever i ≠ j), then the product I of these ideals is equal to their intersection, and the ring R/I is isomorphic to the product ring R/I1 x R/I2 x ... x R/Ik via the isomorphism
f : R/I --> R/I1 x ... x R/Ik x mod I |-> ( (x mod I1), ..., (x mod Ik) )
Search Encyclopedia
|
Featured Article
|