Title :
Testing Reed-Muller codes
Author :
Alon, Noga ; Kaufman, Tali ; Krivelevich, Michael ; Litsyn, Simon ; Ron, Dana
Author_Institution :
Dept. of Math., Tel-Aviv Univ., Israel
Abstract :
A code is locally testable if there is a way to indicate with high probability that a vector is far enough from any codeword by accessing only a very small number of the vector´s bits. We show that the Reed-Muller codes of constant order are locally testable. Specifically, we describe an efficient randomized algorithm to test if a given vector of length n=2m is a word in the rth-order Reed-Muller code R(r,m) of length n=2m. For a given integer r≥1, and real ε>0, the algorithm queries the input vector υ at O(1/ε+r22r) positions. On the one hand, if υ is at distance at least εn from the closest codeword, then the algorithm discovers it with probability at least 2/3. On the other hand, if υ is a codeword, then it always passes the test. Our result is almost tight: any algorithm for testing R(r,m) must perform Ω(1/ε+2r) queries.
Keywords :
Reed-Muller codes; binary codes; polynomials; probability; query processing; randomised algorithms; testing; Reed-Muller code; affine subspace; binary field; codeword; multivariate polynomials; probability; property testing; randomized algorithm; Codes; Decoding; Mathematics; Performance evaluation; Testing; Affine subspaces; Reed–Muller code; binary field; multivariate polynomials; property testing;
Journal_Title :
Information Theory, IEEE Transactions on
DOI :
10.1109/TIT.2005.856958