Efficient dispersal of information for security, load balancing, and fault tolerance
Journal of the ACM (JACM)
ISCA '89 Proceedings of the 16th annual international symposium on Computer architecture
Polynomial and matrix computations (vol. 1): fundamental algorithms
Polynomial and matrix computations (vol. 1): fundamental algorithms
Fast algorithms with preprocessing for matrix-vector multiplication problems
Journal of Complexity
Practical loss-resilient codes
STOC '97 Proceedings of the twenty-ninth annual ACM symposium on Theory of computing
A displacement approach to efficient decoding of algebraic-geometric codes
STOC '99 Proceedings of the thirty-first annual ACM symposium on Theory of computing
The complexity of the matrix eigenproblem
STOC '99 Proceedings of the thirty-first annual ACM symposium on Theory of computing
Nearly optimal computations with structured matrices
SODA '00 Proceedings of the eleventh annual ACM-SIAM symposium on Discrete algorithms
Multivariate polynomials, duality, and structured matrices
Journal of Complexity
Fast reliable algorithms for matrices with structure
Fast reliable algorithms for matrices with structure
Reed-Solomon Codes and Their Applications
Reed-Solomon Codes and Their Applications
Linear time erasure codes with nearly optimal recovery
FOCS '95 Proceedings of the 36th Annual Symposium on Foundations of Computer Science
FOCS '98 Proceedings of the 39th Annual Symposium on Foundations of Computer Science
Hi-index | 0.00 |
We exploit various matrix structures to decrease the running time and memory space of the known practical deterministic schemes for erasure-resilient encoding/decoding. Polynomial interpolation and multipoint evaluation enable both encoding and decoding in nearly linear time but the overhead constants are large (particularly, for interpolation), and more straightforward quadratic time algorithms prevail in practice. We propose faster algorithms. At the encoding stage, we decrease the running time per information packet from C log2 r, for a large constant C, or from r (for practical encoding) to log r. For decoding, our improvement is by the factors C and N/log N, respectively, for the input of size N. Our computations do not involve polynomial interpolation. Multipoint polynomial evaluation is either also avoided or is confined to decoding.