Method and system for providing minimal aliasing error correction code
AssigneeHongik University Industry-Academia Cooperation Foundation
InventorsSung Il Pae, Kon Woo Kwon
Life of this patent
Estimating the term…
Abstract
Disclosed is a method and system for providing a minimal aliasing error correction code. In constructing a single error correction (SEC) code by constructing a parity check matrix H for a data length k applied to a device, as the SEC code is designed to be valid and minimize generation of aliasing by checking some bits rather than all bits when nonzero binary column matrices different from each other are arranged in the parity check matrix, destruction of information can be prevented, and reliability of a device applying the SEC, such as DRAM or the like, can be improved.
Claims9
A system for providing a minimal aliasing error correction code (ECC), the system comprising:
- a memory device including a plurality of memory cells;
- and
- a processor configured to receive a message word having a length of k, generating a codeword having a parity check matrix H and to store the codeword in the memory device, and to perform error correction code decoding according to a syndrome generated using the codeword read from the memory device and the parity check matrix H, and to output a decoded data;
- an encoder to receive a message word having a length of k, and to generate a codeword having an r*n parity check matrix H including an r*k data submatrix M containing k data bits and an r*r check submatrix C containing r check bits, wherein the r*n parity check matrix H includes all nonzero 4-bit columns, and the r*r check submatrix C is an identity matrix and includes invertible nonzero column vectors;
- and
- a decoder, implementing the minimal aliasing error correction code (ECC), configured to include a correction circuit which is configured to receive the codeword and to correct at least some of error bits included in the codeword according to the syndrome generated using the codeword and the r*n parity check matrix H, wherein
- r, k, and n are natural numbers greater than or equal to 1, n=k+r, and k=2r-1.
A method of providing a minimal aliasing error correction code (ECC), the method comprising:
- receiving a message word having a length of k, by a processor;
- encoding by generating a codeword having an r*n parity check matrix H including an r*k data submatrix M containing k data bits and an r*r check submatrix C containing r check bits, by the processor, the encoding further receiving a message word having a length of k, and generating a codeword having an r*n parity check matrix H including an r*k data submatrix M containing k data bits and an r*r check submatrix C containing r check bits, wherein the r*n parity check matrix H includes all nonzero 4-bit columns, and the r*r check submatrix C is an identity matrix and includes invertible nonzero column vectors;
- and
- decoding by implementing the minimal aliasing error correction code (ECC), including a correction circuit that receives the codeword and correcting at least some of error bits included in the codeword according to a syndrome generated using the codeword and the r*n parity check matrix H, by the processor, wherein
- r, k, and n are natural numbers, n=k+r, and r satisfies k=2r-1.
Dependencies are inferred from the claim text.