Error correction and loss recovery of packets over a computer network
US6000053A · kind A · utility
Assignee
Inventors
Key dates
| Filing date | Jun 13, 1997 |
| Grant date | Dec 7, 1999 |
| Priority date | — |
| Expiry date | Jun 13, 2017 |
Classification
- Technology area (CPC H)Electricity
- CPC primaryH03M13/19
- WIPO fieldBasic communication processes
- WIPO sectorElectrical engineering
Abstract
A method of recovering a lost or corrupted data packet of a plurality of data packets transmitted by a transmitting computer system to a target computer system over an unreliable computer network. In one embodiment, the transmitting computer system generates parity information by padding the plurality of data packets to the length of the longest data packet with a suitable bit pattern. The parity information is then generated using a suitable algorithm, such as an exclusive OR (XOR) operation. The parity information and the packet lengths of the data packets are used to form the parity packet. The transmitting computer system then sends both the data packets and the parity packet to the target computer. If the target computer system detects a lost or corrupted data packet, the target computer attempts to reconstruct the lost or corrupted data packet. The received uncorrupted data packets are padded with the suitable bit pattern described above. Note that it is useful to pad up to the length of the lost data packet. The inverse operation of the parity generating algorithm is then applied to the padded data packets and the parity packet to reconstruct the lost or corrupted data packe…
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.