Message Framing and Error Control

Abstract

Two problems which arise in the transmission of digital data are message framing and error control. A technique has been developed whereby polynomial coding can be used simultaneously for both error detection and message framing. With this technique, encoded data blocks can be transmitted asynchronously without incorporating any special signals or bit patterns for "start-of-message" and "end-of-message" indications. Erroneously received blocks are automatically deleted from the accepted data stream. The standard systems use a special signal for a start-of-message (SOM) indication at the beginning of each block of data, and independently create and detect a specified format (usually based on parity check sums) to determine when errors have occurred in transmission of the data. In order to achieve high reliability it is necessary to use sophisticated techniques and to devote a large number of transmitted digits to each of these functions. By considering these two problems together a technique which provides a solution to both was found. This technique uses fewer bits than would be required by many conventional approaches to framing to get the same reliability. The error-detection capability which is included is so strong that the mean time between undetected errors is usually dependent on equipment reliability rather than on channel error statistics.