Parity error correction for band-limited digital signals
First Claim
1. A method of correcting and replacing an erroneous digital source signal sample containing K bits which are compressed to N bits, and added with a parity bit considering parity of M of the N bits which are transmitted and sent to a receiver, wherein N is less than or equal to K, and M is less than or equal to N and wherein signal samples received in the receiver are bit-expanded back to K bit samples as applicable, wherein a parity check is used for ascertaining whether the digital source signal sample received by the receiver is an erroneous signal sample, the method comprising the steps of:
- choosing M most significant bits from expanded samples of K bits in each said erroneous signal sample and producing M inverted test samples each with K bits, wherein each of said M inverted test samples is obtained by inverting a single one of the M bits and keeping all other (K−
1) bits unchanged;
passing each of said M inverted test samples through a selected filter and obtaining a filtered output in each case, and for each of said M inverted test samples measuring a difference in value between each inverted test sample and its corresponding filtered output; and
,selecting that value of the difference which is minimum, and choosing a corresponding inverted test sample of K bits for replacing the erroneous signal.
1 Assignment
0 Petitions
Accused Products
Abstract
An error correction method corrects and replaces erroneous digital signal samples (having N companded bits) in a receiver after ascertaining by parity check that a sample is erroneous. The method chooses M MSBs where M is less than or equal to N, and produces M test samples, each of the M test samples being obtained by inverting a single bit from the M bits, keeping other bits unaltered. Each test sample is expanded and passed through a selected low pass filter (e.g., 15 kHz) to obtain a filtered output and a differential value between the test sample and its filtered output. The test sample producing the least differential value is chosen to replace the erroneous signal sample. The technique is especially applicable in NICAM demodulators receiving 14 bit sample signals (at 32 kHz) companded to (N) 10 bits from which (M) 6 MSB parity encoded bits are chosen for producing test samples.
6 Citations
25 Claims
-
1. A method of correcting and replacing an erroneous digital source signal sample containing K bits which are compressed to N bits, and added with a parity bit considering parity of M of the N bits which are transmitted and sent to a receiver, wherein N is less than or equal to K, and M is less than or equal to N and wherein signal samples received in the receiver are bit-expanded back to K bit samples as applicable, wherein a parity check is used for ascertaining whether the digital source signal sample received by the receiver is an erroneous signal sample, the method comprising the steps of:
-
choosing M most significant bits from expanded samples of K bits in each said erroneous signal sample and producing M inverted test samples each with K bits, wherein each of said M inverted test samples is obtained by inverting a single one of the M bits and keeping all other (K−
1) bits unchanged;passing each of said M inverted test samples through a selected filter and obtaining a filtered output in each case, and for each of said M inverted test samples measuring a difference in value between each inverted test sample and its corresponding filtered output; and
,selecting that value of the difference which is minimum, and choosing a corresponding inverted test sample of K bits for replacing the erroneous signal. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 23)
-
-
11. An error correction technique for correcting and replacing an erroneous digital signal source sample sent by a transmitter to a receiver after ascertaining by parity check that the digital signal source sample received is erroneous, wherein the digital signal source sample has N bits, comprising the steps of:
-
choosing M parity encoded bits which are used by the transmitter for parity, from said N bits in each said erroneous sample and producing M number of inverted test samples each with N bits, M being less than or equal to N, wherein each of said M inverted test samples is obtained by inverting a single one of the M bits and keeping all other N−
1 bits unchanged;passing each of said M inverted test samples through a selected filter and obtaining a filtered outputs in each case, and for each of said M inverted test samples measuring a difference in value between each inverted test sample and its corresponding filtered output; selecting that value of the difference which is minimum, and choosing a corresponding inverted test sample with N bits for replacing the erroneous signal sample. - View Dependent Claims (12, 13, 14, 15, 16, 17, 18, 19, 20, 24)
-
-
21. A method for correcting and replacing an erroneous digital signal sample from a band limited signal source received in a receiver after ascertaining by parity check that the digital signal sample is erroneous, wherein the digital signal sample has 14 bits compressed to N bits and expanded back to 14 bits as applicable, and wherein M bits are considered for parity check, N being equal to less than 14 and M being equal to or less than N, said method being implemented for error correction in a NICAM (Near Instantaneous Companded Audio Multiplexed) demodulator, and comprising the steps of:
-
choosing M parity encoded bits from expanded samples of 14 bits in each said erroneous sample and producing M number of inverted test samples each with 14 bits, wherein each of said M inverted test samples is obtained by inverting a single one of the M bits and keeping all other remaining 13 bits unchanged; passing each of said M inverted test samples through a selected filter and obtaining a filtered output in each case, and for each of said M inverted test samples measuring a difference in value between each inverted test sample and its corresponding filtered output; selecting that value of the difference which is minimum, and choosing a corresponding inverted test sample with 14 bits for replacing the erroneous signal sample. - View Dependent Claims (22, 25)
-
Specification