Coded data generation or conversion – Digital code to digital code converters – To or from minimum d.c. level codes
Patent
1997-04-21
1998-09-01
Williams, Howard L.
Coded data generation or conversion
Digital code to digital code converters
To or from minimum d.c. level codes
371 434, H03M 1300
Patent
active
058016490
DESCRIPTION:
BRIEF SUMMARY
This application claims the priority benefit of U.S. provisional application Ser. No. 60/001,986, filed Aug. 3, 1995.
BACKGROUND OF THE INVENTION
This invention relates to encoding digital data for transmission through partial response channels. More particularly, the present invention relates to implementing a 12/15 matched spectral null encoder.
Information that is passed from a source to a destination can, in general, be described as passing through a channel. Ideally, the channel should transmit the information without modifying it, but in reality, all channels modify information or signals that pass through them. The modifications vary from channel to channel, but in general can be categorized into one or more types. One type of modification, known as a partial response, creates a channel output signal that is a combination of the original input signal and the original input signal delayed by one or more time units. Such partial responses are often described in terms of a transfer polynomial, h(D)=(1.+-.D.sup.N), where h(D) represents the channel's output, "1" represents the original signal and D.sup.N represents the original signal delayed by N times units. For instance, many computer disc drives have channels that exhibit a class-IV partial response, also known as a PR(4) response, which has a transfer function h(D) of (1-D.sup.2) that produces an output signal that is equal to the input signal minus the input signal delayed by two time units.
Partial response transfer functions, like the one above have associated spectral nulls which are frequencies that the transfer function does not transmit. For instance, a channel acting as a PR(4) channel at a data rate X, will not transmit the constituent frequency X/2, known as the Nyquist frequency, nor the frequency zero, known as D.C.
The spectral nulls of a channel are of interest because it has been shown that the reliability of data recovery at the destination can be improved if the channel's input signal is encoded so that the spectral power density of the encoded signal drops to zero at the spectral nulls of the channel. Such coded signals are known as matched spectral null codes, and their properties have been described in detail in such works as R. Karabed and P. H. Siegel's, "Matched Spectral-Null Codes for Partial-Response Channels," IEEE Trans. Info. Th., vol. 37, No. 3, pp. 81-855, May 1991.
Of particular interest are matched spectral null codes with spectral power densities that match the spectral nulls of a dicode channel (1-D). These codes may be used to create two separate encoded signals which may be interleaved to create an encoded signal with spectral power densities that match the spectral nulls of a PR(4) (1-D.sup.2) channel.
In a 12/15 matched spectral null code, an input signal is encoded by converting twelve bit segments of the input signal into fifteen bit codewords. Since there are 2.sup.12 (4096) possible input values, which each require a valid and unique codeword, implementing an encoder to translate 12-bit input values into 15-bit codewords could be accomplished by using a single table of 4096 entries, with each entry containing a 15-bit codeword addressable by a 12-bit value. However, to implement a table of that size would require a large number of components, and in nearly every application of such an encoder, it is desirable to have as few components as possible.
EP-A-0097763 discloses the division of an input value to two encoders which provide preset values which are subsequently concatenated to provide an output value. A disparity control chooses the complement of the output value depending upon the disparity value in order to reduce run lengths.
EP-A-0503863 relates to lexicographically encoding and decoding codes, and discloses a set of coding rules to permit a trellis code to be efficiently employed in a NRZ format.
SUMMARY OF THE INVENTION
In a first aspect the present invention provides a method of encoding an input value as an encoded value, the method comprising: derived from the input value to at least two
REFERENCES:
patent: 5073776 (1991-12-01), Shigemori
patent: 5497384 (1996-03-01), Frederickson et al.
patent: 5537424 (1996-07-01), Karabed et al.
patent: 5608397 (1997-03-01), Soljanin
patent: 5731768 (1998-03-01), Tsang
K. J. Knudson et al., A Concatenated Decoding Scheme for (1-D) Partial Response with Matched Spectral-Null Coding, Proceedings of 1993 IEEE Global Telecomm. Conf. (GLOBECOM '93), pp. 1960-1964 (Nov. 1993).
Gu, Jian et al., A New Approach to Constructing Optimal Block Codes for Runlength-Limited Channels, 8097 IEEE ransactions on Information Theory, 40(1994) May, No. 3., pp. 774-785.
Seagate Technology Inc.
Williams Howard L.
LandOfFree
Matched spectral null encoder/decoder does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Matched spectral null encoder/decoder, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Matched spectral null encoder/decoder will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-274080