Error detection/correction and fault detection/recovery – Pulse or data error handling – Digital data error correction
Reexamination Certificate
2000-07-07
2003-07-01
Tu, Christine T. (Department: 2133)
Error detection/correction and fault detection/recovery
Pulse or data error handling
Digital data error correction
Reexamination Certificate
active
06587987
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to a method and apparatus for extracting reliability information from partial response channels, and more particularly, to the extraction of soft information from partial response channels by using a dominant error event list.
BACKGROUND OF THE INVENTION
Digital transmission and recording systems convey discrete-time data sequences across channels using signals which are variable in nature. A primary goal of any such system is to convey information at the fastest possible rate with a minimum number of errors. Accordingly, numerous approaches for error control have been developed to try to minimize and/or correct errors in transmitted signals, as illustrated in U.S. Pat. No. 6,029,264 to Kobayashi et al.; and “Error Control Coding, Fundamentals and Applications,” by S. Lin and D. Costello, Jr., Pages 1-14, Prentice Hall, 1983; both of which are incorporated herein fully by reference.
A data sequence having a bit interval T and comprising, for example, a series of bits which change in amplitude (e.g., a change from a positive-amplitude bit +A to a negative-amplitude bit −A or vice versa) and/or which remain unchanged (e.g., no change in state of a data bit from one pulse to the next) with the occurrence of each bit interval T is transmitted as a sequence of analog pulses each of a certain duration equal to the bit interval T.
FIG. 1
illustrates an example of an input data sequence 1, 1, −1 (identified respectively as pulses
102
,
104
, and
106
) and the corresponding transmitted output sequence, identified by the dotted line
210
. In this example, a “1” indicates a positive-amplitude bit +A and a “−1” indicates a negative-amplitude pulse −A. As can be seen in
FIG. 1
, the pulses of the transmitted output sequence have rounded edges as compared to the square edges of the input data sequence
102
,
104
, and
106
.
Despite the somewhat imprecise correlation between the input data sequence
102
,
104
,
106
and the transmitted output data sequence
210
of
FIG. 1
, the different pulses are distinguishable from each other nonetheless, using a well-known technique referred to as “sequence detection.” Even if the signal is corrupted by noise, the detection of each bit is accomplished by sampling and storing the value of each bit at the point P in time, corresponding to the peak value X of the transmitted signal. The bit being sampled at a time P(n) is referred to herein as the “present is sample” or “present bit.” By comparing the value of the present sample with the value of the samples immediately preceding the present sample in the data sequence (e.g., the samples taken at time P(n−1), P(n−2), P(n−3), . . . P(n−M+1), where M is the channel memory, i.e., the number of previous samples that influence the present sample) it can be determined if a change in amplitude has occurred. When the present sample is being compared or is otherwise under scrutiny, it is also known as the “bit of interest.”
Although sequence detection offers a reasonable method of error control, a problem occurs when the overall data rate is increased, since the bit intervals T overlap as the total time t is shortened to accommodate the additional data to be transmitted.
FIG. 2
illustrates a situation in which the data rate for the data sequence of
FIG. 1
is increased. As shown in
FIG. 2
, a second bit
204
is transmitted before the transmission of first bit
202
has been completed. Since the bits overlap, the combined pulse, obtained by superposition of the overlapping pulses, results in a wave shape that does not allow the first pulse and the second pulse to be easily distinguished from each other. The combined pulse wave form is identified in
FIG. 2
as dotted line
210
.
This complicated wave form
210
is created because the present sample sampled at time n at the peak value X contains not only the value of pulse
204
at time n but also includes the “tail” of previous pulse
202
. This effect is known as inter-symbol interference, or ISI. Base-band communication systems transmit data at a high rate (e.g. 1000 Mb/s) in comparison to data transmission using the channel frequency band (e.g. 500 MHz). This high rate of data transmission can result in severe ISI.
If the pulses have short tails, the ISI is limited, since at some time interval (i.e., at the point at which the prior pulse has been completely transmitted) a previous pulse will no longer affect the value of the present sample. From the standpoint of detection, it is desirable that the samples of the signal tail be limited to a finite number of integer values to reduce the number of comparisons that must be made. Systems utilizing this property are called “partial response” (PR) systems. Examples of such systems are described in, for example, “Modulation and Coding for Information Storage” by P. Siegel and J. Wolf, IEEE Communications Magazine, pp. 68-89, December 1991; and “Application of Partial Response Channel Coding to Magnetic Recording Systems,” by H. Kobayashi and D. Tang, IBM J. Res. and Dev, vol. 14, pp. 368-375, July 1997, both of which are incorporated fully herein by reference.
The above-mentioned sequence detection technique is commonly used to detect the presence of ISI in a sampled signal. According to this technique, a sequence detector is employed to detect and identify transmitted (or, in the case of magnetic recording, recorded) sequences of pulses which are more likely to be incorrectly transmitted, using knowledge of noise statistics and knowledge of the channel configuration (e.g., the dependence of noise samples on previous noise samples; noise power; and/or partial response polynomials). A common noise model is called Additive White Gaussian Noise (AWGN), and a common channel model is linear time-invariant channel.
A maximum-likelihood sequence detector (MLSD) is a known sequence detector used for an uncoded, linear ISI channel (a channel which is transmitting pulses that are close together and thus having pulses which interfere with each other) with additive Gaussian noise. An example of an MLSD is described in an article entitled “Maximum Likelihood Sequence Detection in the Presence of Intersymbol Interference” by G. D. Forney (IEEE Trans. Inform. Theory, Vol. IT-25, pp. 332-335, May 1979), incorporated fully herein by reference.
A MLSD comprises a whitened matched filter (WMF), having an output that is sampled at the pulse rate of the input signal. The sampled output is input to a Viterbi detector having a trellis structure which reflects the data sequence transmitted over the ISI channel and detects the data. MLSD is not practical, particularly for communication and recording systems, because the structure of the MLSD is prohibitively complex to implement, requiring an unlimited number of states in the Viterbi detector, i.e., the hardware required to implement it is very complex.
PR systems utilize equalization to deal with ISI-related errors. Equalization is the process of reducing distortion over transmission paths by inserting compensation devices, e.g., amplifiers, in the transmission path. Placing an equalizer on the front end of a receiver receiving the transmitted data sequence allows ISI to be reduced and controlled, or even eliminated altogether. A detailed explanation of equalization is provided in the previously-cited article by P. Siegel and J. Wolf.
Complete elimination of ISI using equalization can be accomplished only in older, less sophisticated recording channel systems with low data rates, or in recording systems having low recording densities where the pulses are far enough apart to avoid ISI. In PR systems, a data sequence is reconstructed from a suitably equalized received (readback) signal, without whitened filtering, using a Viterbi detector. The noiseless equalizer output sample resulting from the use of this approach is affected by only a finite number of previous input (transmitted or recorded) samples; thus, the set of equalizer output values has set boundaries.
The wel
Lee Inkyu
Sonntag Jeffrey L.
Vasic Bane V.
Lucent Technologies - Inc.
Tu Christine T.
LandOfFree
Method and apparatus for extracting reliability information... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for extracting reliability information..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for extracting reliability information... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3067710