Cirrus Logic Inc.

U S West, Inc.

Other

Mediaone Group, Inc.

Tensorcomm, Incorporated

Data Fusion Corporation

Marvin L Vis

David E Reed

John K Thomas

Wolfgang Kober

Christopher P Zook

Jay Neil Livingston

Aruna Bayya

William G Bliss

German S Feyh

Stephen Alan Turk

Kevin Jon McCall

Baker Perkins Lee Scott, Iii

Processing circuitry includes an analog filter and filter control circuitry and processes an analog signal having a varying data rate. The analog filter filters the analog signal based on a transfer characteristic to transfer a filtered signal. The filter control circuit processes a data clock signal corresponding to the varying data rate and a reference clock signal to generate a filter control signal to adjust the transfer characteristic in proportion to the varying data rate. The analog filter adjusts the transfer characteristic in proportion to the varying data rate in response to the filter control signal. The processing circuitry may be implemented in a tape, magnetic, or optical drive.

Filter control for varying data rate communication channels

A sampled amplitude read channel is disclosed for magnetic disk storage systems comprising an adaptive non-linear correction circuit for correcting non-linear distortions in the read signal, such as asymmetry caused by the non-linear response of a magneto-resistive (MR) read head. The analog read signal is sampled and the discrete time sample values equalized into a desired partial response prior to sequence detection. The non-linear correction circuit is inserted into the read path prior to the sequence detector and adaptively tuned by a least-mean-square (LMS) adaptation circuit. In one embodiment, the non-linear correction circuit is a discrete-time Volterra filter comprising a linear response for implementing an equalizing filter, and a non-linear response for attenuating non-linear distortions in the read signal. The filter coefficients of both the linear and non-linear sections of the Volterra filter are adaptively adjusted by the LMS adaptation circuit. In an alternative embodiment, the non-linear correction circuit operates in the analog domain, prior to the sampling device, where the cost and complexity can be minimized. The analog correction circuit implements an inverse response to that of the non-linearity in the read signal, and the response is adaptively tuned using an LMS update value computed in discrete-time for a Volterra filter, without actually implementing a Volterra filter. Further, the LMS update value for the analog correction circuit can be implemented using a simple squaring circuit.

Sampled amplitude read channel employing an adaptive non-linear correction circuit for correcting non-linear distortions in a read signal

The architecture of the present invention is premised upon an algorithm involving integration of oblique correlators and RAKE filtering to null interference from other spread spectrum signals. The oblique correlator is based on the non-orthogonal projections that are optimum for nulling structured signals such as spread spectrum signals. In one configuration, space spanned by a first signal associated with a first emitter is orthogonal to an interference space associated with one or more signals of one or more other emitters. RAKE filtering is used to rapidly steer the beam of the multi-antenna system and to mitigate the effects of multipath.

Rake receiver for spread spectrum signal demodulation

The architecture of the present invention is premised upon an algorithm involving integration of oblique correlators and RAKE filtering to null interference from other spread spectrum signals. The oblique correlator is, of course, based on the non-orthogonal projections that are optimum for nulling structured signals such as spread spectrum signals. RAKE filtering is used to rapidly steer the beam of the multi-antenna system and to mitigate the effects of multipath.

Data storage system circuitry comprises randomizer circuitry and grading circuitry. The randomizer circuitry concurrently generates multiple randomized versions of user data. The grading circuitry selects one of the multiple randomized versions of the user data for data storage. Decoding circuitry decodes the selected one of the multiple randomized versions of the user data. In some examples of the invention, the grading circuitry selects the one of the multiple randomized versions of the user data based on zero run characteristics of the multiple randomized versions of the user data. In some examples of the invention, the user data is not encoded with a run length limit algorithm.

Data randomization in a data storage system

A sampled amplitude read channel is disclosed for reading a data sector recorded on a disk storage medium by detecting an estimated data sequence from a sequence of read signal sample values generated by sampling an analog read signal emanating from a read head positioned over the disk storage medium, the data sector comprising a user data segment and an embedded field comprising a known data segment. A sampling device samples the analog read signal to generate the read signal sample values and a trellis sequence detector detects a preliminary sequence from the read signal sample values. A post processor detects and corrects errors in the preliminary sequence using a syndrome generator responsive to the preliminary sequence representing the user data segment. The syndrome generator generates an error syndrome according to a predetermined error detection code. A boundary error compensator compensates for a boundary error event spanning the user data segment and the known data segment. A guided error corrector, responsive to the error syndrome, corrects errors detected in the preliminary sequence.

Sampled amplitude read channel employing a post processor with a boundary error compensator which compensates for boundary error events in a split-field data sector

A sampled amplitude read channel is disclosed for writing data to and reading data from a disk storage medium. A first channel encoder encodes a first j-k bits of a j-bit data block to generate first encoded data, and an ECC encoder encodes the first encoded data and a remaining k-bits of the data block to generate ECC redundancy symbols comprising a plurality of bits. A second channel encoder encodes the remaining k-bits of the data block and the ECC redundancy symbols to generate second encoded data. The first encoded data and the second encoded data are then output as channel data written to the disk storage medium.

Multiple-rate channel ENDEC in a commuted read/write channel for disk storage systems

Data storage system circuitry comprises randomizer circuitry and run-length control circuitry. The randomizer circuitry generates randomized user data. The run-length control circuitry grades the randomized user data to determine if the randomized user data requires run-length encoding. The run-length control circuitry applies the run-length encoding to the randomized user data if the randomized user data requires run-length encoding. The run-length control circuitry may encode decoding information and error correction information for the decoding information into the randomized user data. The run-length encoding may comprise de-randomization or RLL encoding. If de-randomization is used, RLL encoding may be eliminated altogether for the randomized user data.

Data encoding to provide run-length control in a data storage system

A sampled amplitude read channel is disclosed for disc storage systems that extracts early-decisions from a discrete-time trellis sequence detector to generate estimated target values for use in decision-directed timing recovery, gain control, and adaptive equalization. The trellis sequence detector comprises a metric generator for generating error metrics corresponding to a plurality of states of a state transition diagram, and a plurality of path memories which correspond to the paths of a trellis. The path memories store a plurality of survivor sequences which eventually merge into a most likely sequence at the output of the path memories. To reduce the latency in generating the estimated target samples, the trellis sequence detector outputs an early-decision from an intermediate location within the path memories. The early-decision is then converted into the partial response signaling space of the read signal samples. To improve the accuracy in estimating the target sample values, the accumulated metrics of a predetermined number of states are compared and the early-decision value is selected from the path memory having the smallest error metric. Alternatively, a majority-vote circuit evaluates the intermediate values stored in a predetermined number of the path memories and outputs the intermediate value that occurs most frequently. Although the early-decision technique of the present invention requires more latency than a simple slicer circuit, during acquisition the estimated target sample values are not used and therefore the increase in latency is not a significant problem.

Sampled amplitude read channel employing early-decisions from a trellis sequence detector for sampling value estimation

A speech signal divided into frames, each frame having a sound type, and a class is determined for each frame depending on the sound type of the frame. One of multiple filters is selected for each frame depending on the class of the frame. Each frame is filtered according to the filter selected, and the filtered frames combined to provide a filtered speech signal. The system includes filters and software.

Method and system for region based filtering of speech

A method and system for adaptively filtering a speech signal. The method includes decomposing the signal into subbands, which may include performing a discrete Fourier transform on the signal to provide approximately orthogonal components. The method also includes determining a speech quality indicator for each subband, which may include estimating a signal-to-noise ratio for each subband. The method also includes selecting a filter for filtering each subband depending on the speech quality indicator, which may include estimating parameters for the filter based on a clean speech signal. The method further includes determining an overall average error for the filtered subbands, which may include calculating a mean-squared error. The method still further includes identifying at least one filtered subband which, if excluded from the filtered speech signal, would reduce the overall average error determined, and combining, with exception of the filtered subbands identified, the filtered subbands to provide an estimated filtered speech signal. The system includes filters and software for performing the method.

Growing community of inventors

Longmont, CO, United States of America

Marvin L Vis