Back to results
US 6,370,191 B1Granted
Efficient implementation of error approximation in blind equalization of data communications
AssigneeTexas Instruments Incorporated
InventorsShivaling S. Mahant-Shetti, Srinath Hosur, Alan Gatherer
Life of this patent
Estimating the term…
Abstract
A cable modem (20) having a blind equalizer (40) in its equalization function (28) is disclosed. The blind equalizer (40) includes an adaptive equalizer (44) and an approximating update function (42) that provides updated equalization coefficients to the adaptive equalizer (44) using the Constant Modulus Algorithm, wherein the error in the adaptive equalizer output is estimated. The estimates are based upon a determination of the maximum and minimum of the real and i components of symbols output by the adaptive equalizer (44). Efficiency in the computations required for updating the equalizer coefficients is obtained, without sacrificing convergence.
Claims14
A method of setting equalization coefficients for data signals received over a communications channel, comprising the steps of:
- receiving a data symbol from the communications channel;
- applying an equalizer filter to the received data symbol to produce an output symbol having real and imaginary components;
- generating an error estimate for the output symbol by performing the steps of:
- determining the minimum and maximum of the absolute value of the real and imaginary components of the output symbol;
- generating a subtrahend using the determined minimum and maximum;
- and
- subtracting the subtrahend from a modulus value to generate an error estimate;
- updating an equalization coefficient for the equalizer filter from a combination of the received data symbol, the error estimate, and the output symbol;
- and
- repeating the receiving, applying, and updating steps for a plurality of received data symbols.
7 more elements
A method of setting equalization coefficients for data signals received over a communications channel, comprising the steps of:
- receiving a data symbol from the communications channel;
- applying an equalizer filter to the received data symbol to produce an output symbol having real and imaginary components;
- generating an error estimate for the output symbol by performing the steps of:
- determining the minimum and maximum of the absolute value of the real and imaginary components of the output symbol;
- generating a subtrahend by:
- dividing the determined maximum by four;
- generating a difference operand corresponding to the determined minimum minus the divided determined maximum;
- responsive to the difference operand being less than zero, generating a subtrahend corresponding to the determined maximum;
- and
- responsive to the difference operand being greater than or equal to zero, performing the operations of:
- dividing the difference operand by two;
- generating a subtrahend corresponding to a sum of the divided difference operand and the determined maximum;
- and
- subtracting the subtrahend from a modulus value to generate an error estimate;
- updating an equalization coefficient for the equalizer filter from a combination of the received data symbol, the error estimate, and the output symbol;
- and
- repeating the receiving, applying, and updating steps for a plurality of received data symbols.
14 more elements
A communications device for receiving data communications, comprising:
- an interface for receiving modulated signals from a communications channel;
- a demodulator for demodulating the received modulated signals into digital data symbols;
- an equalizer for compensating the received modulated signals for distortion of the communications channel, comprising:
- a blind equalizer, comprising:
- an adaptive equalizer for applying a digital filter to the data symbols, having an input coupled to the demodulator, and having an output;
- an approximating update function, having an input coupled to the output of the adaptive equalizer for receiving output symbols having real and imaginary components therefrom, having an input coupled to the demodulator to receive data symbols therefrom, and having an output coupled to the adaptive equalizer, the approximating update function for updating an equalization coefficient of the adaptive equalizer, by performing the operations of:
- generating an error estimate for the output symbol by performing the steps of:
- determining the minimum and maximum of the absolute value of the real and imaginary components of the output symbol;
- generating a subtrahend using the determined minimum and maximum;
- and
- subtracting the subtrahend from a modulus value to generate an error estimate;
- generating an updated equalization coefficient from a combination of the received data symbol, the error estimate, and the output symbol;
- and
- repeating the receiving, applying, and updating steps for a plurality of received data symbols;
- and
- a decision block having an input coupled to the output of the adaptive equalizer, for generating a detected symbol from the output symbol.
13 more elements
A communications device for receiving data communications, comprising:
- an interface for receiving modulated signals from a communications channel;
- a demodulator for demodulating the received modulated signals into digital data symbols;
- an equalizer for compensating the received modulated signals for distortion of the communications channel, comprising:
- a blind equalizer, comprising:
- an adaptive equalizer for applying a digital filter to the data symbols, having an input coupled to the demodulator, and having an output;
- an approximating update function, having an input coupled to the output of the adaptive equalizer for receiving output symbols having real and imaginary components therefrom, having an input coupled to the demodulator to receive data symbols therefrom, and having an output coupled to the adaptive equalizer, the approximating update function for updating an equalization coefficient of the adaptive equalizer, by performing the operations of:
- generating an error estimate for the output symbol by performing the steps of:
- determining the minimum and maximum of the absolute value of the real and imaginary components of the output symbol;
- generating a subtrahend by dividing the determined minimum by two;
- adding the determined maximum and divided determined minimum;
- raising the sum of the determined maximum and divided determined minimum to a selected power to produce the subtrahend;
- dividing the determined maximum by four;
- generating a difference operand corresponding to the determined minimum minus the divided determined maximum;
- responsive to the difference operand being less than zero, generating a subtrahend corresponding to the determined maximum;
- and
- responsive to the difference operand being greater than or equal to zero, performing the operations of:
- dividing the difference operand by two;
- and
- generating a subtrahend corresponding to a sum of the divided difference operand and the determined maximum;
- and
- subtracting the subtrahend from a modulus value to generate an error estimate;
- generating an updated equalization coefficient from a combination of the received data symbol, the error estimate, and the output symbol;
- and
- repeating the receiving, applying, and updating steps for a plurality of received data symbols;
- and
- a decision block having an input coupled to the output of the adaptive equalizer, for generating a detected symbol from the output symbol;
- and
- a decision feedback equalizer, having an input coupled to the demodulator.
25 more elements
Dependencies are inferred from the claim text.