Abstract: Technologies for jitter extraction are described. A receiver device includes an analog-to-digital converter (ADC) and a signal processing circuit. The signal processing circuit includes an equalizer block to output current data based on samples from the ADC. A clock-recovery (CR) block includes a timing error detector (TED) or a phase detector to measure a sampling offset. The CR block can use the sampling offset to control sampling of subsequent data by the ADC. A jitter extraction block can use the sampling offset to re-sample the current data to obtain re-sampled data based on the sampling offset to remove jitter from the current data.

Abstract: Disclosed are apparatuses, systems, and techniques for deploying and training machine learning models for fast and efficient equalization of signals transmitted over communication channels. In one embodiment, the techniques include processing, using first model(s), a digital representation of a received (RX), via a communication channel, signal to obtain channel loss metrics representative of a difference between the RX signal and a transmitted (TX) signal. The techniques further include obtaining a first set of equalization (EQ) parameter(s), and iteratively obtaining a second set of EQ parameter(s). The techniques further include configuring, using the second set of the EQ parameters, one or more EQ circuits to equalize at least one of the RX signal, the TX signal, or a channel signal.

Abstract: Technologies for jitter extraction are described. A receiver device includes an analog-to-digital converter (ADC) and a signal processing circuit. The signal processing circuit includes an equalizer block to output current data based on samples from the ADC. A clock-recovery (CR) block includes a timing error detector (TED) or a phase detector to measure a sampling offset. The CR block can use the sampling offset to control sampling of subsequent data by the ADC. A jitter extraction block can use the sampling offset to re-sample the current data to obtain re-sampled data based on the sampling offset to remove jitter from the current data.

Abstract: A receiver device includes circuitry to measure an error vector of a pulse amplitude modulation (PAM) sequence in a signal received from a transmitter and control logic coupled to the circuitry. The control logic removes estimated linear components from the measured error vector to generate a non-linear error vector. The control logic further determines, with reference to a set of lookup table (LUT) values, one or more tuning parameters for the PAM sequence based on the non-linear error vector and modifies the set of LUT values according to the one or more tuning parameters. The control logic further provides the modified set of LUT values to the transmitter, which when used by the transmitter to add digital pre-distortion to the PAM sequence, causes the non-linear error to be at least partially removed from the signal.

Abstract: A receiver device includes circuitry to measure an error vector of a pulse amplitude modulation (PAM) sequence in a signal received from a transmitter and control logic coupled to the circuitry. The control logic removes estimated linear components from the measured error vector to generate a non-linear error vector. The control logic further determines, with reference to a set of lookup table (LUT) values, one or more tuning parameters for the PAM sequence based on the non-linear error vector and modifies the set of LUT values according to the one or more tuning parameters. The control logic further provides the modified set of LUT values to the transmitter, which when used by the transmitter to add digital pre-distortion to the PAM sequence, causes the non-linear error to be at least partially removed from the signal.

Abstract: Systems, methods, and devices for performing predictive baseline wander correction are described. A digital signal may be demodulated to obtain an estimated transmitted symbol stream, based on which an amount of baseline wander error may be predicted. The predicted amount of baseline wander error may be used to correct for baseline wander.

Abstract: Technologies for jitter extraction are described. A receiver device includes an analog-to-digital converter (ADC) and a signal processing circuit. The signal processing circuit includes an equalizer block to output current data based on samples from the ADC. A clock-recovery (CR) block includes a timing error detector (TED) or a phase detector to measure a sampling offset. The CR block can use the sampling offset to control sampling of subsequent data by the ADC. A jitter extraction block can use the sampling offset to re-sample the current data to obtain re-sampled data based on the sampling offset to remove jitter from the current data.

Abstract: Embodiments are disclosed for a sequence selective symbol checker for communication systems. An example method includes configuring a symbol checker of a receiver with first binary sequence data generated by a symbol generator of the receiver. The example method also includes comparing, using the symbol checker, second binary sequence data provided by a transmitter to the first binary sequence data to generate error count data related to a number of errors for symbols associated with the second binary sequence data. The example method also includes determining total count data related to a number of symbols associated with the first binary sequence data. The example method also includes determining error ratio data associated with the transmitter based on the error count data and the total count data.

Abstract: Embodiments are disclosed for a sequence selective symbol checker for communication systems. An example method includes configuring a symbol checker of a receiver with first binary sequence data generated by a symbol generator of the receiver. The example method also includes comparing, using the symbol checker, second binary sequence data provided by a transmitter to the first binary sequence data to generate error count data related to a number of errors for symbols associated with the second binary sequence data. The example method also includes determining total count data related to a number of symbols associated with the first binary sequence data. The example method also includes determining error ratio data associated with the transmitter based on the error count data and the total count data.

Abstract: Embodiments are disclosed for full-rate phase detection for a pulse amplitude modulation N (PAM-N) signal. The example method includes sampling an incoming signal in one or more sampling times. The example method further includes determining that an amplitude associated with a current sampling time is within an upper threshold and a lower threshold for each sampling time of the one or more sampling times. The example method further includes upon determining that the amplitude of the current sampling time is within the upper threshold and the lower threshold, determining an amplitude range associated with an immediately preceding sampling time and an amplitude range associated with an immediately subsequent sampling time. The example method further includes determining a transition status representing one of an upward transition, a downward transition, or no transition with respect to the current sampling time.

Abstract: Embodiments are disclosed for tuning a continuous time linear equalizer embedded in a receiver of a communication system. An example method includes receiving an N-point estimation of a channel pulse response associated with an input signal. The method further includes calculating an estimated power spectral density for one or more desired frequency bands. The method further includes updating one or more parameters of a continuous time linear equalizer to adjust one or more power densities of the one or more desired frequency bands based on the one or more estimated power spectral densities calculated. The method further includes calculating, for each desired frequency band, an estimated power spectral density after updating the one or more parameters of the continuous time linear equalizer to adjust the one or more power densities of the one or more desired frequency bands.

Abstract: Embodiments are disclosed for generating a lookup table value for calibrating a lookup table circuit in an optical transmitter. The example method includes generating a calibration signal. The calibration signal encodes a plurality of bits in a number of amplitude levels. The example method further includes transmitting the calibration signal to a module under calibration and transmitting a symbol sequence of defined length to a reference module. The reference module compares the symbol sequence with a distorted signal received from the module under calibration to generate a set of condition count statistics. The example method further includes receiving the set of condition count statistics from the receiver in the reference module and calculating a lookup table value based on the set of condition count statistics. The example method further includes transmitting the lookup table value to a transmitter associated with the module under calibration.

Abstract: Embodiments are disclosed for channel estimation in a receiver of a communication system. An example method includes receiving, via a receiver of a communication system, an input signal. The example method further includes using a first event indicator embedded in an analog circuit of the receiver to slice the input signal to generate a sliced input signal and applying an offset to the input signal to generate an offsetted signal. The example method further includes using a second event indicator embedded in the analog circuit to slice the offsetted signal to generate a sliced offsetted signal. The example method further includes applying a first predefined delay to the sliced input signal and applying a second predefined delay to the sliced offsetted signal. The example method further includes generating a conditional ones signal based on the sliced input signal and the sliced offsetted signal and using the conditional ones signal to calibrate an equalizer embedded in the receiver.

Abstract: Embodiments are disclosed for generating a lookup table value for calibrating a lookup table circuit in an optical transmitter. The example method includes generating a calibration signal. The calibration signal encodes a plurality of bits in a number of amplitude levels. The example method further includes transmitting the calibration signal to a module under calibration and transmitting a symbol sequence of defined length to a reference module. The reference module compares the symbol sequence with a distorted signal received from the module under calibration to generate a set of condition count statistics. The example method further includes receiving the set of condition count statistics from the receiver in the reference module and calculating a lookup table value based on the set of condition count statistics. The example method further includes transmitting the lookup table value to a transmitter associated with the module under calibration.

Abstract: Embodiments are disclosed for equalizing input signals for communication systems. An example method includes receiving an input signal. The input signal encodes a plurality of bits in a number of amplitude levels. The example method further includes converting the input signal to an equalized output signal using a plurality of lookup table circuits. The equalized output signal encodes a plurality of symbols in a number of amplitude levels. The example method further includes feeding the equalized output signal to an output driver circuit.

Abstract: Embodiments are disclosed for equalizing input signals for communication systems. An example method includes receiving an input signal. The input signal encodes a plurality of bits in a number of amplitude levels. The example method further includes converting the input signal to an equalized output signal using a plurality of lookup table circuits. The equalized output signal encodes a plurality of symbols in a number of amplitude levels. The example method further includes feeding the equalized output signal to an output driver circuit.