University of Texas System

National Instruments Corporation

Motorola Corporation

Freescale Semiconductor,inc.

Takao Inoue

James Wesley McCoy

Robert Wendell Heath, Jr

Matthew A Pendleton

Peter R Molnar

Jaeweon Kim

Milos Milosevic

Techniques are disclosed relating to use of digital predistortion in the context of full-duplex radio. In some embodiments, an apparatus includes one or more antennas and is configured to simultaneously transmit and receive wireless signals via at least partially overlapping frequency resources using the one or more antennas. In some embodiments, the apparatus includes receive chain circuitry that is configured to process both wireless signals transmitted by the apparatus via the one or more antennas and over-the-air wireless signals from one or more other computing devices. In some embodiments, the apparatus includes one or more processing elements configured to determine one or more digital predistortion parameters based on the wireless signals transmitted by the apparatus via the one or more antennas and processed by the receive chain circuitry and apply predistortion to transmitted wireless signals based on the one or more digital predistortion parameters.

Digital predistortion for full-duplex radio

Techniques for transmitting data over a communications channel are generally described. A value of a property of an inverse of a channel matrix corresponding to the communications channel may be calculated and compared to a threshold value. If the value of the property a first one of greater than or less than the threshold value, at least one transmit message may be altered using a vector perturbation technique to generate data symbols and the data symbols may be precoded using a channel inversion technique to generate precoded data symbols. If the value of the property is the other of greater than or less than the threshold value, the at least one transmit message may be precoded using a channel inversion technique to generate precoded data symbols.

Transmitting data over a communications channel

Techniques are generally described here for communicating channel state information using predictive vector quantization. In some examples, a method may include measuring channel state information based, at least in part, on signals received over a communications channel. An error vector may be calculated between the measured channel state information and predicted channel state information. The error vector may be quantized, and subsequent channel state information may be predicted based, at least in part, on the quantized error vector.

Communicating channel state information using predictive vector quantization

A simple fast and robust echo canceller for both synchronous and asynchronous multicarrier transceiver systems. A first residual time domain echo component is separated from a receive signal in a first frame and a second residual time domain echo component is separated from a receive signal in the next consecutive frame. The first and second residual time domain echo components from consecutive frames are combined to obtain a combined residual time domain echo component. The combined residual time domain echo component is used to adaptively update coefficients in a transfer function representing an estimate of the echo channel in the multicarrier transceiver system. The separation of the echo components from the receive signal eliminates dependence on the receive signal so that convergence is substantially faster and not signal dependent. Performance of the echo canceller is virtually independent of the receive signal and allows reliable tracking of changes in the echo channel over time.

Method and apparatus for echo cancellation updates in a multicarrier transceiver system

An echo canceler ( ) includes a summing device ( ) that subtracts a correction signal from a received signal, the difference of which represents the far-end signal with an error component. Instead of adapting its coefficients using the output of the summing device ( ), the echo canceler ( ) uses the difference between the input and output of a decision device ( ) as an estimate of the error component alone. The estimate of the error component is then used to adapt the coefficients according to the adaptive least mean squares (LMS) algorithm. In one embodiment, the decision device ( ) forms discrete multi-tone symbols based on the equalized output of the summing device. In this embodiment, the echo canceler ( ) performs an inverse of the equalization step efficiently by replacing a division operation with a multiply operation and a corresponding power-of-two shift operation.

Growing community of inventors

Austin, TX, United States of America

Takao Inoue