Abstract: Aspects of efficient signal compression format schemes are described. In various embodiments, samples are grouped into one or more blocks. One or more scaling factors are determined for the samples in the one or more blocks. In some cases, a block scaling factor is determined for each block. Further, in some embodiments, if a unique scaling factor exists among samples in a block, the position of the sample is identified. In turn, sign extension bits are scaled from the one or more blocks, and samples in the one or more blocks are quantized. In one embodiment, based on a number of sign extension bits scaled from the block and whether the unique scaling factor exists, a scaling code word is encoded for the block. In other embodiments, implicity, adaptive scaling, and/or adaptive quantization, for example, are used to increase the performance of the compression schemes.

Abstract: This invention discloses an optimum form maximum likelihood MIMO detector (SFS ML Detector) that computes the LLR of the most likely received bit sequence from the received signal y, which consist of multiple signal stream from a MIMO channel H and an estimate of H. Through 5 simple steps of QR factorization, computation of partial Euclidean distance, sorting of the partial Euclidean distance, selection of surviving hypothesis by a set of algorithmic rule and computation of the full Euclidean distance of the survivors, and computation of the maximum likelihood from the surviving full Euclidean distance, the LLR of the received bit sequence can be obtained with a significantly lower number of computation and comparison than that is implied in the theoretical form of the maximum likelihood (ML) detector. There is no loss in performance of the SFS ML detector from the theoretical ML detector.

Abstract: Aspects of efficient signal compression format schemes are described. In various embodiments, samples are grouped into one or more blocks. One or more scaling factors are determined for the samples in the one or more blocks. In some cases, a block scaling factor is determined for each block. Further, in some embodiments, if a unique scaling factor exists among samples in a block, the position of the sample is identified. In turn, sign extension bits are scaled from the one or more blocks, and samples in the one or more blocks are quantized. In one embodiment, based on a number of sign extension bits scaled from the block and whether the unique scaling factor exists, a scaling code word is encoded for the block. In other embodiments, implicity, adaptive scaling, and/or adaptive quantization, for example, are used to increase the performance of the compression schemes.

Abstract: A method used in a time domain equalizer is provided. A method comprising the steps of: providing a time domain equalizer comprising a feed forward equalizer and a feedback equalizer; and using a conjugate gradient iteration in order to calculate a set of coefficients of the time domain equalizer.

Abstract: A method for training of a non-updated decision feedback equalizer is provided. The method comprising the steps of: providing a sequence of frames adapted to be received by a receiver; provide a sequence of synchronization frames interposed between a predetermined number of frames; and using at least part of the sequence of synchronization frames to train a decision feedback equalizer (DFE), thereby speeding up system convergence or making system convergence possible.

Abstract: In a receiver of a multi-leveled variable sideband communications system, a method is provided that comprises the steps of: dividing the receiver into a real portion and a complex portion; and providing a decision feedback equalizer (DFE) processing data substantially in the real portion.

Abstract: A method used in a time domain equalizer is provided. A method comprising the steps of: providing a time domain equalizer comprising a feed forward equalizer and a feedback equalizer; and using a filter circuit or structured implementation to incorporate conjugate gradient iteration in order to calculate a set of coefficients of the time domain equalizer. Whereby matrix times vector operations is converted into filtering using the filter circuit.

Abstract: A method used in a time domain equalizer is provided. The method comprising the steps of: providing a time domain equalizer comprising; and extracting a real part of an input or a derivative of the input to the time domain equalizer and using the only real part of the input in the time domain equalizer to derive an output of the time domain equalizer.

Abstract: A method comprising the steps of: providing a known sequence comprising a plurality of data points; and curve-fitting the plurality of data points to calculate coarse frequency offset.

Abstract: A method comprising the steps of providing a slicer for slicing real values of an equalizer output; and cross correlating an equalizer input with an output of the slicer is provided.

Abstract: A decision feedback equalizer is configured to equalize an input signal to generate a recovered output signal. Linear feed-forward filter circuitry is configured to provide a linearly filtered output signal based on the input signal. Composite trellis decoder circuitry configured to process a combined signal that is based on a combination of at least the linearly feed-forward filtered output signal and on output of linear or non-linear feedback filter circuitry, in accordance with state metrics generated by processing a composite trellis diagram relative to the combined signal, to provide a trellis-decoded output signal as input to the linear or non-linear feedback filter circuitry. The composite trellis decoder circuitry is further configured to provide a particular phase output of the combined signal, based on the state metrics, as the decoded output signal.

Abstract: A decision feedback equalizer is configured to equalize an input signal to generate a recovered output signal. Linear feed-forward filter circuitry is configured to provide a linearly filtered output signal based on the input signal. Composite trellis decoder circuitry configured to process a combined signal that is based on a combination of at least the linearly feed-forward filtered output signal and on output of linear or non-linear feedback filter circuitry, in accordance with state metrics generated by processing a composite trellis diagram relative to the combined signal, to provide a trellis-decoded output signal as input to the linear or non-linear feedback filter circuitry. The composite trellis decoder circuitry is further configured to provide a particular phase output of the combined signal, based on the state metrics, as the decoded output signal.