Abstract: A chess variant includes (a) one of an 8 by 8 board, a 8 by 10 board, or a 10 by 10 board; (b) a conventional chess piece set; and (c) a novel non-conventional chess piece set, wherein the pieces of the non-conventional chess piece set perform non-standard movements with rules associated therewith to govern how the pieces move and capture upon the board. Additionally, the set of rules governing how the rules of movement for the piece can undergo a change when a non-conventional chess piece reaches the furthest rank, akin to the process of promotion in standard chess.

Abstract: A chess variant includes (a) one of an 8 by 8 board, a 8 by 10 board, or a 10 by 10 board; (b) a conventional chess piece set; and (c) a novel non-conventional chess piece set, wherein the pieces of the non-conventional chess piece set perform non-standard movements with rules associated therewith to govern how the pieces move and capture upon the board. Additionally, the set of rules governing how the rules of movement for the piece can undergo a change when a non-conventional chess piece reaches the furthest rank, akin to the process of promotion in standard chess.

Abstract: Signals in a multi-channel, impaired communication system are post-processed at the receiver. A triangular matrix Decision Feedback Demodulator (DFD) at the receiver extracts channels without requiring delivery of receiver parameters to the transmitter. Multi-Input Multi-Output (MIMO) processing matrices and DFD parameters are computed by first applying matrix transformations to diagonalize the noise covariance matrix of the multiple channels received at the receiver. QR decompositions (i.e., decompositions into orthogonal and triangular matrices) are then applied to the main channels to obtain triangular channel matrices. The noise-diagonalizing transformations and QR decompositions are then combined to form the MIMO postprocessing matrices and DFD parameters. MIMO postprocessing matrices and DFD parameters are computed from training data and then adapted during live data transmission.

Abstract: A method and apparatus are disclosed for the prediction and optimization of a communications system. The present invention provides for the prediction and optimization of the performance of a communications system comprising the steps of inputting a plurality of channels, predicting a performance of each channel using a plurality of parameters to characterize the performance of the channel, and possibly optimizing the parameters of each channel according to a design criteria.

Abstract: Signals in a multi-channel, impaired communication system are post-processed at the receiver. A triangular matrix Decision Feedback Demodulator (DFD) at the receiver extracts channels without requiring delivery of receiver parameters to the transmitter. Multi-Input Multi-Output (MIMO) processing matrices and DFD parameters are computed by first applying matrix transformations to diagonalize the noise covariance matrix of the multiple channels received at the receiver. QR decompositions (i.e., decompositions into orthogonal and triangular matrices) are then applied to the main channels to obtain triangular channel matrices. The noise-diagonalizing transformations and QR decompositions are then combined to form the MIMO postprocessing matrices and DFD parameters. MIMO postprocessing matrices and DFD parameters are computed from training data and then adapted during live data transmission.

Abstract: A method that sends upstream a collection of data samples measured from a DSL line.

Abstract: Events that occur in a number of in domain communication channels where each channel is used by a communication service, are detected. A Bayesian Belief Network (BBN) defines a probabilistic cause-effect relationship between each cause and each effect on a victim channel. The probability of each of a number of possible causes as being a cause of interference in the victim channel is determined, by propagating observations of the interference backwards through the BBN.

Abstract: A method and apparatus are disclosed. The method includes one or more of the following: compiling statistical models of physical layers of a communications system; creating a priori distributions of cross-talk transfer functions; storing the models and the a priori distribution in a storage medium; and using the models and the a priori distributions to diagnose probable causes of events detected in said communications system.