Abstract: Neural Normalized MinSum (N-NMS) decoding for improving frame error rate (FER) performance on linear block codes over conventional normalized MinSum (NMS). Dynamic multiplicative weights are assigned to each check-to-variable message in each iteration to efficiently provide training parameters of N-NMS that support N-NMS for longer block lengths. Embodiment are described for neural two-dimensional normalized MinSum (N-2D-NMS) decoders requiring fewer training parameters. The N-2D-NMS approaches for example use the same weight for edges with the same check and/or variable node degree. Simulation results indicate that that this LDPC decoding performs similarly to previous techniques while substantially reducing the amount of training necessary.

Abstract: Low density parity check (LDPC) codes (LDPCCs) have an identical code blocklength and different code rates. At least one of the rows of a higher-rate LDPC matrix is obtained by combining a plurality of rows of a lower-rate LDPC matrix with the identical code blocklength as the higher-rate LDPC matrix.

Abstract: A method and apparatus is disclosed for use in a communication system for encoding data to achieve more efficient coding rates and high data transmit rate, or a lower bit error rate. An encoder may receive a portion of the data to be encoded and process the portion of the data to create encoded data. The encoded data and the portion of the data not sent to the encoder are presented to a transform module configured to perform matrix multiplication on the inputs with a transform matrix. The resulting transformed data is mapped, using a mapper, either via an algorithmic approach or a manual mapping approach to generating a constellation mapped output. 2-dimensional encoding may occur.

Abstract: A method for trellis constellation shaping is disclosed. In one embodiment, this method comprises receiving two or more input bits and filtering at least one of the two or more input bits to create two or more filtered output bits. The step of filtering at least one input bit introduces at least one extra bit. The method then encodes the input bits which are not filtered to create encoded bits and stores the encoded bits and the filtered bits in a buffer. A processing element may be configured to perform Viterbi type processing on the filtered bits to create processed bits. The method combines at least one of the processed bits with at least one of the buffered bits to create a combined bit set from the buffered bits and then performs mapping on the combined bet set to thereby map the combined bit set into a constellation.

Abstract: The quality of speech being transmitted in a telecommunication channel is enhanced by adaptively compensating for frequency dependent attenuation in the telecommunication channel. The frequency dependent attenuation in the telecommunication channel is dynamically determined and, then, compensated for by adaptively selecting and inserting an appropriate frequency dependent gain function in the telecommunication channel.

Abstract: A transmission arrangement is disclosed which facilitates compression of modem modulated facsimile signals for transmission on digital facilities. The presence of a facsimile call is detected and facsimile control messages are obtained by demodulating and interpreting the facsimile call control protocol. A controller is responsive to the obtained control messages for controlling a voiceband encoder to encode the control portion of the facsimile call and for controlling a so-called facsimile page demodulator to demodulate any facsimile page portions of the facsimile call to obtain a baseband digital bit stream. The presence of prescribed tone signals, i.e., echo protection tone (EPT) signals, is detected in the facsimile call. EPT control messages representative of the EPT signals are generated. The encoded control portion of the facsimile call and the baseband digital bit stream representative of the demodulated, i.e.