Abstract: A system and method for communicating information using circular multidimensional signal space constellations using uncoded modulation allows increased data rate and bandwidth by eliminating trellis encoding, thus reducing the number of computational cycles required for coding. Furthermore, the use of circular signal space constellations lowers the transmit peak factor, thus consuming less power. Using uncoded modulation with circular constellations allows an approximate 10-20% performance improvement for a given central processor unit. Additionally, circular constellations allow the easy and reliable transmission of special marker symbols such as "End of File", "Start of Message", "End of Transmission" or "Increase or Decrease Data Rate".

Abstract: An improvement to a half duplex multipoint communication environment enables the full duplex communication of information between a control modem and any one of a plurality of remote modems at any given time. This is accomplished by providing a system for echo cancellation in a multipoint communication environment whereby a first modem having a first echo canceler for canceling the echo on a transmission line and a plurality of additional modems each having an echo canceler are connected to the same copper wire pair. The presence of an echo canceler in each modem connected to the transmission path enables the full duplex exchange of information between the control modem and one of the remote modems at any given time.

Abstract: An improvement to a half duplex multipoint communication environment wherein a digital subscriber line (DSL) receiver employs an adaptive comb filter and a decision feedback equalizer (DFE) to efficiently suppress or eliminate low frequency periodic impulse noise and crosstalk generated in the vicinity of the device receiver.

Abstract: In a system that uses a dither signal in the production of a transmitted signal, the recoverability of an original trellis code is maintained while forming the dither signal using a modulo value that is equal to the distance between two adjacent symbols. This is accomplished by forming individual modulo counts for each of the orthogonal components produced by the transmitter's 3-tap FIR filter. The modulo counts and the bits from the trellis encoder are used to substitute the constellation subset identified by the trellis encoder with another constellation subset. The substituted subset is used for transmission and results in recovery of the original trellis code by the trellis decoder in the receiver.

Abstract: A communications system simultaneously transmits both a primary data signal and a secondary voice signal in such a way that the dynamic range of the secondary voice signal is increased even though deliberate errors are introduced into the transmitted data portion of the signal. In particular, the communications system includes an error correcting technique like channel encoding and the constellation signal space is divided into a number of regions, where at least one of the regions overlaps with another region. The primary signal is channel encoded to select a particular one of the number of regions, the region being represented by a reference signal point value. The voice signal is encoded to provide a signal point vector, which is added to the reference signal point in such a way that the resultant signal point is located in the overlapping region thereby deliberately introducing an error into the resulting transmitted signal point stream.

Abstract: The present invention discloses an improvement over existing techniques for transmitting data over voice-band telephone channels by automatically adapting the amount of warping or compression that is applied to a sequence of signal points. A sequence of warped signal points, each of which is related to a respective signal point of a predetermined base constellation according to a warp function, is received in a decoder via a transmission channel having a non-linear component. After each of the received signal points is unwarped using substantially an inverse of the warp function, the average dispersions of the received inner and outer signal points about corresponding sequences of expected signal points are calculated. The difference between the average dispersion of the inner and outer points is then computed and used to update the amount of warping or compression in order to further minimize the effects of the channel non-linearity. The desired amount of warping is communicated to the transmitting encoder.

Abstract: A method and apparatus for optimizing parameters selectable by an analog modem for operation on any bandwidth-constrained channel is provided. Through the addition of circuitry for performing an optimization process, the modem is configured for identifying the optimum bandwidth and maximum data rate for use over a selected bandwidth-constrained channel. In the process of optimizing the parameters for operation on the selected bandwidth-constrained channel, a measurement of the signal and noise spectrum present on the selected channel is determined. And from this measurement, the optimum bandwidth, maximum data rate for the transmission of data over the selected bandwidth-constrained channel and a modulus for performing a modulus conversion of the data for transmission are provided. The modem is then trained to operate using these optimized parameters for transmission of the data.

Abstract: A technique for use in communications systems utilizing busy and idle modes. In the busy mode, symbols representative of data are transmitted to system end users. In the idle mode, during which no data is communicated to end users, idle mode symbols are transmitted. Such idle mode symbols are not used to represent data and are only transmitted during the idle mode. Advantageously, such idle mode symbols maintain adaptive apparatus, such as AGC circuits, NEXT cancellers and equalizers, in their properly converged state. In the disclosed embodiment, the idle mode symbols are the innermost symbols in a signal constellation and transition from the idle to busy mode is represented by the transmission of a predetermined number of outermost symbols in the signal constellation. Use of such symbols for idle-to-busy mode transitions advantageously provides ready detection of such transitions. In addition, use of the innermost symbols for the idle mode reduces spectral emission and crosstalk.

Abstract: The relative amounts of additive white noise and proportional noise in a communication channel are compensated for by providing adaptive precoding and preemphasis to signal point signals. The amounts of precoding and preemphasis are controlled using scalars to modify the magnitudes of the precoding and preemphasis signals that are combined with the signal point signals. The values of the scalars are transferred between the transmitter and receiver, and the values can be changed without losing synchronization between the transmitter and receiver.

Abstract: In order to mitigate against the multiplicative noise effects caused by known non-linearities in a trellis coded data transmission system, a signal constellation is formed by starting with a base constellation whose number of signal points and whose geometry are selected in accordance with conventional criteria and then warping that constellation by adjusting the positions of its signal points in accordance with a warp function which is the inverse of the known component of the non-linear characteristic of the transmission system. Because the constellation warping is deterministic, it is possible for the receiver to "unwarp" the received signal points prior to applying them to the Viterbi decoder. In preferred embodiments, the trellis code is of a type in which the dominant error event is a trellis path error.

Abstract: The effect of non-linear distortion on a received signal point is reduced by the use of a circular limiter. In particular, the circular limiter computes the distance of the received signal point from the center of the signal point constellation. If this distance exceeds a desired limit, then both dimensions of the received signal point are reduced by the same factor to maintain the original phase. This reduction in the dimensions of the received signal point decreases the corresponding error signal that is calculated by a Viterbi decoder. As a result, this further reduces the value of the accumulated path metrics of the Viterbi decoder which results in an improvement in system performance. In addition, the circular limiter also realizes a slight improvement in system performance against additive white Gaussian noise. The value of the limiting distance can either be predefined, or determined adaptively to match channel conditions of the communications system.

Abstract: A modem which represents the transmitted primary data stream using signal points selected via a trellis code is provided with the further capability of transmitting secondary data. In particular, a standard constellation of signal points is expanded to include a number of signal points which is greater than that needed to transmit the primary data. Particular primary data word values are representable by either of a predetermined pair of signal points associated with that value--one signal point of the pair being a standard signal point and the other being an expanded signal point. The specific one of the two signal points of the pair that is transmitted at any given time is chosen as a function of whether the value of the next bit in the secondary data stream waiting to be transmitted is a "0" or a "1".

Abstract: Fractional rate modulation is accomplished by separating incoming data into frames of bits. Each frame in partitioned into bit words of unequal lengths. The words are divided by a modulus to obtain remainders. The remainders are transmitted using, for example, QAM modulation. At the receiving end, the process is reversed. The ratio of baud rate of the system to the incoming bit rate defines the modulus as well as the number of QAM modulation points. A "data rate throttling" technique takes advantage of the capabilities provided by the foregoing to allow the rate at which data is channel mapped and transmitted over a communication channel to be easily varied in small increments.

Abstract: In order to mitigate against the multiplicative noise effects caused by known non-linearities in a trellis coded data transmission system, a signal constellation is formed by starting with a base constellation whose number of signal points and whose geometry are selected in accordance with conventional criteria and then warping that constellation by adjusting the positions of its signal points in accordance with a warp function which is the inverse of the known component of the non-linear characteristic of the transmission system. Because the constellation warping is deterministic, it is possible for the receiver to "unwarp" the received signal points prior to applying them to the Viterbi decoder.

Abstract: Viterbi decoder performance in a data communication system using 2N-dimensional channel symbols N>1 can be further enhanced by an interleaving technique which uses a distributed trellis encoder in combination with a signal point interleaver.

Abstract: Fractional rate modulation is accomplished by separating incoming data into frames of bits. Each frame is partitioned into bit words of unequal lengths. The words are divided by a modulus to obtain remainders. The remainders are transmitted using for example QAM modulation. At the receiving end the process is reversed. The ratio of the baud rate of the system to the incoming bit rate defines the modulus as well as the number of QAM modulation points.

Abstract: In a full-duplex modem, correlation of an echo estimate signal with an echo-canceled signal is used to determine the presence or absence of a remote signal.

Abstract: A modem which represents the transmitted primary data stream using signal points selected via a trellis code is provided with the further capability of transmitting secondary data. In particular, a standard constellation of signal points is expanded to include a number of signal points which is greater than that needed to transmit the primary data. Particular primary data word values are representable by either of a predetermined pair of signal points associated with that value--one signal point of the pair being a standard signal point and the other being an expanded signal point. The specific one of the two signal points of the pair that is transmitted at any given time is chosen as a function of whether the value of the next bit in the secondary data stream waiting to be transmitted is a "0" or "1".

Abstract: Fractional rate modulation conversion is accomplished by separating incoming data into frames of bits. Each frame is partitioned into bit words of unequal bit lengths. The words are divided by a modulus to obtain remainders. The remainders are then multiplexed into sequential bauds of common modulus. The apparatus can be used for data compression as well as efficient modulation in bandwidth restricted channels.

Abstract: A method and system for communicating data over a communications channel such as a telephone line includes the conversion of asynchronous data transmitted from a diagnostic microcomputer through a modem into synchronous data for transmission over the communications channel. The method includes the use of error vectors to shift the asynchronous data into a continuous synchronous stream of data bits.