Bell Telephone Laboratories

At&t Bell Laboratories

American Telephone & Telegraph Co., At&t Bell Laboratories

Yau-Chau Ching

James D Tomcik

David Gavin Messerschmitt

James C Candy

Allan Michael Hofmann

Ben-Zion Gotz

Nian-Chyi Huang

Circuits at a remote transmission site receive digitized voice and data samples from incoming channels. The voice samples are compressed. The samples are formatted into fixed-size data blocks with signaling and control information for transmission to a switching node. Additional bits of the voice samples are dynamically dropped, if necessary, to fit the data into the fixed-sized blocks. The signaling and control information allows receiving circuits at the switching node to determine the number of bits allocated to each sample. These receiving circuits form packets for individual channels by attaching switch routing packet headers to the individual channel data. Packet receiving circuits on the output side of the switch reformat the received packets into fixed-size nonpacket data blocks for transmission.

Integrated voice/data network

Adaptive bit allocation to a set of spectral transform coefficients is optimized by setting a predetermined first (maximum) threshold, then allocating one bit to those coefficients above the threshold, no bits to those below. If the total number of bits has not been allocated, the threshold is lowered an 'integer' step to a second threshold, and again one bit allocated to those coefficients above the second threshold. The procedure is iterated until all the total bits are allocated or, if too many bits are allocated after the last integer step, a 'fractional' step threshold is used.

Bit allocation generator for adaptive transform coder

Known adaptive differential pulse code modulation (ADPCM) coders typically convert a nonuniform input signal to a linearized representation thereof prior to processing the signal. Usually the linearized signal requires more bits than the nonuniform signal. To mitigate the expense of apparatus for manipulating signals having many bits as well as to dissipate a scale size mistrack between the coder and a decoder, the instant ADPCM coder, responsive to an improved adaptive control signal, converts (300, 600) the nonuniform input signal to special semiuniform representation thereof which representation is processed (400, 500, 600) by the ADPCM coder. The semiuniform signal permits use of fewer bits than the prior art linearized representation. Also, the adaptive control signal is provided by a scale index generator including rounding-up and rounding-down means, for dissipating scale size mistrack. Further, the semiuniform signal permits much of the ADPCM process to be confined to relatively inexpensive read only memory (ROM) devices (300, 400, 500, 800).

Adaptive quantizer apparatus for differential coding of nonuniform

In a digital speech interpolation system, speech inactivity time is utilized to reduce the bit rate on a link between transmitter and receiver by compressing digital characters from a plurality of trunks onto a lesser plurality of channels. If the number of trunks having active speech samples exceeds the number of channels, an overload may exist. To mitigate overload, it is common to truncate one or more bits from a transmitted character. However, bit truncation leads to signal degradation. Of course, it is desirable to abate the degradation consistent with mitigating the overload. Unfortunately, the number of bits truncated from a character on a first trunk is typically more or less than the number of bits truncated from a character on a second trunk. As a result, the signal degradation may not be uniform. To obtain a more uniform signal degradation, the instant priority rotation arrangement rotates the starting point for assigning bits to a digital frame responsive to the activity status of the trunks. Specifically, an active trunk is assigned to one of a plurality of groups of trunks. The starting point is taken from a different group for each output frame. Thereby, the starting point is adaptively adjusted on a frame-by-frame basis according to which trunks are active during the frame. By so rotating the starting point among active trunks rather than among all trunks, the instant arrangement achieves a more uniform signal degradation during periods of overload than was obtained with the prior art.

Digital speech interpolation trunk priority rotator

An analog signal to be converted into a high resolution digital signal is first converted to a coarsely quantized signal by a direct feedback coder. A group of successive coarsely quantized signals are then processed by a converter which produces the high resolution digital signal. In the converter, the processing is performed by an arrangement of two accumulators. The first accumulator provides an output signal indicative of the running sum of each group of coarsely quantized signals. The second accumulator has the capability of providing an accumulation of either the negative or the positive of the output signal of the first accumulator. A negative accumulation is used for a first portion of the group and a positive accumulation is used for the second portion of the group to provide a net sum. The net sum is a triangularly weighted accumulation of the successive signals in the group which occasions self-cancelling of the quantizing noise inherent to the direct feedback coder. Each net sum is used to form a code word of several digital bits in a high accuracy signal. Utilization of a third portion between the first and second portions of the group in which the second accumulator is idle provides a trapezoidal weighting in the net sum. The latter weighting provides a further improvement in resolution and accuracy.

Analog to digital converter with noise suppression

The idle-channel noise of a delta modulator is controlled by controlling the relative step size imbalance, 0. The instantaneous magnitude of the delta modulator output signal is converted to a multibit word and a second multibit word, representing a predetermined number, N, is subtracted therefrom. The dc component of the difference word is converted to analog form, amplified and inverted, and fed back to the delta modulator input stage. As a result, .theta. .apprxeq. -N. The delta modulated signal may be converted to a pulse-code modulated (PCM) signal using a digital integrator having a leak, .beta., in which case .theta. .apprxeq. -.beta.N. At the same time, the dc component of the PCM signal is advantageously reduced substantially to zero.

Method and arrangement for controlling delta modulator idle-channel noise

A digital speech interpolation (DSI) system advantageously utilizes speech inactivity time to reduce the bit rate by compressing digital characters from a plurality of trunks onto a lesser plurality of channels. A signaling arrangement is typically employed therein to signal a receiver as to the activity of a trunk. If the number of active trunks exceeds the number of channels, an overload may exist. Known arrangements for mitigating overload typically include apparatus responsive to an activity signal for truncating one or more bits from the digital characters and for transmitting the truncated characters. Unfortunately, quantization noise is increased and digital precision decreased in such arrangements. The hereindisclosed system includes an improved automatically adaptive arrangement for advantageously incorporating nearly instantaneous companding (NIC) and priority trunk rotation in a plurality of frames, called a multiframe. Thereby, a mitigation of overload as well as a lessening of quantization noise relative to known DSI arrangements is attained.

Growing community of inventors

Morganville, NJ, United States of America

Yau-Chau Ching