Bell Telephone Laboratories

Hayes Microcomputer Products, Inc.

At&t Bell Laboratories

John A Copeland, Iii

Stewart E Miller

John W Jerrim

Scott C Swanson

Tien P Lee

Robert H Krambeck

David F Strawn

Martin H Sauser, Jr

Andrew G Dentai

Richard J Burton

Larry D Becker

Toby L Nixon

Robert E Weatherford, Jr

A data flow controller (150) for monitoring and automatically controlling the flow of serial data from a remote transmitter to a host device. A serial communications card (11) provides an interface between a remote transmitter connected to a serial port connector (20) and a host device (10). The card (11) contains a UART (14) which has a buffer. A counter (151) counts the number of bytes received by the UART (14) since the last time that the host (10) read all the data in the buffer. If the number exceeds a predetermined portion of the buffer capacity the counter output (Q11) will go high, thereby disabling the counter and sending a control signal (DTR, RTS) to the remote transmitter to stop sending data. Once the host (10) has read all the data in the buffer the UART (14) provides a signal (-RXREADY) which resets the counter (151), thereby causing the output (Q11) to go low, thereby allowing the remote transmitter to resume sending data. The counter (151) is held in a reset state until the buffer has received at least one byte of data. Data flow control is therefore obtained without the intervention of the host (10).

Method and apparatus for serial data flow control

A method for communicating with a plurality of hosts connected to an X.25 network and for selectably using a character as a control character or a data character. Each character received from the X.25 network is inspected to determine if it is one of a set of reserved characters. If so, the character is encoded before being sent to its final destination. Data intended for the X.25 network is inspected for the presence of a predetermined control character. If the control character is present, the next character is inspected to determine whether it is a command character or an encoded data character. If this next character is an encoded data character, it is decoded and provided to the X.25 network. If this next character is a command character, then the command is executed. The method allows the use of any character as either a command character or data character. Data transfer between the DTE and DCE is assigned a PAD and a virtual channel number which corresponds to a particular host on the X.25 network. The channel number may, but need not, be the same as the X.25 device address for that host. The method therefore allows the DTE and DCE to simultaneously communicate with a plurality of X.25 hosts over a single telephone line or physical transmission link.

Method for sending a plurality of data channels over a single

A method and apparatus for compressing data, particularly useful in a modem. The preferred method is implemented in a microprocessor within a modem, and dynamically adapts to changing data statistics. Parallel encoding and decoding tables are provided at the encoder and the decoder, and are updated for each character processed. Each table has a plurality of digital compression codes associated with characters of an alphabet. In response to an item of data presented for encoding, a compression code which corresponds to the character presented for encoding is selected using the encoding table. The selected compression code is provided as an output. Periodically, the association between the codes and the characters of the alphabet in the table is adjusted as a function of the frequency of occurrence of characters of the alphabet, over a plurality of characters. As the frequency of occurrence of characters presented for encoding changes, the more frequently occurring characters become associated with the shorter codes in the encoding table. By performing the same steps in the decoding table, the compression codes are used at the decoder to select a character of the alphabet, which is provided as a decoded output.

Adaptive data compression method and apparatus

The preferred embodiment (11) of the modem comprises a control unit (13), a memory (27), a switch (14), and a modem engine (17). The modem engine (17) establishes communications with a modem (30) using conventional handshake methods. The control unit (13) then initiates a special handshake sequence composed of nonprintable, opposing characters to the modem (30) via the switch (14) and the modem engine (17). If the modem (30) completes the special handshake sequence then the control unit (13) and the modem (30) exchange the desired information. If the modem (30) does not complete the special handshake sequence then the preferred embodiment (11) functions as a conventional modem. The use of nonprinting, opposing characters for the special handshake sequence prevents the special handshake sequence from adversely affecting the external devices (10) (34).

Modem with improved handshaking capability

A dual wavelength optical source includes a monolithic integrated pair of series-opposition connected diodes, each fabricated for emitting light having a different wavelength than the other. Polarity of a common bias current conducted through the diodes is reversed for activating alternatively light emission from the diodes. Active emission regions of the diodes are stacked upon one another for efficiently coupling the resulting beams of the different wavelengths of emitted light through a surface into the core of a single mode optical fiber.

Dual wavelength optical source

Two-dimensional semiconductor chips are stacked to form a three-dimensional array in which coupling between chips is effected optically. This permits the use of smaller chips, with a corresponding higher yield, and serves to reduce the chip area required for interconnecting leads. It also reduces the internal interconnection path lengths which, at present, limit the speed of operation.

Optically coupled integrated circuit array

Single mode lasers are advantageously stabilized in order to prevent a side mode from growing to a level where partition noise effects degrade high speed performance. To obtain such spectral stability, the powers in the two adjacent side modes are measured, and the operating temperature of the laser varied in response to changes in their relative powers. Optimally, the gain curve peak is made to coincide with the preferred cavity mode.

Spectrally stabilized laser

An electro-optic integrated circuit is disclosed wherein the long electrical connections normally present on a large scale integrated circuit are replaced by an optical waveguide layer. A plurality of epitaxial layers are grown on a single substrate and at least three of the plurality of epitaxial layers are grown with bandgaps that are suitable for optical sources, detectors and waveguiding. These primary layers are separated from each other by a barrier layer having a bandgap greater than either of the adjacent primary layers. Two of the layers adjacent to the substrate are grown to accommodate electrical devices that can be used to couple electrical signals to the optical source layers and to amplify electrical signals provided by the optical detection layer.

Multilayered optical integrated circuit

Large current gains and high degrees of sensitivity to impinging primary photons are realized in photon feedback photodetectors embodying the invention. A photocurrent generated by an internal photodiode (10, 11) in response to the primary photons (6) causes secondary photons to be emitted by internal serially connected luminescence diodes (12, 13; 14, 15). Secondary photons traveling away from the photodiode are redirected by a reflector (16) to impinge on the photodiode and thereby sustain the photocurrent. Gains of the order of 100 are realized by these photodetectors.

High sensitivity photon feedback photodetectors

Longitudinal mode control is achieved in a heterojunction semiconductor laser (201-208) by doping the active region (203) of the laser with a deep level electron or hole trap. The trap is chosen to have a carrier capture cross section .sigma..sub.e and an optical cross section .sigma..sub.o such that the ratio of P, the average number of photons per cubic centimeter, to P.sub.s is between 0.1 and 100 where P.sub.s is equal to (N.sigma..sub.e V/.sigma..sub.o C.sub.o), N is the carrier density, V is the carrier thermal velocity, and C.sub.o is the speed of light in the material. In a specific embodiment the active region is bombarded by photons to achieve deep level electron traps in the active region.

Mode stabilized semiconductor laser

A semiconductor laser is disclosed wherein the active region has been doped with deep-level electron traps either by proton bombarding the active region or by doping with an impurity, such as oxygen, iron, or chromium. The density of traps is such that an optical absorption parameter of greater than 30 cm.sup.-1 is achieved. This laser, when combined with an ordinary photodiode, exhibits overall optical gain thereby permitting an array of optical logic circuits.

Trap doped laser combined with photodetector

In a digital communication system using a switching device as a detector-discriminator, information about the amplitude of the input signal is lost. As a result, there is no way to adjust the threshold of the switch as a function of signal strength to achieve minimum error rate. This problem is resolved by transmitting, along with the data, reference pulses whose amplitude is intermediate that of the data '0' and '1'. Automatic threshold level control in such a system is provided by means of a coincidence circuit (13), which gates out any of the reference pulses that are detected by the detector-discriminator switch (11), and an integrating circuit (14) which averages the detected pulses over a period of time. The output from the integrator is used to bias the switch in a manner to optimize the threshold level.

Discriminator threshold level control circuit for a digital transmission

An input voltage overload protection semiconductor structure useful with MOS circuitry consists of a p-region in an n-substrate with p+ type regions formed on both sides of the p-region and an n+ type region centrally located in the p-region. Input signals are applied to the first p+ region. The gate of an MOS structure to be protected from voltage overload is connected to the second p+ type region. A power supply used with the MOS structure is connected to the n+ region. This structure provides significantly greater load protection than the standard resistor-diode-resistor circuit.

Semiconductor overload protection structure

A light-activated light-emitting device has at least one p-n junction provided with electrodes for confining light-emission to an area of the junction. It has been determined that light-emission can be activated by light impinging on the junction outside this confined area, so two optical fibers are provided, one being an input fiber for bringing activating light to the nonemitting sensitive part of the junction and the other fiber being an output fiber coupled to the light-emitting area. When the device is a p-n-p-n light-activated light-emitting switch provided with an RCL reset control circuit, a very inexpensive unidirectional optical pulse regenerator is obtained. The device in its various forms is advantageously suited for use in each of many stations along optical fiber data busses or in optical logic arrays because the unidirectional feature prevents light feedback between adjacent devices and consequently avoids spurious switching of a preceding device.

Unidirectional optical device and regenerator

This invention furnishes a semiconductor shift register circuit characterized in that each stage of the register contains a dynamic 'master' portion feeding a static 'slave' portion. The register circuit is driven by a suitable clock voltage circuit such that if the signal source for the clock circuit should stop, then the information in each stage is retained and is stored in the static portion. In this way, the overall performance of the shift register is static in the sense that the information in the register is not lost during temporary stoppage of the driving clock; yet, fewer electrical component elements are required per entire stage than in the case of a fully static stage.

Growing community of inventors

Fair Haven, NJ, United States of America

John A Copeland, Iii