Abstract: A method of fabricating optical preforms that yield optical fibers having minimal scattering centers by substantially eliminating bubbles and crystallites that form during preform fabrication. The method utilizes glasses having low viscosity in their liquid state to eliminate large bubbles that contain ambient gasses. Small bubbles that have sizes comparable to the wavelength of light are eliminated by simultaneously vacuum-pumping and slow-cooling the liquid glass used to fabricate the preform before pouring the glass into the preform cast mold. Finally, bubbles due to the formation of vacuums in the preform are eliminated by applying a temperature gradient across the cast mold to induce a steep vertical gradient in viscosity in the liquid glass when the liquid glass is poured in the mold.

Abstract: A smart pixel array is disclosed wherein a single optical beam can be used to transmit information to and from each subarea or cell of the pixel array. An optical P-I-N diode with an intrinsic MQW region is present in each subarea, and the optical beam is oriented to impinge on the diode. A first transmission gate couples information to be transmitted to the diode during a transmit interval, and information received by the diode is coupled through a second transmission gate to a receiver circuit during a receive interval. A third transmission gate couples a reset potential to the diode during spacing intervals between the transmit and receive intervals.

Abstract: An integrated circuit including a circuit for determining shift overflow in a binary digital circuit having an n-bit shift data and an m-bit shift amount. The device has a logic array for producing an n-bit output from the m-bit shift amount; a conversion circuit for selectively converting the sign of an n-bit shift data; and a combination of OR and AND logical gates for logically combining the selectively converted n-bit shift data and the n-bit output producing an overflow output.

Abstract: In a digital filter having tap coefficients, a gain element is employed to scale the filter output. The gain element is controlled by an error monitor element which runs an adaptive process in accordance with the invention. Such a process causes each tap coefficient value to be changed so as to reduce power consumption in the filter. On the other hand, the process ensures that the filter output maintains an acceptable signal to noise ratio (SNR), despite losing the bit precision of the filter as a result of the change of the coefficient values.

Abstract: A power factor correction system with an EMI line filter at the input includes circuitry to sense the capacitor current of the EMI filter to improve the accuracy of the power factor enhancement. The circuitry consists of a current sensing means connected after the EMI filter, and a voltage sensing capacitor which derives its input signal from a pair of diodes separate from the conventional 4-diode rectifier bridge, which are connected either before or after the EMI filter. The current-sensing means, which may be a resistor, and the voltage-sensing capacitor are tightly coupled to the current control loop of the power factor correction system to compensate for the current in the EMI filter capacitors.

Abstract: A method and apparatus for reducing the output bit rate in a video object planes sequence encoder. An approximation of the shape of an object on a current frame is generated and used, along with a predicted image of the object from a subsequent frame and an estimate of the displacement of the object, to predict an image of the object in a current frame. A stripe is formed around the predicted image and the picture elements contained in the stripe are assigned texture values. The predicted image and stripe are subtracted from the current frame to form an error signal which is masked by the approximate shape of the object and provided to an entropy coder for transmission at an output bit rate.

Abstract: A joint data (features) and channel (bias) estimation framework for robust processing of speech received over a channel is described. A trellis encoded vector quantizer is used as a pre-processor to estimate the channel bias using blind maximum likelihood sequence estimation. Sequential constraint in the feature vector sequence of a speech signal is applied for the selection of the quantized signal constellation and for the decoding process in joint data and channel estimation. A two state trellis encoded vector quantizer is designed for signal bias removal applications.

Abstract: An accessory connector and adapted assembly (30) that utilizes a connection to the reference voltage line (19') of the interface (19) between a personal computer (10) and its keyboard or other accessory (14), thereby avoiding the need to provide an independent power supply circuit for peripheral devices and accessories such as a microphone (20), speakers or the like. The adapter assembly (30) is configured with a first interface (28) having, by way of illustrative example, a male connector portion (28A) dimensioned for insertion into the keyboard port or PS2 port (19) of a personal computer and a female connector portion (28B) dimensioned to accommodate the male termination (18) of a conventional keyboard cable (17). The adapter assembly further includes a second interface (26) for providing power to the one or more external peripheral devices.

Abstract: A MOSFET amplifier device is biased to exhibit a substantially constant transconductance over a range of variations in power supply, temperature and process. A bias circuit includes a pair of MOS field effect devices, one of which is biased in a triode operating region with a constant reference drain-to-source terminal voltage and with a constant first reference drain-to-source terminal current. The other field effect device is biased in saturation by a circuit that derives a gate-to-source terminal bias voltage from the gate terminal voltage of the first device and the first reference voltage. A second reference current flows into the drain terminal of the second device, and a corresponding bias current is derived, for example, by mirroring the second reference current. The mirrored reference current is used to bias a MOS amplifier device, to maintain a substantially constant transconductance over the mentioned range of variations.

Abstract: The invention is a speech recognition system and method for transmitting information including the receipt and decoding of speech information such as that modeled by hidden Markov models (HMMs). In this invention, the state likelihoods of the modeled state sequences contained within the speech information are assigned penalties based on the difference between those state likelihoods and a maximum possible state likelihood. Once penalties have been assigned, the modified state sequence with the modified state likelihoods having the highest cumulative state likelihoods is used in further speech recognition processing. In this manner, state sequences having no extremely poor state likelihoods are favored over those having both extremely high and extremely poor state likelihoods.

Abstract: A history-based prefetch cache which includes a time queue. The time queue correlates past events with cache misses in a microprocessor. The time queue is set to N cycles, N being a predetermined, arbitrary or programmable amount. The prefetch cache is a prefetch target buffer which receives inputs from a time queue and a cache and determines if an event is present in the cache. If an address is not present in the cache it is prefetched based on past events and inserted into the prefetch target buffer so that the microprocessor will not miss it the next time.

Abstract: A method and device for increasing the bandwidth of a drop line in a power line carrier communication system for facilitating high speed transmission of communication data. The drop line is divided into two sections, each containing both high frequency communication data components and low frequency power components. A high pass filter is provided for passing the high frequency components to a receiver device and a low pass filter is provided for passing the low frequency power components to an electrical system in a residence. The low pass filter is also used for blocking, from the drop line, high frequency noise generated by electrical appliances operating from the electrical system.

Abstract: The specification describes techniques for the manufacture of optical gratings in optical fibers. The grating pattern is written into the core of the fiber without removing the fiber coating. Coating compositions with high transparency to the actinic (writing) radiation but which are UV curable are described in detail. The coating compositions contain a UV photoinitiator that absorbs sufficient UV radiation to effectively cure the polymer but is relatively transparent to UV radiation used for writing the grating. The photoinitiator is one or more compounds selected from a specified group of aliphatic and cycloaliphatic ketones.

Abstract: A closure for optical fiber cables has a cover member and an end plate having an O-ring seal for sealing the end plate to the cover member by means of over-center latches. The inner surface of the end plate has a backbone member to which are mounted a fiber organizer and splice tray, a cable sheath grip, and a central strength member anchor for use with cables having central strength members. The end plate has openings for grommets through which cables pass into the interior of the closure, and a grommet and port plug retaining member fits over the outer ends of the grommet or grommets or grommet and port plug. The backbone member, which may be integral with the end plate, has a wall and flange portion that bears against the inner end of the grommet or grommets toward the closure interior.

Abstract: An equalizer/decoder for a communication channel presenting impairments to error-free reception of transmitted symbols includes a modified Viterbi decoder operating on samples of received signals in a plurality of equivalent subchannels. Estimates of each of the subchannels are updated using a locally best estimate in the Viterbi trellis processing, thereby avoiding delay in updating channel estimates in a rapidly changing channel such as a digital cellular communications channels. A leaky predictor channel updating technique also proves advantageous in an alternative embodiment.

Abstract: A ROM (read only memory) is disclosed. For via-ROMs, an isolation transistor is used to isolate adjacent pairs of memory devices instead of the more conventional field oxide isolation. The gate of the isolation transistor is grounded, insuring that conduction does not take place. For a GASAD ROM, a field oxide isolation is used.

Abstract: Noise reducing circuits for electronic receiving instruments, such as telephone receivers in headsets or handsets that are used in noisy locations, provide compensation of the set's receiver unit-to-error microphone transfer function to enhance the noise reduction. Further circuits provide pre-conditioning of the incoming voice signal to make the noise cancellation more effective. The tendency of these noise cancelling circuits to oscillate is substantially lessened by added circuitry which rapidly detects onset of oscillation and momentarily reduces the noise cancellation without interrupting the incoming speech path altogether.

Abstract: A user of a telephone terminal with an interactive display is provided complete access to all feature choices in a hierarchical menu through a minimum number of menu keys. These menu keys include a "Next" key, a "Back" key, a "Select" key and a "Home" key. These menu keys occupy minimal space on the housing of the telephone terminal, yet enable the user of the terminal to easily scroll to and select every feature in the interactive display. The user is able to quickly develop a mental model of the feature choices and easily navigate through these menu selections. Context-sensitive functionality also is provided for using the interactive display. As a user completes a command function, the feature choice presented to the user for this function is the one most likely to be accessed by the user, thus advantageously avoiding the necessity of having a dedicated key for allowing the user to select this feature choice.

Abstract: Pseudorandom Boundary-Scan testing of a interconnects within a system (10) is accomplished by first generating a pseudorandom pattern and then filtering the pattern before application to the system to modify values within each pattern that could cause a signal conflict. Each response generated by the system upon application to the system is also filtered to mask non-deterministic values with the response. The filtered response is compared to a set of signatures representing a fault-free condition to detect faults within the system.

Abstract: A bipolar transistor with MOS-controlled protection for a reverse-biased emitter-base junction is disclosed. A bipolar transistor and a MOS transistor are configured with the drain and the gate electrically coupled to the emitter, and the source and body electrically coupled to the base. A reverse-bias at the emitter-base junction, which is less than a breakdown voltage for the emitter-base junction, activates the MOS transistor which substantially reduces the resistance between the emitter and the base. Preferably, a first semiconductor region provides both the drain and the emitter, and a second semiconductor region provides both the body and the base, for reduced surface area on an integrated circuit chip.