Abstract: A system and method for operating a digital voice recognition processor with a non-volatile flash memory unit as the primary storage medium utilizes a wandering buffer scheme to reduce the number of memory writes performed to any portion of the flash memory unit during training and recognition operations, thereby reducing the incidence of the flash burnout effect.

Abstract: Methods and apparatus for stabilizing or otherwise controlling optical signal power levels in a multi-channel optical communication system. A multi-channel optical signal is detected in a single receiver, and the resulting detected signal is supplied to a correlator. The correlator correlates the detected multi-channel signal with a data signal associated with a particular channel to be controlled in order to generate a measure of the optical signal power of the particular channel. The measure is utilized in a feedback control loop to control an optical signal power level of the particular channel. The correlator may generate the measure of optical signal power by mixing the detected signal and the data signal in an RF mixer, by multiplying the detected signal and a data signal in the form of a unique identifier associated with the particular channel, or by using other suitable correlation techniques.

Abstract: We have invented methods of and devices for introducing unequal error protection into spread spectrum communication systems. One embodiment of the invention addresses certain inefficiencies of the standard known as IS-95 used, for example, in some mobile direct sequence code division multiple access ("DS-CDMA") phone systems. The invention allows for increased system capacity and/or improved quality. More specifically, spread spectrum multiple access ("SSMA") coding may be combined with the concept of unequal error protection ("UEP"), resulting in an UEP SSMA coding process. IS-95, in order to make significant portions of a signal highly immune to errors, has imposed the same high degree of error immunity on less significant portions of the signal (when, in fact, a lower degree would suffice) and, thus, wastes bandwidth. The present invention's coding process (i.e., UEP SSMA) utilizes bandwidth in a closer to optimal manner than known methods.

Abstract: A method for encoding handwritten symbols operates upon penstroke data received from a device capable of sampling a stylus position at discrete intervals. Each handwritten symbol is segmented into an ordered sequence of discrete strokes. An arc length and initial and final tangent angles are evaluated for each of these strokes. Each stroke is encoded in the form of a parameter set comprising position coordinates of the initial and final endpoints of the stroke, the arc length, and the initial and final tangent angles. In specific embodiments of the invention, the segmentation is based, in part, on properties of the handwritten symbol when it is expressed as a curve .phi.(s), wherein s represents arc length and .phi. represents the net turning angle.

Abstract: An apparatus for providing a clock signal having a fifty-percent duty cycle that comprises a signal input having a plurality of pulses. A delay means is provided for producing a delayed output signal having a plurality of delayed pulses, each of the plurality of delayed pulses being delayed relative to one of the plurality of pulses. A means for measuring a first delay between one of said plurality of delayed pulses and a next of the plurality of pulses is provided. A means for measuring a second delay between one of the plurality of pulses and a corresponding delayed pulse of the plurality of delayed pulses is provided. A means for comparing the first delay to the second delay is provided. A means for increasing the delay of the delay means if the first delay is greater than the second delay and decreasing the delay of the delay means if said first delay is less than the second delay is provided.

Abstract: A method and apparatus for generating and analyzing a reduced-order model of a linear circuit. The method and apparatus generates the symmetric multi-port transfer function of an RLC circuit. The invention employs a novel symmetric block-Lanczos-type procedure, termed SyMPVL for Symmetric Matrix Pade via Lanczos, to reduce original circuit matrices to a pair of banded symmetric matrices. When the circuit comprises only two of the three RLC components, the matrices are also positive definite. These matrices are typically much smaller than the original circuit matrices and determine a reduced-order model of the original multi-port transfer function of the circuit. The reduced transfer function represents a matrix-Pade approximation of the original multi-port matrix transfer function.

Abstract: A flag generating circuit that uses a feedback mechanism to set or reset a flag associated with two systems with asynchronous clocks is provided. Upon receipt of a set flag (or reset flag) signal, the circuit immediately isolates the signal after setting (or resetting) the flag to prevent race conditions between the systems. The clock associated with the setting system is synchronously started when waiting to set the flag and synchronously stopped when waiting for the flag to be reset. The clock associated with the resetting system is synchronously started when waiting to reset the flag and synchronously stopped when waiting for the flag to be set.

Abstract: A method and apparatus is presented for establishing an equivalence between the Hamming branch metric scheme and the Manhattan branch metric scheme. A signal format converter is presented to convert input signal in a first number system such as binary or two's complement, into a second number system. The signal format converter implemented at the input to a Viterbi decoder allows the use of a Viterbi decoder designed for a Manhattan metric sceme in an application where the signals are represented in two's complement. Similarly a Viterbi decoder designed for a Hamming metric scheme can be implemented in an application where the signals are represented in unsigned binary. The signal format converter can be used with both a hard decision Viterbi decoder and a soft decision Viterbi decoder.

Abstract: An apparatus and method is provided that supports continuous media for conventional networked workstations and PC's. Described are user-level mechanisms and policies designed to give good, efficient multimedia service under the mild assumption that the operating system provides a preemptive real-time scheduling class that can be used to give CPU cycles to the multimedia processes in preference to other processes that are not time sensitive. There are no modifications to the operating system kernel and isochronous networks are not required. It suffices for an application to state that it wants to play a particular stream of a type known to the server (e.g. a file containing MPEG-1 video 320.times.240 pixels, 8-bit color, 30 frames per second), or for the application to specify a frame rate and an index describing the offset of each frame in a file.

Abstract: A polarization separator/combiner for separating a light beam having two orthogonally polarized components into two parallel orthogonally polarized light beams which is especially useful as a reflective substitute for conventional walk-off devices or other birefringent separator/combiners. The light beam passes through a polarizing interface which reflects one component of the light and transmits the other, which is reflected off a parallel reflective interface and passes back through the polarizing interface to emerge parallel to the reflected component. A first embodiment of the invention comprises a flat plate having a coating on the front surface which reflects substantially all of one of the components of the light beam, while transmitting substantially all of the orthogonally polarized component.

Abstract: The present invention provides, for use in an integrated circuit structure having a prior level that includes a foundation dielectric formed over a conductive polycrystalline material, a capacitor comprising first and second electrodes having a capacitor dielectric formed therebetween. The first electrode is formed immediately over the prior level and extends beyond a common area of the first and second electrodes and connects the capacitor to the prior level outside of the common area. The capacitor is free of a direct electrical contact with the prior level; that is, the capacitor is not connected to the prior level by a window or other interconnect structure that extends directly from the capacitor itself within the common area. Electrical connection of the capacitor to the prior level is made outside the common area of the capacitor.

Abstract: A first-order differential microphone array with a fully steerable and variable response pattern. One illustrative embodiment of the present invention comprises a microphone array consisting of 6 small pressure-sensitive omnidirectional microphones flush-mounted on the surface of a 3/4" diameter rigid nylon sphere. The microphones are advantageously located on the surface at points where included octahedron vertices contact the spherical surface. By selectively combining the three Cartesian orthogonal pairs with scalar weightings, a general first-order differential microphone beam (or a plurality of beams) is realized which can be directed to any angle (or angles) in three-dimensional space. The microphone array may find use in surround sound recording/playback applications and in virtual reality audio applications.

Abstract: A slicer circuit receives a binary signal to be sliced at the midpoint of its amplitude with the portions of the sliced signal above the slicing level corresponding to a binary 1 and the portions below corresponding to a binary 0. The slicer has a fixed threshold level for slicing and a variable offset voltage is combined with the voltage level of the received binary signal to maintain the mid-point of the binary signal applied to the slicer at the slicer fixed threshold slicing level, which preferably is at a zero voltage level.

Abstract: A process and system for synchronizing a video signal to the sampling period of a client system. The synchronization may be suitably implemented by adding or deleting trailing lines from each frame of video data to adjust the frame to match the sampling period of the client system. The synchronization system minimizes the lag between the time a video signal is delivered to the client system and the time the client system responds. The low lag time enhances the responsiveness of the client system to human inputs, and improves the suitability of the overall system to human use. The synchronization system can be achieved using suitable low-cost hardware and is therefore particularly well suited for video game or virtual reality systems designed for home use, increasing the feasibility of employing video inputs in such systems.

Abstract: The invention concerns digital processing of analog signals in a telephone set. A signal path runs between a telephone line and the handset of the telephone. A digital signal processor is interconnected within this signal path, together with appropriate analog-to-digital, and digital-to-analog, converters. This approach allows digital processing of the information carried by the signal path. The digital processing facilitates, for example, (1) a high degree of separation of the handset's microphone signal from the received signal, and (2) independent control of volumes of the received signal and sidetone.

Abstract: A phase-locked loop (PLL) has a phase detector (PD), a charge pump, a loop filter, and a voltage-controlled oscillator (VCO). The PD generates DOWN pulses based on differences in phase between an input signal and a feedback signal. The charge pump has an UP current source that generates a DC UP current and a DOWN current source that generates a DOWN current based on the DOWN pulses received from the phase detector. The charge pump generates a charge-pump current based on the DC UP current and the DOWN current. The loop filter receives the charge-pump current and generates a loop-filter voltage based on a net accumulation of charge from the charge-pump current. The VCO receives the loop-filter voltage and generates an output signal whose frequency is based on the loop-filter voltage, wherein the feedback signal is generated from the output signal.

Abstract: A connector adapter and dust cover therefor are pivotably connected by means of a pivot pin on the adapter and pivot arms on the dust cover whose distal ends engage the pivot pin. The pivot pin has a cam member located intermediate its ends, and the dust cover has a spring arm between the pivot arms whose distal end engages the cam member. In the closed position, the distal end of the spring arm engages the cam below its high point. In the fully open position, the distal end of the spring arm engages the cam beyond its high point. In both positions, the spring arm bears against the cam to resist dislodgment of the dust cover from either position.

Abstract: In an intranet, servers provide services for a price. A user is allotted a budget consisting of a limited amount of electronic credits to pay for the services. When the user requests a selected service from a server, the server offers a service price which can be accepted or rejected by the user, depending on the importance of the service to the user. The service price is determined based on, for example, the workload attendant to the requested service, its quality of service (QOS) requirements and current conditions pertaining to the server, in accordance with a pricing strategy. The latter is implemented to achieve network-wide objectives such as reducing network congestion by balancing load among the servers.

Abstract: A method and system for delivering data over a communication network which ensures proper payment of royalties while preserving free access to data for purposes such as browsing or research. An exemplary method in accordance with the present invention includes the steps of providing a partially-degraded version of the data over the network, without payment of a royalty fee, to a customer at a user terminal connected to the network; and providing a higher quality version of the data to the customer over the network if the customer is entitled to receive the higher quality version. The determination as to whether the customer is so entitled may be made by, for example, determining if a royalty fee payment has been received. The partially-degraded version of the data is substantially recognizable when displayed, printed, played, run or otherwise utilized by the customer at the user terminal, and may be generated by, for example, deleting or encrypting a portion of the undegraded data.

Abstract: A method and apparatus for interactive call identification of a call to a called party. Identifying messages, either text and voice, can be entered on a call-by call basis by the calling party to provide information regarding the call to enable the called party to screen the incoming call. Such identifying messages can include descriptive information about the identity of the caller, the subject matter of the call or any other user-user information. The identifying message is transmitted in the form of a real-time message sent by the caller and received by the party being called in advance of their accepting the call. The identifying message is created on a call-by-call basis by the calling party. The called party receives the identifying message and, after receiving the message, the called party can accept, forward or reject the incoming call based on the information provided in the identifying message. The calling party has the option of using either real-time or pre-recorded message functions.