Abstract: A continuous speech decoder that is built up of multiple layers. Each of the layers uses independent knowledge sources and rules, but all the layers cooperate to quickly decode the speech input into words. A first layer is concerned with acoustic data, a second layer with phone data of speech and a third layer concerns word data and word sequences. By separating these layers, the higher layers can be made time independent and asynchronous. Thus the asynchronous layers can process data quickly and give fast support to the first layer which keeps a dynamic record called a dynamic network of the most likely continuous speech results. The speed and separation of this decoder allows better memory efficiency and better decoder results compared to previously known continuous speech decoders.

Abstract: A domain name server, DNS, for Internet Protocol or Intranet Protocol networks, that has a non-cached domain name server connected to a cached domain name server, and a query server that is connected to both the non-cached DNS and the cached DNS. The query server balances the query load between the DNSs. The query server compares counts of outstanding queries of each DNS and assigns the next query to the DNS with the fewer outstanding counts, unless the query is a special type known as an authoratative query. All authoratative queries are processed by the non-cached DNS. This is because the DNS information stored in the non-cached server is more carefully updated so its information has a higher level of correctness than the information in the cached DNS. Periodic updating of the non-cached DNS keeps its information at this higher level of correctness.

Abstract: The system of the invention uses ATM switches as high performance Internet routers by using standard ATM signaling to set up cut-through paths. The cut-through paths are Switched Virtual Paths which are shared with similar flows toward the same destination and a similar Quality of Service (QOS). The concurrency of multiple flows is achieved by using a different Virtual Channel Identification (VCI) for each flow. The cut-through paths are reused to minimize the overhead associated with setting up and tearing down the paths. Depending on the needs of the traffic, different SVCs having different QOS and different bandwidths can be used. Customer service needs and traffic measurements data are used to identify potential cut-through paths. Once a cut-through path is set up, IP mechanisms are used to update the routing tables of related routers and ATM hosts.

Abstract: A screening and monitoring capability for switch based voice messaging systems that allows a called party to hear the caller and the caller's voice as the caller leaves a message and break in to start a normal telephone discussion if the caller or the caller's subject warrants such action. The method and apparatus of this invention provide for an advanced intelligent network service feature that controls the bridging of the connections to the voice mail and the called party's telephone station. The service feature turns the voice mail off and tears down the connection if the monitoring called party speaks.

Abstract: A method of and apparatus for constructing a control system and a control system created thereby which is based on multiple finite state machines, each in turn based on a finite state machine with constant code (FSMcc). An input preprocessor module normalizes input signals as required by the finite state machine logical processing unit. An output postprocessor module changes the normalized finite state machine outputs to real signals required by the controlled process or apparatus. The input preprocessor allows for a different finite state machine required for an application to be executed by the same constant code contained in the FSMcc microcode storage. An application logic table for the FSMcc is constructed using a specific organization of data representing application logic conditions. The ability to control a process or apparatus with multiple finite state machines allows the designer to use the FSMcc technology more widely and in more complex situations.

Abstract: A method and apparatus for setting up and executing trigger detection points that implement a previously unimplemented ETSI intelligent network capability set 1. Thus, the requirements of this ETSI intelligently networking standard are realized in a fast and innovative way in hardware and software.

Abstract: Call blockages can now be identified down to the line and group of lines that caused the call blockage. No more long statistical studies followed by guessing in order to alleviate blockages. This method and apparatus operating in concert with recently added load balancing processes report specifically what caused each call blockage at what time. With such specific information, each line or group of lines that needs to be moved or added to a new line terminating equipment are quickly identified and accomplished, thereby keeping service to the average customers at acceptable levels. The amount of specificity represents a breakthrough in call line and traffic management.

Abstract: A radio frequency cover and complementary shaped hole that securely fit together to provide electro magnetic and electrostatic shielding of adjustment access holes. The cover preferably is cold headed and the complementary hole is either stamped, machined or cast for inexpensive mass production. Each cover preferably has a socket or slit on its top surface to receive and be guided by a tool, which also secures the cover in an access hole within a second access hole in radio frequency and/or electronic systems. The covers include self tapping thread formers to further reduce manufacturing time and costs.

Abstract: In the system of the invention the adjunct device is connected to the switching system over two connections. The first connection is a voice channel consisting of an analog line capable of transmitting voice but not call data such as called party number, calling party number or the like. The second connection is a data connection via an Ethernet channel between the switch and adjunct that transmits call data but is incapable of transmitting voice. The call data is logically associated with the voice call at both the adjunct and switching system. As a result, transaction data related to a voice call can be forwarded to an agent station or other terminal when the voice call is transferred from the adjunct to the agent station. This system eliminates the need to recreate the transaction data when a call is transferred while using inexpensive analog connections for the voice call.

Abstract: A method and apparatus using a combined MLLR and MCE approach to estimating the time-varying polynomial Gaussian mean functions in the trended HMM has advantageous results. This integrated approach is referred to as the minimum classification error linear regression (MCELR), which has been developed and implemented in speaker adaptation experiments using a large body of utterances from different types of speakers. Experimental results show that the adaptation of linear regression on time-varying mean parameters is always better when fewer than three adaptation tokens are used.

Abstract: Pin fraud of battery reversal pulse metered sophisticated coin telephones is detected, an announcement of the detected fraud made, and the fraud is stopped if the detected fraud continues by the method and apparatus of the invention. By carefully monitoring within the switch office the loop current that is flowing in the 48 volt battery circuit, especially in the more positive terminal portion of the 48 volt battery circuit, loss of current to a fraud ground at a remote coin telephone can be detected. Once a fraud ground is detected, an announcement is made to at least the calling station where the pin fraud was detected, which says in effect "The pay telephone you are using is charging improperly for the service provided. If such improper charging continues the present call will be discontinued." A convenient location in the switch office and an inexpensive component configuration make the addition of this pin fraud detection, announcement and prevention apparatus very cost effective.

Abstract: Multiple subsystem I/O (Input/Output) buses are coupled to one or more system buses of a computer system by interface circuits which perform necessary decoding of memory space and I/O (Input/Output) space for allocation of portions of the memory space and the I/O (Input/Output) space to each I/O (Input/Output) bus. The interface circuits also translate fixed addresses within each I/O (Input/Output) bus to permit proper operation of the I/O (Input/Output) buses with the computer system. The interface circuits are programmed by the computer system to define the allocated memory spaces and I/O (Input/Output) spaces for the corresponding I/O (Input/Output) buses. Programming of the I/O (Input/Output) buses is performed at the time of system configuration by writing appropriate values into configuration registers incorporated into each of the interface circuits.

Abstract: The proposed model aims at finding an optimal linear transformation on the Mel-warped DFT features according to the minimum classification error (MCE) criterion. This linear transformation, along with the (NSHMM) parameters, are automatically trained using the gradient descent method. An advantageous error rate reduction can be realized on a standard 39-class TIMIT phone classification task in comparison with the MCE-trained NSHMM using conventional preprocessing techniques.

Abstract: A method and apparatus for accessing data of a telecommunications interface control RAM to meet the two different requirements of the two types of devices accessing the control RAM. Data for output ports are typically aligned such that time-slot 0 for each port leaves the network at the same time. The hardware therefore requests that the control RAM be organized such that the first N locations correspond to time-slot 0 of all output ports, the next N locations to time-slot 1 of all ports, and so on with the final N locations corresponding to the last time-slot of all ports. The software addressing under processor control, on the other hand, uses data structure groups which are a function of the communication ports.

Abstract: A call scheduling and assignment method and apparatus that schedules each call request on an application processor of a multiple processor system that will process the call request in the smallest amount of time. This method and apparatus then places response time as the important criterion over equal utilization of application processors in a multiple processor system. This method and apparatus recognizes that some application processors are able to respond faster than others, even when operating at the same or higher call load as other processors in the system. For such very common situations, the faster processors can and should be allowed to process a proportionally greater amount of the call processing load. By doing so, the calling customer receives a shorter wait and the system owner receives more call revenue for the same system equipment.

Abstract: A class of features related to voicing parameters that indicate whether the vocal chords are vibrating. Features describing voicing characteristics of speech signals are integrated with an existing 38-dimensional feature vector consisting of first and second order time derivatives of the frame energy and of the cepstral coefficients with their first and second derivatives. Hidden Markov Model (HMM)-based connected digit recognition experiments comparing the traditional and extended feature sets show that voicing features and spectral information are complementary and that improved speech recognition performance is obtained by combining the two sources of information.

Abstract: The called party can temporarily defer answering a call simply by pressing a button on the station set. When the button is pressed, the called party's station set sends a data message to the system's voice response unit directing the voice response to play a "please wait" message to the calling party rather than sending the calling party to the called party's voice mail. The voice response unit plays the "please wait" message to the calling party and sends a call park message to the switching system hosting the station set. The call park message to the switching system directs the switching system to park the call at a specific location as is known in the art and a corresponding call park message to the station set transmits the location of the parked call to the station set. When the called party desires to answer the call, a message is sent from the station set to the switching system requesting that the call parked at the identified location be connected to the station set.

Abstract: A method and apparatus that takes advantage of the connectivity of a unit to speed up communication of needed files to the unit. The method uses each port as a channel to input or pump a portion of the file. If a unit has N equally fast ports, it could possibly pump a file in N portions and take 1/N amount of pumping time. This saves the time of craft personnel performing the load without any degradation of the file. Once all portions are received, the unit has sufficient processing capability to rejoin the portions back into the needed file. The method also allows for the possibility that a port or channel will not be operating properly and automatically sends the portions through the remaining available channel(s) and port(s).

Abstract: The system of the invention uses ATM switches as high performance Internet routers by using standard ATM signaling to set up cut-through paths. The cut-through paths are Switched Virtual Paths which are shared with similar flows toward the same destination and a similar Quality of Service (QOS). The concurrency of multiple flows is achieved by using a different Virtual Channel Identification (VCI) for each flow. The cut-through paths are reused to minimize the overhead associated with setting up and tearing down the paths. Depending on the needs of the traffic, different SVCs having different QOS and different bandwidths can be used. Customer service needs and traffic measurements data are used to identify potential cut-through paths. Once a cut-through path is set up, IP mechanisms are used to update the routing tables of related routers and ATM hosts.

Abstract: A continuous speech decoder that is built up of multiple layers. Each of the layers uses independent knowledge sources and rules, but all the layers cooperate to quickly decode the speech input into words. A first layer is concerned with acoustic data, a second layer with phone data of speech and a third layer concerns word data and word sequences. By separating these layers, the higher layers can be made time independent and asynchronous. Thus the asynchronous layers can process data quickly and give fast support to the first layer which keeps a dynamic record called a dynamic network of the most likely continuous speech results. The speed and separation of this decoder allows better memory efficiency and better decoder results compared to previously known continuous speech decoders.