Abstract: A learning method supervised by a binary teacher signal for a binary neural network comprises at least an error signal generator 10 for weighting factor updating, which generates an error signal for weighting factor updating having an opposite polarity to that of a difference signal between an output unit signal of the binary neural network and the binary teacher signal on an output unit whereat a binary output unit signal coincides with the binary teacher signal, and an amplitude which decreases by increase of distance from the binary teacher signal, when an absolute value of the difference signal is smaller than a threshold, generates an error signal which has the same polarity as that of the difference signal and an amplitude smaller than that of the difference signal, when the absolute value of the difference signal is larger than the threshold, or generates an error signal which has an amplitude equal to or smaller than that of the difference signal on an output unit providing a wrong binary output unit

Abstract: An authentication system whereby authentication load can be distributed in the network without sharing secret information of users is provided.The system has a single master authentication center arranged in the network, the master authentication center sharing with the user a user secret key, and a plurality of slave authentication centers sharing with the master authentication center respective secret keys different from the user secret key. The master authentication center authenticates the user by using the user secret key and issues a certificate information which certifies legitimation of the user, to the user if the user is authenticated as a legitimate user. The slave authentication center authenticates the certificate information from the user and issues a permission information which allows an access to a specified server or an application server in the network, to the user if the user is authenticated as a legitimate user.

Abstract: In a radio communication system using time division multiple access system (TDMA) for access control method, modulation signal sent by a slave station is frequency-dispersed during a TDMA burst period by using a chirp signal which frequency changes continuously on time axis so that signal power density on communication link is decreased and interference signal affected by the slave station to another radio communication system is reduced. According to the present invention, modulation signal transmitted by each slave station is multiplied to a chirp signal which frequency changes continuously on time axis during a TDMA burst period so that carrier frequency which carries the modulation signal changes continuously on time axis to disperse on frequency axis to a predetermined bandwidth thereby decreasing signal power density of transmit signal from the slave station.

Abstract: A speech endpoint detection unit determines that a first condition is met when a matching score of a partial sentence accepted by a grammatical rule unit is the highest of all partial sentences, and that a second condition is met when a duration time of input speech determined to coincide with a silent standard pattern is longer than a predetermined time. A speech endpoint is determined when the first and second conditions are both met and a speech endpoint detection signal is sent to a word prediction unit and a recognition result output unit. By requiring that both first and second conditions be met, a speech endpoint can correctly be detected even when a long silent period is present in the course of a sentence.

Abstract: Light beams emitted from an optical fiber are converted into parallel light beams by a lens, and then allowed to pass through a Faraday device, followed by being converged to an incident end of the optical fiber by a lens. The Faraday device is applied with a magnetic field in a direction of the optical axis thereof from a coil. Moreover, the Faraday device is applied with a magnetic field in a direction perpendicular to the optical axis by a pair of magnets. When an electric current flowing in a coil 74 is changed from I to -I, the direction of a synthesized magnetic field can be rotated. Therefore, by arbitrarily changing the electric current flowing in the coil, polarization of transmission light can be controlled.

Abstract: An automatic facsimile signal analyzing system includes a switch unit, a service trunk, a signal identification trunk, a facsimile communication determination unit, and a control unit. The signal identification trunk restores a facsimile control signal from a modulated and encoded facsimile signal on the set speech path and performs collection of communication state information and detection of a start of facsimile communication on the basis of the restored facsimile control signal. The facsimile communication determination unit permits an operation of the signal identification trunk when the classification result from the service trunk represents a type other than the voice signal and analyzes the communication state information from the signal identification trunk.

Abstract: This invention provides an optical amplifier which has low noise characteristics for a wide range of input signal electric power. The optical amplifier has a plurality of cascaded optical amplifying units, a signal level detector, and a pump-source controller. The pump-source controller changes controlling objects from a pump-source corresponding to a first one of the optical amplifying units to a pump-source corresponding to a second one of the optical amplifying units according to a detection level of the signal level detector from a low level to a high level. When the detection level is outside of a control object range, the optical amplifier controls an output from the pump-source to stabilize. When the detection level is within the control object range, the optical amplifier controls an output from a pump-source of each of optical amplifying units according to the detection level.

Abstract: Validation of an input protocol is divided into two steps. In the first processing step, a state. transition graph of each process including only executable state transitions of each process is generated according to transition information which the protocol specification has, and at the same time, protocol errors related to behavior of the process is detected. In the second processing step, reachable system states among system states defined by combinations of individual process states and individual channel states are successively generated according to transition information which the state transition graph of each process has, and a system state of the generated reachable system states which cannot transit further is detected as a deadlock. In the protocol validation method, every system state can be erased from the memory when a deadlock detection for it is completed, and searching the generated system states is not necessary, and some additional techniques for more efficient validation can be applied.

Abstract: The present invention proves an electronic dealing system, which electronically performs matching processing between information on transaction orders placed by order side customers and information on transaction orders placed by hit side customers so as to establish transactions the electronic dealing system detects orders for which hit requests have been made, but for which transactions have failed to be established for a long period of time and eliminates those orders from the electronic dealing market. Further, when transactions fail to be established due to factors other than the transaction price and the transaction amounts, other customers which would help establish the transactions between order side and hit side customers are found so as to increase the number of transactions which are established.

Abstract: In a payer-variable exchange system, exchanges 2A and 2B and payer change control devices are provided corresponding to telephones 1A and 1B used by communicators A and B, payer change function units 4A and 4B of the payer change control devices 3A and 3B transmit and receive signals with the payer change function units 4A and 4B of the other parties, a communication amount for which the communicator itself is the payer is calculated by the charge calculation units 5A and 5B, and the same operation is repeated every time a payer change request is received.

Abstract: An efficient packed encoding rules (EPER) for ASN.1 for efficiently encoding the value to each type of abstract syntax definition, in which structure of an encoding data is divided into three fields of a bit field, an octet field following the bit field, and an offset field selectively added to before the bit field, in the bit field, bit data is set in the order of the abstract syntax definition, in the octet field, octet data in a unit of octet is set in the order of the abstract syntax definition, in the offset field, a length information of bit field is set and, in the bit field, padding is made so that a total length of the offset field and the bit field is an integer multiple of 8 bits, and the offset field is added when the length of bit field is not determined from the abstract syntax definition, thereby eliminating problems of PER.

Abstract: A moving image with a plurality of continuous frames is processed by using a plurality of image processing units (5-1 through 5-n), which operate simultaneously. A frame identification designator (4) attaches an identification number cyclically to each frame of the input image signal. A frame identification selector (6) in each image processing unit inputs an image frame selectively based upon the identification number, and transfers the selected frame to a corresponding image processing module. The image processing units (5-1 through 5-n) typically carry out either the same processing on respective image frames, or a plurality of processing to a single frame. An additional processor network (8) may be coupled with the image processing units for high-level processing including recognition and/or understanding of an image.

Abstract: In a satellite mobile communication system in which at least one mobile earth station communicates with a land earth station which is coupled with a network, through a satellite, the land earth station measures frequency offset in a request signal and/or a response signal in a received signal from the mobile earth station due to frequency offset by a transponder in the satellite and/or a local oscillator of said mobile earth station, and doppler shift due to movement of the satellite. The measured frequency offset is forwarded to the mobile earth station through a singalling channel or a communication channel through a satellite. The mobile earth station compensates frequency offset by adjusting oscillation frequency of a local oscillator in the mobile station based upon the measured frequency offset sent from the land earth station.

Abstract: In the preliminary authentication stage, the mobile station (35) is authenticated by sending from the home network (30) to the roamed network (31), a plurality of pairs of first random numbers (RNDa.sub.1, . . . , RNDa.sub.n) and calculation results (SRESa.sub.1, . . . , SRESa.sub.n) of the cipher function (f), which calculation is performed at the home network using the secret key (ki) and the first random numbers (RNDa.sub.1, . . . , RND.sub.an), by sending, from the roamed network to the mobile station, third random numbers formed by coupling second random numbers (RNDb.sub.1, . . . , RNDb.sub.m) produced at the roamed network with the first random numbers (RNDa.sub.1, . . . , RNDan), by sending, from the mobile station to the roamed network, calculation results (SRESa.sub.1, . . . , SRESa.sub.n, SRESb.sub.1, . . . , SRESb.sub.

Abstract: A frequency offset compensation method has the steps of raising the detected signal to the M-th power to remove a modulation factor from the received signal, and accumulating the M-th power signals for N (N is an integer) symbols to derive a phase component of the accumulated M-th power signals. Then, the derived phase component is divided into M so as to obtain an estimation value with respect to phase shift due to frequency offset which represents frequency deviation from a true carrier frequency contained in the received signal, and the received signal is multiplied by the obtained estimation value in a form of conjugate complex number to remove the phase shift due to the frequency offset from the received signal.

Abstract: A communication system includes a bidirectional tree network having a base end and a plurality of terminating ends, a central apparatus connected to the base end, a plurality of terminating equipment connected respectively to the terminating ends, and a plurality of subscriber terminals accommodated respectively in the terminating equipment. Each of the terminating equipment has a multiplexer for multiplexing information transmitted from the subscriber terminal accommodated therein to the central equipment, and a signal selection device for discriminating and selectively receiving only information which is destined for the subscriber terminal accommodated in this terminating equipment from information transmitted from the central apparatus to all the subscriber terminals via the tree network.

Abstract: A mobile communication system has a cell structure constituted by integrating macro cells and micro cells, and at least one mobile station. The cell structure is made to cover a service area by a plurality of the micro cells each having a predetermined size and to cover the same service area by the single macro cell having a larger size than the size of the micro cell. Each of the micro cells includes a micro cell base station for transmitting at least radio control channel at a predetermined frequency and slot. The mobile station includes a measuring unit for repeatedly measuring received level of the radio control channels from the micro cell base stations, an estimation unit for estimating its moving speed in accordance with changing rate of the measured level, and a determining unit for determining, in accordance with the estimated moving speed, that control of a radio link between the mobile station and the base station is executed by a micro cell unit or by a macro cell unit.

Abstract: A variable length code system which is instantaneously decodable in both forward direction and backward direction is generated either in asymmetrical form or symmetrical form. The code thus generated has high transmission efficiency, and capable of backward reconstruction of video signals even when some bits are lost during transmission. The symbols to be encoded are prepared (1) with occurrence probability, and they are provisionally encoded to a non-reversible variable length code (2). Then, the provisional code is transformed to an asymmetrical reversible code (4) through equivalent transform process (9), reverse symbol tree process (10), and bit supplement process (12) when suffix condition is not satisfied. The provisional code is also transformed to a symmetrical reversible code (3) by assigning a symmetrical code word to each symbol on a symbol tree (FIG. 5). At least one of symmetrical code or a asymmetrical code is selected (5), and is provided as a final resultant code.

Abstract: Direct encoding of composite video signal which includes both luminance signal (Y) and chrominance signal (C) is carried out without separating luminance signal from chrominance signal. A motion of a block in a frame (8) which provides a motion vector (MV.sub.x, MV.sub.y), and a from a preceding frame is detected in a motion detector reference block is defined in said preceding frame which is locally decoded according to said motion vector. A block to be encoded in a current frame is converted through Hadamard conversion H (1). A composite motion compensation (7) provides a prediction block according to said reference block, which is subject to phase compensation of color sub-carrier, according to remainder of the value MV.sub.x +MV.sub.y when said value is divided by 4assuming that sampling frequency of video signal is four times as high as color sub-carrier frequency.

Abstract: A public telephone system includes public telephone sets assigned with different specific numbers and allowing speech communication using cards and a center apparatus connected to the public telephone sets through a telephone line. The center apparatus includes a list number forming unit for forming one list number from numbers of a plurality of invalid cards produced by illegally rewriting the cards, a registering unit for registering each formed list number in association with a series number, management number, and telephone number of each of the telephone sets, a storage unit for storing the list number and the series number, management number, and telephone number of each telephone set, all of which are registered by the registering unit, and a list number transmitting unit for calling, through the telephone line, each public telephone set corresponding to the telephone number registered in the storage unit and transmitting a list number corresponding to each telephone set.