Abstract: A mobile terminal detects a network formed on a communication media when the mobile terminal is to be connected to said communication media by transmitting control signals to and receiving control signals from the communication media and performs a connection control according to a protocol of a trunk network when the detected network is the trunk network or to a protocol of an ad hoc network when the detected network is the ad hoc network. When the mobile terminal can not detect a network to which the mobile terminal can be connected, the mobile terminal sets a new ad hoc network identifier for forming a new ad hoc network, after determining that the new ad hoc network identifier is not already in use another network.

Abstract: In a computer network formed by subnetworks in which a multicast scope is defined as a reachable extent of multicast packets, each host in the scope transmits a multicast packet containing a scope field indicating that transmission of the multicast packet beyond the scope is forbidden and a boundary flag indicating whether only those hosts within the scope are allowed to participate in a group activity or mobile hosts outside the scope are also allowed to participate in the group activity. Each router connected to the subnetworks of the scope receives a multicast packet and prevents it from being transmitted beyond the scope if the scope field is contained in the received packet. A home agent located within the scope receives a multicast packet and transmits it to a mobile host visiting the external subnetwork if the boundary flag of the received packet indicates that participation in the group activity is allowed.

Abstract: A mobile terminal and method of controlling the same, each that can integrally move between an infrastructure network and an ad hoc network. The mobile terminal can interconnect to an infrastructure network and an ad hoc network while moving. In an address management process, a connection network identification process, and a destination address capture process which are needed to establish communications with the terminal connected to a network, an infrastructure network connection procedure is integrated with an ad hoc network connection procedure. The integrated configuration consists of an ad hoc/ infrastructure network address management means, an ad hoc/ infrastructure integration move management means, an ad hoc/ infrastructure network integration destination address capture means. In each process, it is identified whether or not the mobile terminal is connected to an ad hoc network or infrastructure network, so that the procedure corresponding to the connected network is used.

Abstract: An originator selects at least one sub-network from a plurality of sub-networks to establish a scope of a multicast group and multicasts group characteristic information, which includes a specific group identifier to the scope. The sub-networks include nodes to accept a mobile host. The mobile host is related to one of the nodes to be managed by it. Each node receives the group characteristic information when it is in the scope and judges whether or not to transfer the group characteristic information to the mobile host in response to the content of the group characteristic information. The mobile host judges where the group characteristic information is sent from when it receives the group characteristic information. Then the mobile host judges whether or not the originator admits the mobile host to participate in the multicast group on the basis of the group characteristic information.

Abstract: An ad hoc LAN is set up in an environment in which there simultaneously exist a backbone-connected LAN connected by a high-speed backbone network and an ad hoc LAN not connected to the backbone network, wherein the ad hoc LAN is capable of communicating simultaneously with both LANs. An ad hoc LAN is defined as one multicast group used within the backbone network and backbone-connected LAN. This makes it possible for a terminal, which is connected to the ad hoc LAN and backbone-connected LAN simultaneously so that it may communicate with both LANs, to participate in the ad hoc LAN in the same manner as it would participate in a multicast group of a conventional backbone-connected LAN. As a result, the terminal construction is simplified and the terminal can be connected to an ad hoc LAN and backbone-connected LAN simultaneously.

Abstract: A method for allocating a geographical identifier is executed in a network for supporting a plurality of movable hosts being moved among a plurality of sub networks as keeping communications therebetween. When movable hosts 100, 101 moves to a sub network 110, an agent 120 of the sub network 110 allocates the same geographical identifier depending on the sub network 110 to the movable hosts 100 and 101 and stores logical identifiers and link layer addresses of the movable hosts 100 and 101. The packets destined for the movable hosts 100 and 101 are sent to the agent 120. The agent gets to know the link layer address corresponding to the logical identifier of a destination movable host described in a network layer header of the received packet from the registration table, generates the link layer header of the destination movable host, and add the link layer header to the received packet before transferring the received packet.

Abstract: A transmission side generates a first frame comprising a field for storing error correction coded information and a field for storing identification information representing error correction coded information, and generates a second frame comprising a field for storing non error correction coded information, a field for storing identification information representing non error correction coded information and a field for storing an error-detecting bit. A reception side detects whether or not a received frame has an error based on bit contents of the field storing the error-detecting bit of the received frame. If the received frame has an error, a frame type is detected based on the identification information. If the detected frame type is the first type the received frame is output to a process stage. If the detected frame type is the second type the received frame is abandoned.

Abstract: A location information management table 14 stores information on the location of the host both at present and prior to the movement. The host is movable based on the transmission address contained in the frame transmitted therefrom. The frame addressed to the host is transferred to the segment presently connected and to the segment that has been previously connected based on the table 14. Even though the host locates halfway among a plurality of segments, stable communication can be realized. When the host moves away from other segment toward the center of one segment, the frame is transferred to one segment at a time when the timer terminates its counting operation, thus deleting unnecessary traffic.

Abstract: A message routing system for use in a mobile packet communication system where communication between mobile terminals is conducted by evaluating headers of data messages. The source and destination of each data message is determined by evaluating the inherent address and position address of each data message without each header containing any association data relating the inherent address and the position address. By facilitating such a message routing process, the amount of communications between mobile terminals is reduced.

Abstract: In a computer network where each mobile host is assigned a subnetwork-independent logical identifier (L-ID), the mobile host acquires a subnetwork-dependent geographical identifier (G-ID) from an agent-existing subnetwork and sends the G-ID to the home subnetwork of the mobile host for location registration if a beacon received therefrom is a first one or different from a previous one or acquires a G-ID from an agentless subnetwork and sends the G-ID to the home subnetwork if no beacons are received. At intervals, a multicast packet is sent from the mobile host, containing the acquired G-ID to the agentless subnetwork. A similar multicast packet is received from a second mobile host and the G-ID contained therein is stored in a database corresponding to the L-ID of the second mobile host. When establishing a session, a data packet is sent from the mobile host to the network along with a G-ID corresponding to a L-ID of the data packet stored in the database.

Abstract: A mobile packet communication system has a plurality of mobile terminals movable between a plurality of auxiliary networks. Each of the mobile terminals has an inherent address processing layer for adding a packet header based on the inherent address thereof to transmission data from a higher layer, and a positional address processing layer for adding a packet header based on the positional address thereof to the transmission data to which the packet header based on the inherent address has been added by the inherent address processing layer. The packet headers contain no header information representing an association between the inherent address and the positional address. The mobile packet communication system facilitates a control process for maintaining data relative to an association between addresses and reduces an amount of communications required by such a control process.

Abstract: In a diversity receiver provided with diversity branches (51 and 52) which have a predetermined distance therebetween and each of which produces a received signal sequence in response to reception of a radio signal directed to the diversity receiver with a symbol rate, each of samplers (56 and 57) samples the received signal sequence with plural times of the symbol rate to produce a sampled serial signal sequence. Each of serial-to-conversion circuits (58 and 59) converts the sampled serial signal sequence into parallel signals. With reference to the parallel signals, channel impulse response estimation circuits (61 and 62, or 63 and 64) estimate channel impulse responses, respectively. In accordance with each of the estimated impulse response signals and each of the parallel signals, branch metric calculation circuits (66, 67, 68, and 69) calculate branch metrics. A branch metric combining circuit (71) combines the branch metrics into a combined branch metric.

Abstract: A maximum likelihood sequence estimator implemented to cope with an intersymbol interference from a single symbol employs a necessary channel impulse response estimation vector of two components for a decision to estimate a maximum likelihood sequence. A channel impulse response estimator implemented to estimate a channel impulse response in consideration of intersymbol interferences from up to two symbols outputs a channel impulse response estimation vector of three components. The three-component vector is input to a converter, where it is converted into the two-component vector to be output to the maximum likelihood sequence estimator.

Abstract: A receive condition in burst transmission varies in units of a burst or a plurality of symbol times. The present invention provides a soft decision signal outputting receiver which can simply produce correct reliability information even if the receive condition varies while employing a simple reliability information production method. A demodulator demodulates a burst signal to obtain a hard decision signal. A time-varying reliability information production circuit receives information of a demodulation process from the demodulator 11 and produces reliability information regarding the hard decision signal. The time-varying reliability information production circuit receives signal quality information which reflects a receive condition for each burst or each plurality of symbol times, and changes, based on the signal quality information, a production method for reliability information or elements such as threshold levels to be used in a production method for each burst or each plurality of symbol times.

Abstract: An adaptive equalizer estimates a channel impulse response h in a channel impulse response estimator and outputs a result to a channel impulse response convertor. The channel impulse response convertor removes interference components from the channel impulse response h, outputs a channel impulse response h' to an equalizer. The equalizer sets up, based on the channel impulse response vector h', internal parameters, and equalizes, based on the internal parameters, reception signals. By configuring as above, distortion by intersymbol interference and a deterioration of reception performance by co-channel interference waves can be prevented.

Abstract: In a digital data demodulating apparatus, a sampling circuit samples a received signal at a speed N.times.K (N>1, K>1; integers) times a symbol rate to output received signal sequences. N received signal sequence selection circuits estimate channel impulse responses from the respective received signal sequences to obtain channel state data, and output a control pulse and the estimated channel impulse response values. A received signal sequence selection controller outputs a switch control signal on the basis of the channel state data. N selectors select demodulation received signal sequence candidates from the received signal sequences on the basis of the control pulse. A first switch selects/outputs a received signal sequence to be demodulated on the basis of the switch control signal. A second switch selects/outputs an estimated channel impulse response value estimated from the received signal sequence to be demodulated on the basis of the switch control pulse.

Abstract: A maximum likelihood sequence estimation apparatus decodes a digital data signal. The apparatus includes: a sampling device to perform sampling and to output a received signal by sampling pulses having a constant time interval T and N different sampling phases; an operation device to operate a branch metric of the received signal sampled by the N different sampling phases; and a device to perform maximum likelihood sequence estimation of the branch metric.

Abstract: A receiver for digital radio communications has an equalizer which is powered off for reduced power consumption when received radio waves undergo relatively small intersymbol interference. When intersymbol interference is significantly large, the equalizer is powered on to remove the intersymbol interference for better signal reception. The receiver has a signal quality decision circuit which powers on the equalizer when the quality of the received signal is poor, and powers off the equalizer when the quality of the received signal is good. The receiver consumes a relatively small amount of electric power overall.

Abstract: When time-division multiplexed digital data signals are to be demodulated, symbols in a time slot and symbols in the next time slot are stored in memory. The digital data signal in each time slot includes a training sequence. Channel estimators extract channel information regarding stored signals from the signals themselves as well as from the training sequences in succeeding time slots and output the channel information to a channel information generator which combines the channel information from the channel estimators into channel information regarding the signal in the time slot. A demodulator receives the combined channel information and the stored signals and demodulates the signal in the time slot.

Abstract: An adaptive sequence decoder comprises a maximum likelihood sequence estimator for deriving a decision sequence of a received signal in response to a control signal. Channel estimators of zero-th to m-th order of channel impulse response are connected to the MLSE to produce a first signal representing a respective order of a replica of a transmitted signal and a second signal representing a respective order of channel impulse response. The first signals from the channel estimators are summed to produce the replica of the transmitted signal for comparison with the received sequence to produce an error signal, and the second signals are summed to produce the control signal. At periodic intervals, the MLSE control signal is initialized and the zero-th order channel estimator is reset to the initialized control signal, and the other channel estimators are reset to respective initial values to prevent time dispersion.