FR2645690A1 - Device for initialising a communication channel between a subscriber set and a base station in a communication system - Google Patents

Abstract

The system comprises a base station 104 in a network with a multitude of subscriber sets 41, where control information is communicated between the base station and the subscriber sets via a radio control channel at a frequency chosen by the base station from a multitude of frequencies. The base station transmits control messages comprising a single network number via the control channel. Timing of the communications transmitted by a subscriber set via a communication channel between a base station and the subscriber set is the subject of a refinement during initial establishment of the communication channel.

Description

 The current. The invention relates to subscriber communication devices and, more particularly, to the initialization of a communication channel between a subscriber station and a base station in such a device.

 A subscriber communication system of the prior art is described in United States Patent No. 4,675,863 of June 23, 1987.

 The present invention provides a subscriber communication device in which a base station is incorporated into a network with a multitude of subscriber stations, and wherein control information is communicated between the base station and the subscriber stations. subscribers by a radio frequency (VCR) channel at a frequency selected by the base station from a multitude of predetermined frequencies.The base station transmits the control messages via the VCR channel. The control messages include a network number belonging to the base station.

Each subscriber unit processes the network number of the control message from the VCR channel to allow the subscriber unit to process the control message according to whether or not the subscriber unit is in the same network as the subscriber station. the base station.

 Each subscriber station may also operate in a search mode in which it searches for the VCR frequency by transmitting a VCR acquisition message at each of the predetermined frequencies, each acquisition message including an identification number which belongs to the station. subscriber. The base station processes the subscriber identification number in an acquisition message received via the VCR channel to determine if the subscriber unit is in the same network as the base station, and transmits an acknowledgment to the subscriber station that the VCR channel has been acquired by the subscriber station when such subscriber identification number processing indicates that the subscriber's station is in the subscriber's station. the network leads only the base station.

 The present invention also provides a system.

subscriber communication method, wherein the timing of communications transmitted by a subscriber station on a given communication channel between a base station and the subscriber station is refined during the initial establishment of the communication channel . The base station includes a master clock for providing a system timing signal.

The subscriber unit comprises an internal synchronization generator for producing a subscriber station synchronization signal for synchronizing the signals that are transmitted over the given communication channel between the subscriber station and the base station, and provides a refinement signal indicating the timing of the internal synchronization signal. During the initial establishment of a communication channel between the base station and the subscriber unit, the latter transmits the refinement signal via the given communication channel, from the subscriber station to the base station, and the base station processes the refinement signal from the subscriber unit in connection with the system timing signal to determine the value of an offset between the timing of the system timing signal and that of the refinement signal.

The base station communicates the determined value of the offset to the subscriber station; and the subscriber unit processes the offset value communicated by the base station to adjust the synchronization signal of the subscriber station to reduce the offset.

 The present invention further provides a subscriber communication device in which the DC signaling information and so many voice data are transmitted through an assigned channel between a line that couples a base station to an office. central and a line interface that couples a subscriber station to a subscriber terminal. The system processes the DC signaling information for communication by the assigned channel between the line and the line interface by detecting this information on the line and / or line interface and conditions the detected signaling information. for communication by the assigned channel instead of the voice data signals.

The present invention will be better understood from the following description made in conjunction with the accompanying drawings in which
Fig. 1 is a block diagram of a preferred embodiment of the subscriber unit of the present invention;
Figs. 2A and 2B show the processing programs for establishing communications between a base station and a subscriber station in the same network as the base station;
Fig. 3 shows processing programs for refining the timing of transmissions at the base station of subscriber station signals;
Fig. 4 shows processing programs for transmitting DC signaling information over an assigned channel of voice data communication.

 A glossary of the abbreviations used in the brief is given below.

CRC Cyclic Redundancy Check
ECV Control Package Track
ETD Remote Connection Processing Set
HF Radio Frequency
MDT Multiplexed to Divis Ion in Time MIC Modulation by Encoded Pulse SXPEE Programmable Dead Memory Programmable Erasable Electri
cally
MU Single Word
MUX Multiplexer
NIA Subscriber Identification Number
NIR Network Identification Number
PBB Baseband Processor
PLER Linear Prediction Excited by Residual RElU Refinement of Unique Word
SF Fractional Synchronization
TCA Subscriber Order Task
TCV Track Command Task
TEC Combined State Buffer
UCV Vocal Codec Unit
VCR Radio Control Channel
In connection with FIG. 1, a preferred embodiment of the subscriber communication system of the present invention comprises a base station 104 and a plurality of subscriber stations 41. This preferred embodiment is useful with a base station described in the patent application of
United States of America: "Base Station
For Wireless Digital Telephone Svstem "(Base station for digital telephone system without it) and we will use the same references as in this application to designate identicues components.

 The base station 104 includes a central office 13, a remote connection processing unit (DTE) 14, a master clock 18, a multiplexer assembly (MUX) 19 and a track module 20. The central office 13 is connected to a desk The central unit 13 is coupled to the channel module via a T1 junction line 28 and to the multiplexer 19. The multiplexer 19 proceeds to multiplex different communication channels in different sectors on the line. 28. Each channel module 20 comprises a channel control assembly (ECV) 23, an encoder-decoder unit (UCV) 24 and a modem 106.

ECV 23 places communication paths in different high frequency paths. The UCV set 24 conditions the voice communication signals conveyed by the communication channels. The modem 106 allows the transmission and reception of voice signals and data communication signals by an assigned RF channel. The ECV 23 transfers the communication signals between the assigned HF communication channel and the assigned communication channel in the affected sector on the trunk line 28. The DTE 14 monitors the sectors of the trunk 28 and the status of the HF voids and then assigns communication channels to predetermined sectors and predetermined RF channels in accordance with the predetermined assignment program. The ECV set 23 exchanges control messages with the subscriber stations 41 through radio control channel (VCR) in a given sector of a predetermined RF path.

 Each subscriber unit 41 includes a modem 107, a baseband processor 112, and an internal synchronization generator 113. The processor 112 is connected by a two-wire interface line 27 to a subscriber terminal, such as a telephone 115 and / or a data processor 116. The processor 112 comprises two modules implemented by software, a subscriber control task module (TCA) 100 and a channel control task module (TCV) 105. TCV 105 is responsible for word synchronization and frame alignment, collision detection and resolution, and error detection.

The ECV set 23 and all the TCV 105 tasks that listen on the VCR path must exhaustively verify a valid VCR message in each VCR sector.

The TCV task 105 executes it by scanning a single word (MU) in a window - 4 symbols around a nominal MU location, based on the synchronization of the mother system. The ECV 23 listening on the VCR channel scans the single word in a window - 3 symbols around the nominal MU slot. A search algorithm shifts the data until it finds the MU profile or until all the possibilities are exhausted. As soon as the MU profile is found, the VCR message is considered valid only if the VCR checksum is correct.

 The net TCA 100 task implements a VCR frequency search algorithm. The purpose of this search is to enable the subscriber unit 41 to find the base station 104 belonging to the same network as the station 41 as quickly as possible, and to prevent the subscriber station 41 from trying to communicate. with known incorrect networks. Each base station 104 has a unique network identification number (NIR). Each subscriber unit 41 has a unique 24-bit subscriber identification number (NIA). This number is stored in an electrically erasable programmable read-only memory (vM.PEE) of the subscriber unit 41. All subscriber identification numbers of a part network are stored in the network database at the same time. base station 104.

 VCR search is either active or passive.

The active search is executed only when waiting for a call. The subscriber unit 41 is accepted for the network and determines its unique identification number only through the active search. When a call is not on hold, the set performs a passive search using its known unique identification number (NIR) to acquire the correct VCR path again.

 If all possible VCR frequencies have been tried unsuccessfully in each search mode, TCA 100 tries to reset the hardware.

This could possibly erase a system fault that prevents the subscriber unit 41 from getting synchronization. Reset resets the modem again. The update of the modem adapts its filters to the current conditions of the environment. If the handset is off-hook when all frequencies have been tried without success, the task
TCA 100 causes the output of a possible busy VCR tone, fast to the handset 115.

 Whenever task 100 performs a reset, it reads the identification number NIA and the network identification number (NIT) in the electrically erasable programmable read only memory. If there is no NIR identification number in the erasable ROM, it is reset by default. When the task 100 gets synchronized on a VCR frequency in a passive search, it compares the received NIR with the internally stored NIR and rejects all the VCR frequencies with non-matching NIRs.

Active VCR search is initiated only when a call is on hold. When the call waiting state ends, the user hangs up or the set enters a suspended state. Then the active search becomes a passive search. If the task 100 has unsuccessfully tried all the VCR frequencies, it switches to a suspend state and sends the handset 115 a new order tone. This clears the waiting state of the call and causes the search mode to change from active to passive. ors-
that the task 100 determines its affiliation to the network, the search ends.

The task 100 determines the network affiliation of the subscriber's station and the NIR identification number by the normal procedure of establishing a call. Task 100 performs a frequency search. Each time the task 100 obtains synchronization on a VCR frequency, it sends a message "VCR call request". If the base station 104 recognizes the
NIA, it answers either with a "call connection" message if it wishes to complete the call, or with an "erase indication" message with the new order erase code if it is too busy to complete the call. In either case, the search is terminated and the NIR number in the data area of the VCR message is saved in an electrically erasable programmable read-only memory by station 41 for retention in memory during interrupts. power.

If the base station 104 does not recognize the NIA number, it sends an erase indication message with the unknown subscriber delete code to the subscriber station 41. The task 100 then produces the next frequency search of the VCR channel. The absence of acknowledgment from the base station also causes the task 100 to produce the next frequency for the search.A new frequency may also be requested by the task.
TCV 105 following a loss of synchronization.

 After finding the correct network, task 100 performs a passive search each time it loses VCR synchronization, or passes VCR from a voice channel. It also performs the passive search if the network number is not confirmed, but the waiting state of a call is cleared. If the subscriber unit 41 detects a handset off-hook (service request) it starts an active search. The following events cause the task 100 to produce the next VCR frequency in the active search mode: (a) a requesting a new frequency from task 105 due to an AM hole failure or a VCR sync loss; (b) a return to the control channel from a voice channel, or (c) a VCR sxtnchronization obtained on the wrong network.

 To increase the speed of the passive search, the task 100 stores up to six frequencies that correspond to its base station 104. When a search, active or passive, is requested, the frequency generation algorithm alternates between frequencies to from a stored VCR frequency table and an incremental frequency counter. This gives priority to the most probable frequencies and accelerates the acquisition of the base station after a brief failure.

 Whenever task 100 gets the synchronization on a VCR, it looks for a match between its stored NIR identification number and its received NIR identification number. If there is no match, task 100 has acquired the synchronization on the bad network, and it produces a new frequency on which it tries to obtain synchronization. If the identification numbers are matched, then task 100 found the correct network and the search ends.

The general programs executed in the active search mode are summarized in connection with Figure 2A. The task 100 in the subscriber unit 41 executes a program 120, in which an ac acquisition message which comprises the NIA of the subscriber unit 41 is transmitted sequentially on each of the frequencies
Given VCR which are used by the base station 104 of the assigned network of the subscriber's station. The set ETD 14 in the base station 104 executes a program 121 to determine whether the NIA number contained in the received acquisition message has a frequency
Given VCR corresponds to an identification number NIA in a list of NIA identification numbers stored in the base station.If the identification number NIA in the acquisition message transmitted by the subscriber station corresponds to the one of the numbers stored in the base station 104, the latter then executes a program 122 for transmitting an acknowledgment message "to the subscriber station on the frequency
VCR given. The acknowledgment message includes the identification number of the base station. The subscriber unit 41 responds to the acknowledgment message "by executing a program 123 which allows the subscriber unit 41 to process the control messages.

The subscriber unit 41 also responds to the acknowledgment messages by executing a program 120 storing the identification number NIR in the subscriber station memory.

 If the identification number (NIA) in the acquisition message transmitted by the subscriber station does not correspond to any of the NIA identification numbers stored in the base station 104, the latter transmits a message Negative acknowledgment "at the set by the given VCR frequency Upon receiving a negative acknowledgment message, the subscriber set 104 executes a program 125 to change the given VCR frequency and repeats then the program 120 for transmitting an acquisition message at the given VCR frequency changed.

 The general programs executed in the passive search mode are summarized in connection with Figure 2B. The subscriber unit 41 executes a program 127 for sequentially receiving control messages transmitted by each of the VCR frequencies used in the network to which the subscriber station is assigned. At a given VCR frequency, the base station 104 executes a transmission program 128 of a control message which has the identification number NIR. For a control message received at a given VCR frequency, the subscriber station 41 executing a program 129 for determining that the NIR identification number in a received control message corresponds to the NIR identification number stored in the subscriber station. If the numbers match, the subscriber unit 41 executes a program 130 which allows this station to process the control messages from the base station 104. If the numbers do not match, the subscriber unit 41 executes a program 131 for changing the given VCR frequency to which the station subscriber receives command messages; and the program 129-comparing the NIR identification numbers is repeated.

 Refinement of synchronization is performed at the beginning of each voice connection made by the assigned communication channel. The intention is to finely synchronize the transmission symbols of the subscriber station to bring it to + 3% of the master clock of the symbols of the base station.

 To achieve a tolerance of -3%, fractional time offset values "At" are collected for a number of frames at the subscriber station.

Each transmission burst from the base station provides another data point in the sample list of the fractional time offset values. Periodically, the average of the "At average" samples is calculated to produce an estimate of the offset. This estimate is used to adjust the internal synchronization generator of the subscriber station to bring it closer to the desired value. This process is continued until the base station detects that the synchronization of the subscriber station is within 3% of the correct value of symbol synchronization.

 The ECV set 23 of the base station automatically enters the refinement operation when a voice channel is assigned to it. Set 23 instructs modem 106 to initialize said operation and transmits refinement bursts. Each burst contains power information, symbol timing, and fractional synchronization for the subscriber unit 41.

The ECV set 23 of the base station successfully receives a subscriber refinement burst if a unique refinement word (RMU) is found and the control
CRC is verified as correct. If, at any time, the base station set 23 fails to receive a subscriber burst, the next base station transmission burst contains a zero for symbol synchronization. In addition, if the base station determines or the quality of the link of a subscriber burst has fallen below a predetermined level, the station indicates it in a command nulltiplet at the subscriber station. initializing a bit "SF ignored". The subscriber station then rejects the fractional time information contained in the burst.

 The refinement operation terminates successfully when the base station reads three successive fractional time values at the master timing signal from the master clock 18.

Completion of successful refinement is signaled to the subscriber station via the order byte by initializing a "StopRef" bit. The subscriber station acknowledges completion by clearing a "ContRe" bit; during the next burst of reverse lane.

The subscriber station then enters the voice operation.

When detecting the acknowledgment, the base station enters the voice operation.

 The refinement is suspended by the base station after 67 frames (3.0 seconds) if the goal of + 3% has not been reached. This is signaled to the subscriber station via the oder byte by initializing an "AbortRef" bit. It is acknowledged by the subscriber station in the same way as for the stop refinement. The subscriber station then stops the voice channel. When detecting the acknowledgment, the base station clears the voice channel.

 The base station transmits the termination order a second time if it fails to receive the acknowledgment of the subscriber station after the first transmission (ie RMU is not found or CRC is bad). If the base station still fails to receive the acknowledgment of the subscriber station after the second transmission, it automatically enters the voice operation if it has transmitted a "StopRef" bit or clears the voice channel if it has transmitted an "AbortRef" bit.

 The channel control task (TCV) 105 automatically enters a refinement operation upon reception to the assignment of a voice channel.

Upon reception of the base station refinement bursts, the subscriber unit uses the contents of a "Pwr" byte to correct its transmission power and a fractional synchronization byte to correct its symbol synchronization.

 The offset values of the fractional synchronizations (Li) from the base station are stored as they arrive. When five valid values are received, the subscriber calculates a variance of samples to determine their distribution.

In the case where the variance is too large, additional samples are collected. As soon as the variance is small enough, or when the count of valid samples reaches 16, the average of the samples (average A) is calculated and used to adjust the fractional synchronization signal transmitted to the base station. Following the adjustment, the buffering operation is repeated one more time.

 The subscriber station (TCV) 105 task successfully receives a refinement burst from the base station if the RMU is found and the CRC is correct.

The subscriber station ignores bursts of the base station that are not received successfully. The subscriber station also ignores the values of the fractional synchronization when it is instructed to do so by the base station. There is only one case where the subscriber station ignores the value of the power in the burst. This is the value of the power in the first burst received successfully (ie, this power adjustment could cause a "spike" effect in the next burst of the reverse path).

 The refinement operation is determined successfully on order from the base station. The voice operation begins immediately after the subscriber unit has acknowledged receipt of the termination command from the base station.

 The refinement is suspended after 78 frames (3 seconds) on order from the base station. In this case, the voice channel is defeated immediately after acknowledgment of the order from the base station. The subscriber station suspends the refinement of itself after receiving 77 frames (3.5 seconds) of bad bursts of refinement.

This bias allows the subscriber station to receive the "AbortRef" command that the time of the voice channel has expired and is defeated.

 Before an adjustment of the fractional synchronization at the subscriber station 41, the sample variance must fall below a certain threshold.

The determination of this threshold is somewhat arbitrary, but the following analysis provides us with a plausible threshold value.

 It is a fact that 75% of the samples in a random process are located in two standard deviations from the mean. Therefore, if the standard deviation calculated in the interval between -5%, + 5%] is found twice, it is known that 75% of the samples are within ± 5% of the average of the samples.

This provides reasonable confidence that the average of the samples is accurate and can be used for reaction fit.

Since the size of an adjustment step is T / 200, where T is the time of a symbol, the ir.terval / / -5%, + 5% / corresponds to / -10, + 10 / in incremental steps. Therefore, the standard deviation should be in the range -5, + 5 /, or, which is equivalent, the sample variance should be less than 25. The variance of the samples is easier to compute. than the standard deviation, so that it is used without the actual implementation. The formula is

<tb> V = <SEP>(<SEP>#<SEP><SEP> - <SEP><SEP>(#<SEP><SEP>(#<SEP> ti <SEP> - <SEP>#<SEP> t <SEP> average) <SEP>} <SEP> / (n-1) <SEP> Eq. <SEP> 1] <SEP>
<tb> where
"V" is the variance of the samples,
"E t." is the i-th sample of the value
calculated for time lag
fractional,
"n" is the sample size
"Average way" is the average # t calculated for
n samples.

 The refinement approach described here allows the acceleration of the operation under good conditions, while providing robust operation under adverse conditions. If the fractional time estimates are good, the refinement is completed in less than 4 frames (180 ms). In less ideal conditions, an average of 16 full frames may have to be computed taking about 19 frames (855 ms). The worst conditions would cause the algorithm to be at its upper limit of 67 frames (3 seconds) but it seems unlikely that voice operation is even possible under such extreme conditions (ie that is why refinement is suspended if the maximum count is reached).

 The general programs executed by the base station 104 and the subscriber unit 41 to complete the refinement of the synchronization are summarized in connection with FIG. 3. The subscriber unit 41 executes a program 134 for transmitting successive bursts of bursts. of refinement signals which are synchronized by the internal synchronization generator 113.

 The ETD set 14 of the base station executes a program 135 for processing each refinement signal burst in relation to the system timing signal from the master clock 18 to determine an offset value b = for each burst between the timing of the system timing signal and the timing of the refinement signal.

 The base station ECV 23 executes a program 136 to determine whether a predetermined number "n" of successively determined offset values Δt is below a predetermined U value. When the set 23 of the base station determines that the predetermined number "n" of successively determined offset values Δt is below the predetermined "U" value, it ε ε a program 137 of transmission to the subscriber unit 41 of a "refinement stop" signal. The PBB 112 of the subscriber unit 41 responds to the refinement stop signal by executing a program 138 which terminates the transmission of the refinement signal which returns an acknowledgment signal to the base station, and then executing a program 138a. The PBB 112 then executes a program 139 which allows normal communications through the given communication channel with the base station 104.

 The ECV 23 of the base station responds to the acknowledgment signal in the program 138 by executing a program 138b which allows normal communications through the communication channel with the subscriber station 41.

 The set 23 also executes a timing program 141 of the duration "D" of the program 136, in which it is determined if all the n successive values of the shift A t are smaller than the value "U" p-determined. such a determination has not been made within a predetermined time "S" ie D> S), the set 23 of the base station executes a program 142 which transmits an Abort signal (suspension) to the station. Subscriber 41. The PBB processor 112 of the subscriber station 41 responds to the Abort signal by executing a program 143 which returns an acknowledgment signal to the base station, and then executing a program 144 which undoes the signal. The ECV 23 of the base station responds to the acknowledgment signal from the subscriber unit 41 by executing a program 145 which undoes the communication channel given to the base station. based.

 Before the expiration of the predetermined time which is defined by the program 141 to time the program 136 by determining whether a predetermined number "n" of successively determined offset values Δt are smaller than the predetermined "U" value (c ie, D <S), and prior to determining that a predetermined number "n" of successively determined offset values ss t are less than the predetermined "U" value, the ECV set 23 of the base station executes a program 147 which transmits the offset value bt determined at the subscriber station 41.

 The PEB processor 112 of the subscriber station 41 executes a program 148 for calculating the average offset value (average E t) from the "m" last offset values ss t coming from the base station (less than the value the offset is not checked by being accompanied by a bit "SF ignored as described above) .This processor 112 further executes a program 149 to determine whether a predetermined number" p "of offset values A t checked, received, is within a predetermined tolerance R of the average offset value (average E t), calculated in accordance with the program 148.

 If the processor 112 determines, in accordance with the program 149, that a predetermined number of "P" reception offset values are within the predetermined tolerance R of the average offset value (average dt ), it executes a program 150 which adjusts the timing of the internal synchronization generator by the value calculated for the average offset (average E t).

 If the processor 112 determines, in accordance with the program 149, that a predetermined number "P" of received reception offset values are not within a predetermined tolerance of the average offset value (average ht ), the processor 112 executes a program 151 for counting the number of such negative determinations, and when a predetermined count "Q" corresponding to the predetermined duration is reached, the processor 112 executes the timer adjustment program 150 of the generator of internal synchronization by the calculated value for the average offset (! t means).

The subscriber communication system carries the DC signaling information between the interface 27 of the two-wire lines at the subscriber station 41 and the two-wire lines 26 at the central office 25. The information transferred to the The "reverse channel" direction between the subscriber unit 41 and the base station 104 includes changes in the monitoring state, the dial pulse counts, and the handset signals. Direct-channel direct-current signaling has features such as synchronous calling, distinctive calling, and room-by-room operation.
We want to provide transparency
as large as possible within the MTD nature of the system. The transparency of the signaling can be measured by quantifying the following performances: signaling path reliability, signaling delay, and signaling resolution.

 To optimize these parameters, the system uses a waveform coding method to digitally transfer the DC signaling information from the subscriber line interface 27 to the central office lines 26.

 The changes in the state of the handset are monitored by the baseband processor 112 in the subscriber unit 41. A timer interrupt in the baseband processor allows the handset state to be sampled every 1, 5 ms, or 30 times per MDT frame. Each sample is stored as a single bit (on-hook or off-hook) in handset status buffer (TEC) 114. Buffer 114 contains 60 sample bits, although typically only about 45 bits of these bit positions be actively used. The remaining bits provide an elastic overflow capability of the buffer. The nominal 45 bits provide a window of 67.5 ms for the handset status information. The TCA task 100 uses the buffer 114 to determine changes in the handset. Check status such as service requests, responses, and disconnections.

When a call is active, the buffer is also monitored for DC signaling events.

 A DC signaling event can occur only during an active voice operation. The buffer 114 is checked for events once per MDT frame (every 45 ms).

An event is detected by using a group co-. Starting at the 16th bit and going up to the 45th bit in the buffer 114, the group count is incremented for each handset bit hung up and down for each handset bit off hook. If the count reaches a threshold, defined by a terminal group count (Tcc) i a signaling event is declared. There is no chance that the group count will become negative or exceed Tcc. The group count is also kept within the frame boundaries, so that the flow of handset-related samples is seen as a continuum.

The group counting technique has the effect of detecting groups of on-hook handset states in the buffer 114 even in the presence of deformed signals. Acceptances of parasitic information are rejected on the basis of the choice of Tcc
As soon as no signal-event is detected, the next transmission burst is used as a command burst. Voice information in the salvo is replaced by signaling information using the current modulation level of the voice. The oldest 30 bits of the TEC data, representing 45 ms of handset status, are encoded in the burst.

If consecutive signaling events are detected in the TEC buffer 114, the control bursts continue to be sent in consecutive frames. Occasionally, an additional command burst is required that follows a sequence of one or more command bursts, although no signaling event is reported in that frame. The only condition in which an additional command burst is required is pro
when the previous command burst terminates with a handset bit hung up, leaving the base station 104 on-hook. If an additional command burst is required, the processor 112 of the base station 104 must ensure that the last state of the handset corresponds to a handset off hook so that the set (UCV) 24 returns to the off-hook state for the combined.

The first six words of each command block are specialized on an arbitrary profile of the flag This flag profile allows the command block to be detected during a normal voice operation
Following the flag profile are 14 words of the DC signaling data. The words are organized into seven sets, each set comprising two information words. The least significant bit of each word contains no information, and is arbitrarily initialized to zero. However, these bits are used, and should be, in error detection. The other 15 bits of each of the words in a set collectively contain 30 bits of handset status information. The information is stored chronologically between the first word and the second word in the set and between the weight and the least significant bit in the words. To avoid false decisions due to repeated but incorrect profiles, each set undergoes an Exclusive OR operation with a single bit profile.

 The set (UCV) 24 of reception in the base station 104 determines the presence of a control block, as opposed to a voice block, by a simple majority vote decision on the words of the profile of the flag at the head of the block. If the voting threshold is exceeded, the block is declared as an order block.

The PLER synthesis continues during the processing of the control block and the normal PLER data is replaced by a PLER silence.

 As soon as there is detection of a command block, the information on the signaling state it contains is also decoded using a simple majority voting decision. The OR Exclusive Transformation must be eliminated before the voting operation. If the fast does not exceed the threshold, the block is redenté and no change is brought to the state of the combines.

 As soon as the 30-bit content of the buffer 114 is decoded by the set (UCV) 24, it is translated into Tl A / B signaling bits. In the case of the state of a two-wire handset, the TEC bits correspond exactly to the required 30 bits of the A-bit signaling data.

 The T1 bits A are placed in a first-in-first-out queue for transmission by a PCM channel to the base station 104. The corresponding multiplexer 19 provides the CPU of the set (UCV) 24 with an interrupt just before the A-bit T1 signal frame, allowing the processor to inject the appropriate signal bit into the correct MIC byte.

 When no control block is able to fill the A bit queue, the oldest state is repeated indefinitely. In the case of a signaling operation the task (TCA) 100 ensures that the last state in Xe buffer 114 corresponds to a handset off-hook.

 Following the establishment of a call, the set (ECV) 23 in the base station 104 initializes the set (UCV) 24 to the off-hook handset state.

In the case of launching a call, the set 23 places the set 24 in the off-hook handset state just before the completion of the refinement. In the case of completion of a call, the set 23 places the set 24 in the handset off-hook state after detecting a response. Command bursts are not used for these control state transitions.

 Once a voice operation is established, the command bursts are used to transmit signaling events to the base station 104. If a disconnection is detected at the subscriber unit 41, the signaling state of the set ( UCV) 24 of the base station is left in the on-hook state, while the call is cleared via an erase request burst VCR.

By appropriately choosing the signaling parameters, it is possible to adjust the system performance. To aid in the detection and correction of errors, the flag profile and the TEC majority votes are taken on eight-bit segments (aligned with the byte boundaries). The majority vote of the flag profile is taken on 12 complete bytes. For TEC 114, there are four independent major votes, one for each multiplet it contains. If one of the votes is defective then the totality of the votes is considered as having failed. The values chosen for the parameters are the following ones:
Counting a terminal group - 15
Majority Vote Flag Vote - 6 at
within 12 (bytes)
Majority vote TEC - 4 within 7 (mul
tiplets)
The selection of the end group count represents a compromise between the rejection of the acceptances of a parasitic information and the faithful reproduction of the signaling pulses. The minimum duration of a significant on-hook pulse is 19 ms, produced by a dial operation of 20 pulses per second with a cutoff of 58%. With a group count of 15, acceptances of parasitic information less than 22.5 + 1.5 ms will be rejected. This rejection threshold is well below the required 29 ms, which corresponds to 18.5 times a sample. The threshold also implies that a control block transmits only if at least 50% of the MDT frame is taken with the handset on the hook. The 45-bitTEC contains 67.5 ms of signaling status information, providing 22.5 ms of pre-analysis data for the buffer to make a "go-ahead" decision. pre-analysis, it would not always be possible to send the leading bits of the handset state transition in a timely manner.

 The threshold of the flag profile is central in order to avoid false detections of the command block and the missed command blocks. Although undesirable, the false detection of a command block during a normal voice operation is not critical for the system. A false detection only results in a salvo of silence of 45 ms and the distant possibility of the acceptance of parasitic information in the lines of the central office. Much less acceptable would be the loss of a command block, or even worse, a command burst, as this causes the ability of the subscriber to fail to perform a sionalization. With these points in mind, the threshold of the flag profile is set to six (fixed position) byte matches (multiplet limit within 12 possible. The probability that this will occur in a random noise (the PLER data ap -12 appears as a white noise) is 3.2 X 10 12 -With a block transmission period of 22.5 ms, such a match has an expected rate of occurrence of once per 200 years of voice operation con-.

bare. The analysis of control block loss is somewhat more difficult, especially if one assumes that errors occur in bursts, but it is suggested that this detection method provides a good reality.

 The threshold of the majority vote TEC makes it possible to correct the errors in the signaling data.

Since the signaling bits are stored in sets, corresponding TEC words are separated by many bit positions. This natural nesting allows a burst error to obliterate three whole sets while not corrupting the majority vote.

 Voice paths that provide acceptable voice quality will also provide highly reliable signaling using this technique. The resolution of the signaling for the system is 1.5 ms. This corresponds to a T1 A / B signaling resolution and therefore represents an acceptable level. The signaling delay in the system is approximately 80 ms. This delay consists of the TEC window of 67.5 ms, a transmission time of six milliseconds and the processing time of the base station. These measurements make the signaling transparency of the system comparable to that of existing digital loop-carrier systems.

 Identically, the signaling information can be transferred from the lines 26 in the base station 104 to the line interface 27 in the subscriber unit 41.

 The general programs executed by the base station 104 and the subscriber unit 41 for detecting and transmitting the signaling information over an assigned voice channel are summarized in connection with FIG. 4, is either the base station 104 or the subscriber station 41 in accordance with the site at the origin of the signaling information and the receiving station 156 is the other of the two stations .

 The transmitting station 155 executes a program 158 for signal monitoring at the line / line interface of the central office; performs a 1S9 program for buffering the monitored signals in accordance with program 158; executing a program 160 for detecting the signaling information from the buffered signals in accordance with the program 159; executes a program 161 for conditioning the detected signaling training for communication by the affected channel instead of the voice data signals by formatting the detected signaling information into a control block having a flag profile in addition to the signaling data; and executes a control block transmission program 162 in the form of a control signal burst in the assigned communication channel instead of the voice information.

 The receiving station 156 executes a program 166 for determining the presence of a control block in a signal burst received through the assigned communication channel by recognizing the flag profile in the signal burst.

The receiving station 156 then executes a program 167 for reformatting the signaling information in the control block to a standard format for transfer to the central office / interface lines. Finally, the receiving station 156 executes a reformatted signaling information transfer program 168 to the central office / interface lines of the receiving station 156.

 The present invention is not limited to the embodiments which have just been described, it is instead capable of modifications and variants that will occur to those skilled in the art.

Claims (9)

 a means coupled to the internal synchronization generator for processing the offset value communicated by the base station in order to adjust the synchronization signal of the set station and to reduce the offset.  where the subscriber  means for communicating to the subscriber unit the determined offset value; and  means for processing the refinement signal from a subscriber station in relation to the system synchronization signal so as to determine the value of any offset between the synchronization of the system synchronization signal and the synchronization of the signal of the system. refinement;  where each base station further comprises  means that can be implemented during the initial establishment of a communication channel between a base station and a subscriber station in order to transmit the refinement signal via the respective communication channel between the subscriber station and the base station;  means for providing a refinement signal synchronized by the synchronization signal of the subscriber station;  an internal synchronization generator for producing a subscriber station synchronization signal for synchronizing the signals that are transmitted by a given communication channel between each subscriber unit and a base station; and  each subscriber station comprising:  a mother clock for providing a system synchronization signal; and  each base station comprising  A subscriber communication device comprising a plurality of base stations (104), each in a separate network, the base station of each. reed being in selective communication with a plurality of subscriber stations (41) and having a node for transmitting control information to subscriber stations by a radio control channel (VCR) at a frequency selected by the station basic among a multitude of predetermined frequencies;  2 - Device for co. = Communication according to claim 1,  wherein the means for transmitting the raffinate signal of the subscriber unit transmits the successive bursts of bursts of the refinement signal;  wherein the base station processing means determines an offset value for each burst received from the refining signal; and  wherein the subscriber station processing means processes the predetermined number of successively received offset values to adjust synchronization of the synchronization signal of the subscriber station.  The communication device of claim 2, wherein each base station includes means for terminating communication of the determined offset values when a predetermined number of successively determined offset values are below a predetermined value.  4 - Communication device according to claim 3,  wherein the communication means of the base station communicates a stop refinement signal to the subscriber station when a predetermined number of successively determined offset values are below a predetermined value; and  wherein the means of transmitting the refinement signal of the subscriber station responds to the reception of the stop refinement signal by ceasing transmitting the refinement signal.  5 - A communication device according to claim 4, wherein the subscriber station responds to reception of the stop refinement signal by permitting normal communications through the encryption channel given to the subscriber station.  The communication device according to claim 2, wherein the base station comprises means for undoing the given co-channel when the determined offset values remain above a predetermined value for a predefined duration.  The communication device according to claim 6, wherein the communication means of the base station communicates a suspension signal to the subscriber station when the set offset values remain above the predetermined value for a predetermined time;  the subscriber station responds to the reception of the suspension signal by sending an acknowledgment signal to the base station and by disabling the communication channel given to the subscriber station; and  the base station responds to the reception of the acknowledgment signal by undoing the communication channel given to the base station.  8 - A communication device according to claim 2, characterized in that the processing means of the subscriber station calculates an average value of the offset from a predetermined number of processed offset values and provides the calculated average offset value so .adjusting the internal synchronization signal generator of the subscriber station when the predetermined number of processed offset values is within a predetermined tolerance of the calculated average value for the offset.  9 - Communication device according to claim 8, characterized in that the processing means of the subscriber station provides the calculated average value for the offset in order to adjust the internal synchronization signal generator of the subscriber station after a predetermined time when the predetermined number of offset processed values are not within the predetermined tolerance for the calculated average offset value.