Synchronization method for radio transmitters in a local e.g. nationwide paging network.

Method for synchronizing radio transmitters in a regional, eg national paging network. - Synkroniseringsförfarande för ett lokalt, t.ex. riksomfattande personsökarnäts radiosändare.

The invention relates to a method for synchronizing radio transmitters of a regional, e.g. national paging network, which paging network is connected to the public telephone network and comprises paging messages in the forwarding order:

- a network controller with means for receiving, processing and sending messages in digital form and means for manually entering network parameters,

- base station controllers which communicate with the network controller and have means for receiving, processing and transmitting messages in digital form,

- radio transceiver pairs of base stations, the transmitters of which, under the control of said controllers, transmit radio signals for both paging and synchronization of base stations covering a common coverage area.

First, the flow of paging information within the system is generally considered. The paging is started by selecting the remote paging network identification number and the paging receiver identification number made from any telephone in the public telephone network. The call is routed based on the remote paging network identification number to the remote paging terminal, which receives the called number and forwards it to the network controller of the base station network in its immediate connection. The data transmission between the remote paging terminal and the network controller is serial digital information at 512 baud. The format of the data transfer in this transfer is in accordance with the POCSAG (Post Office Code Standardization and Advisory Group) code structure.

The network controller, group controllers, and base station controllers are connected to each other via two-way modem connections. The data2 transmission method is a synchronous message-oriented protocol operating at a speed of 2400 baud, in which the correctness of the data transmission is checked. The message frame itself conforms to the SDLC standard, which is a structure according to CCITT Recommendation X.25 level 2

The network controller sends the remote paging numbers received from the remote paging terminal to the so-called as a remote paging message to group controllers, which in turn forwards them to base station controllers. With the base station controllers, the messages are decoded and the search numbers are formed into codewords according to the POCSAG standard. The code words are formed into a record, which is sent by radio to be interpreted by the receivers carried by the retrieved persons. Thus, paging messages can optionally be sent to individual paging receivers at any time from any telephone set in the public telephone network. The search receiver has both a display and an audio signal output. The conventional paging may optionally comprise e.g. four different messages, i.e. the paging receiver generates one of four different alarm tones, each of which has a pre-agreed purpose. It is also possible to use the system to send 10-digit number information to the paging receiver's number display (indicating, for example, the caller's telephone number).

The base station controller controls the radio transmitter section so that the information bits of the POCSAG records are transmitted in series at 512 baud to the receivers. In radio transmission, direct FSK (Frequency Shift Keying) is used for modulation with a deviation of +/- 4.5 kHz. The carrier frequency is 146.325 MHz. A positive frequency deviation represents logic 0 and a negative frequency deviation represents logic 1.

In a relatively dense base station network, the problem with the radio transmission method described above is that the remote paging receiver may hear transmission information from several transmitter stations simultaneously as sum information, the accuracy of which is compromised the more its components are phase-shifted. The problem caused by phase differences

7145 has been studied under laboratory conditions. These studies have shown that the reception of messages is disturbed by the worst-case operation of the two transmitters under the following conditions:

(a) the intensity of the radio signals at the receiver is uniform within 3 dB

b) the data signals have a phase difference of at least 1/4 bit.

a) The condition of the item is practically fulfilled when the receiver is just in the common coverage area of ​​the two transmitter stations. In terms of synchronization, the situation is worst when the attenuation of the signals from the transmitters is mainly based on distance attenuation (free path attenuation). Taking into account the transmission powers used, the wavelength and the practical signal-to-noise ratio, such a worst-case coverage area can be estimated to be about 3 km.

In order for the data signals (512 bit / s) to have a phase difference of no more than 1/4 bit, the transmitters should be synchronized with each other so that the data signals from the antennas of nearby stations do not deviate by more than +/- 239 ps in the coverage area about +/- 5 ps). The synchronization accuracy of distant transmitters is irrelevant to interference.

Thus, the problem of common coverage area interference can be eliminated by synchronizing the paging transmission signals so that there are no significant phase differences between the stations in close proximity.

Several different types of solutions have already been studied in the past for synchronizing the entire base station network. Examples of these are time division transmission technology and the so-called line synchronization, the main advantage of which is easy theoretical implementation.

Time-sharing transmission technology is not really a question of synchronization at all, because it only circumvents that problem. The solution is based on the fact that the messages are transmitted during four time slices with about 1/4 the number of base stations at a time. If the stations operating simultaneously in each transmission are correctly selected, they do not have common coverage areas and the problem is avoided. The division into four is based on the fact that it is possible to implement the above with a minimum of four time slices. The main disadvantage of this technique is that the transmission capacity of the entire base station network is only 1/4 compared to the synchronized solution.

Line synchronization means that the transit time delays in the data transmission connections between the remote paging center and the base stations are arranged to a certain standard value. If the information to be transmitted is transmitted synchronously in such a system, the information leaving the base stations is also synchronized. The disadvantages of the method are e.g. the following:

- to limit the structure and scope of the data transmission network, as it operates in practice only on direct connections without the use of modern

- is difficult to maintain and service.

The structure is limited by the fact that anything other than a direct connection between the base station and the terminal is practically out of the question; e.g., a multi-level wide area paging network is not possible. The use of moderns is also out of the question because they operate in a fairly low frequency range (1000 Hz to 2000 Hz) and their interactions at both the transmitting and receiving ends are not synchronized (phase locked). The errors they cause are too rough for the synchronization requirements. Service and maintenance are caused by the fact that delay-stable connections often do not exist in practice and the situation created by special equalizers changes as a function of time and environmental conditions, causing a constant need for adjustment and maintenance. The connections must also be selected in this case, and arbitrary back-up connections in the event of a fault are therefore out of the question.

The radio connection between the base stations is the most excellent data transmission mechanism for synchronization, because the temporal phenomena occurring in it are sufficiently accurate, relatively stable and well known. In principle, radio synchronization between base stations is known, but in practice it has not yet been possible to develop a radio synchronization method which can also manage wide network synchronization and perform synchronization with such precision that the required synchronization interval becomes long enough so that synchronization is not significantly reduced.

Synchronization by radio is performed with the same frequency as the actual transmission of remote paging messages. By reserving the radio path for its own use, synchronization reduces the transmission capacity of the remote paging. This disadvantage can be minimized in two different ways: a) by trying to keep the need for synchronization as small as possible, i.e. as long as possible between synchronizations, and b) by performing synchronization as quickly as possible. In addition, synchronization must be performed in such a way that the search receivers do not misinterpret the synchronization as a search.

The object of the invention is to provide a radio synchronization method for a paging network of the above-mentioned type, which - without imposing special requirements on the hardware implementation of the paging network - enables even extensive network synchronization management and synchronization with such precision that the synchronization

To achieve this purpose, the synchronization method according to the invention comprises the following measures: a) a synchronization plan is performed by the common network controller (PNU) of the base stations, comprising:

- set the synchronization route

- select a base station to start synchronization

- assigning a route identifier to each base station

714 5 2

- determining the distances between base stations along the route

(b) a synchronization command is sent from the common network controller (PNU) of the base stations to the base station controllers (TSIs) over the communication network, which synchronization according to the plan includes:

- whether or not the receiving station of the command is the initiator of the synchronization

- a route identifier by which the station to be synchronized recognizes and accepts only the synchronization signal received along the specified route

- information about the travel time delay of the synchronizing signal

c) transmitting a synchronization signal from the transmitter of the base station selected as the synchronization initiator in accordance with the synchronization command;

d) receiving the synchronization signal at the receivers of the base stations in its coverage area and comparing the identifier of the received signal with the identifier given in the synchronization command

e) if the comparison in d) shows that the received synchronization signal is intended for synchronization of the receiving base station, synchronizing the clock / signal transmitter of the receiving base station with the synchronization signal,

f) switching off the transmitter of the base station designated as the initiator and switching on the transmitters of the newly synchronized base stations to transmit a synchronization signal of the second synchronization sequence,

g) synchronizing the base stations to be synchronized during the second synchronization sequence according to d) and e), with the addition that the synchronizing base station's own receiver listens to the signal transmitted by the same station and before transmitting the actual synchronization message the phase of the signal transmitted is corrected, i.e. corresponding to the sum of the delays at the transmitter and receiver and the transit time delay given in the synchronization command,

h) performing the necessary number of synchronization sequences until the last base stations have transmitted a synchronization signal which is no longer received by the other base stations, but performing a phase correction of the clock / signal transmitter of the own station to compensate for the delay and transit time delay in the transceiver pair.

In order to achieve the object of the invention, it is also important that the network controller executes a synchronization plan comprising determining the synchronization order of the base stations based on the mutual location of the stations so that the number of sequences required for synchronization is as small as possible. and that the network controller performs determining the time of transmission of the actual pagings so that the synchronization of the transmission is independent of the structure and characteristics of the digital communication network.

The synchronization accuracy can be improved by entering information about the distances between the base stations to the network controller, and that the synchronization command sent from the network controller to the base station controllers via the digital communication network includes information about the distance of the synchronizing base station from the base station to be synchronized. wherein in the above-mentioned method steps g) and h), the travel time delay between the base stations generated in the previous synchronization sequence is added to the phase correction of the signal to be transmitted.

According to a preferred embodiment of the invention, the route identifier is the start time of the synchronization execution, which in the case of the base station selected as the initiator is the time when the initiating station transmits the synchronization signal and in the case of other base stations the time when the synchronization message arrives at the base station.

7a

In the following, an embodiment of the invention will be described in more detail with reference to the accompanying drawings, in which Figure 1 shows a block diagram of a paging network according to the invention and the connection of the paging network to the public telephone network.

Fig. 2 shows a block diagram of a network controller belonging to the paging network according to Fig. 1.

Fig. 3 shows a block diagram of a group controller belonging to the paging network according to Fig. 1.

Fig. 4 shows a block diagram of a base station controller of the paging network according to Fig. 1.

Fig. 5 shows the transceiver pairs of the base stations of the paging network according to Fig. 1, as well as their data transmission connections and the delays occurring therein.

Figure 6 schematically shows the generation and compensation of the phase difference caused by delays.

Figure 7 shows the time distribution of the synchronization signal transmitted and received by the base stations.

Referring to Figure 1, the general structure of the paging network will be described first. Search calls are made over the Public Switched Telephony Network (PSTN). Invitations can be sent from any telephone exchange in the public telephone network.

The remote paging center PT (Paging Terminal) acts as an interface between the public telephone network and the paging network. All search invitations are reviewed and accepted by the Remote Search Center PT. For this purpose, it shall keep a record of all holders of paging receivers and their specific forms of service.

The network controller The PNU (Paging Network Unit) controls the entire paging network, which consists of Transmitter Group Controllers (TGCs) and Transmitter Site Interfaces (TSIs). Search network operation and maintenance functions are handled through I / O devices connected to the network controller PNU. Network synchronization is also initiated by the network controller PNU.

In the block diagram of Figure 2, the right master central processing unit (CPU85B) controls the entire network controller and comprises memory, timers and serial interfaces for the alarm printer and display terminal. The slave central processing unit (second CPU85B) receives paging messages from the remote paging center PT and queues them in the buffers of the MEM64 memory. The paging messages sent to the group controllers TGC are processed by a number of special I / O controllers (SCC25), each of which includes a microprocessor and a buffer memory. The network controller regroups the paging message streams and feeds them in queue format for transmission to the group controllers TGC. The network controller controls the operation and synchronization of the search network. It instructs the transmitters to start a periodically repetitive radio synchronization process.

Figure 3 is a block diagram of group controllers, showing that the group controller is based on the same telecommunications modules used in the network controller.

One SCC25 is used to connect to the network controller PNU (both channels are used if the PNU is duplicated). With the remaining 7 SCC-25s, up to 56 base stations can be connected to the TGC (four base stations can be connected to one serial channel). The TGC of group controllers is responsible for controlling one node in the network; they check the correctness of the data packets received from the network controller, feed them in queue form towards the base stations and ensure the correct reception of the data packets at the base stations. Group controllers TGC also play an important role in the synchronization preparation phase as a transmitter of command information to base stations. Small search engines can also be built without group controllers.

Figure 4 shows a block diagram of the base station. The base station is divided into two separate parts, namely the base station controller TSI (Transmitter Site Interface), which is the furthest part of the paging network, and the transceiver equipment, which contains the actual radio frequency devices. The TSI base station controller comprises two modules: a TSI85A central unit and a standard V.26 card modem. The TSI85A is a further developed version of the standard CPU 85B central unit board, which is adapted to control the radio transmitter and alarm lines (cf. Figure 5). The TSI85A includes all the circuits needed to provide a smart base station. One of its most important tasks is to perform all the functions required for successful synchronization. The support weapon ground controller recognizes and receives from the multipoint modem line those messages that are specifically intended for it. The paging information contained in the messages is converted to the POCSAG format, according to which the RF transmitter is then controlled. The message received from the group controller may also be of a nature other than paging information, e.g. a command to the base station controller to notify its status to the group controller, activate or deactivate the base station, participate in network periodic radio synchronization, etc.

Next, the execution of the radio synchronization of the paging network by the method according to the invention will be described.

The network controller PNU is responsible for initiating synchronization and related planning. For this reason, the following information is given to the network controller as parameters during system configuration:

- desired synchronization interval

- the identifiers of the other base stations in the coverage area of ​​each base station and the distances to them at a resolution of 1 km

The network controller PNU determines from the received coverage information (distances between base stations) whether a nationwide system can be synchronized as a single area using existing connections between base stations. If this is not the case, at the same time those independent so-called isolated areas that can be synchronized internally but have no connection to other areas. The application of the so-called isolated areas will be described in more detail later. It should be noted in this connection that the smallest such isolated area is a single base station with no connection to other base stations. Given the reception conditions between base stations (in addition to station distances, receiver sensitivity, used antenna heights, etc.), it can be concluded that there is virtually no problem of mutual isolation between such isolated areas for the paging receiver, so they can be ordered to operate independently. In practice, this means that there can be several so-called synchronization initiators - each in an isolated area of ​​their own.

Thus, based on the above, the synchronization design phase also includes the search for a starter in each isolated area. The criterion in this definition task is to select such an initiator that the number of messages transmitted from one station to another during synchronization is minimized. This means that in each synchronization sequence, the number of base stations to be synchronized in parallel at the same time is on average as large as possible.

In fact, the synchronization is triggered by the network controller PNU transmitting a command over the communication network to the base station controller (s) of its choice to start the synchronization at a certain point in time. The other base station controllers of the paging network or its isolated area, in turn, are given a command to prepare to receive the incoming synchronization information at a certain point in time according to a plan made in advance by the network controller. For base stations in a word synchronization sequence, this time is the same, however, with a very rough accuracy (1/16 sec.) Compared to the accuracy at which the 512 baud signal must be synchronized. The synchronization sequence lasts for 8/16 seconds, so the time points of the commands received by the stations to be synchronized in succession differ by 8/16 seconds. The time instant acts as an identifier of the synchronization message during the synchronization, thus ensuring that the synchronization signal came along the planned route.

In addition to the time, the receiving stations of the synchronization message received by radio are conveyed the distance from which the synchronizing station is located, whereby the signal travel time delay can be compensated computationally at the receiving end, as will be described in more detail later. The identifier (time) in the synchronization message ensures that the synchronization route and the corresponding distance correlate correctly.

Before starting the actual synchronization, however, after the synchronization command messages transmitted in the communication network, the network controller PNU informs the base station controllers TSI of its own time with a resolution of 1/16 second. The purpose of this time is to keep the time of the synchronized network and the time of the network controller approximately the same (to the nearest second). However, the network is precisely synchronized specifically to the time of the initiator. Possible repairs

7145 2 in the approximate time before the start of the synchronization period, the base station controllers always have a resolution of 1/16 second, whereby in connection with the resynchronization information about a running error (of the order of tens or hundreds of microseconds) is obtained.

Receiving base stations switch off their transmitters and their synchronization receivers, respectively, shortly before the start of synchronization. At the start of synchronization, the selected initiator sends a 32-bit synchronization message. The same message is repeated for safety. The receiving stations synchronize to this message as will be described later.

All those receiving stations that hear this message correctly and recognize the time information in the message as the same (route identifier) ​​that came from the network controller during the synchronization start command synchronize the operation of their internal clock to the change edges of certain bits of the message.

The receiving stations then turn off their sync receivers and turn on their transmitters. Exactly 5/16 seconds of time have been set aside for these operations in the synchronization sequence.

Stations thus synchronized, in turn, become synchronizing stations that send a 32-bit message twice, respectively, at the start of the next synchronization sequence. The initiator will no longer participate in the broadcast at this point.

The synchronization sequence described above is repeated according to a predetermined plan so many times that the synchronization proceeds as a wave motion over the entire isolated area. Thus, during the whole synchronization phase, each station synchronizes once to the pre-announced message and sends once the synchronization message itself.

The following is a closer look at the synchronization message to be transmitted between the two base stations.

The synchronization between the two base stations thus takes place in such a way that the synchronizing station sends a special 32-bit synchronization message. The content of the message is as follows:

IMF! REV! 3! CNTR! CRC

MF: Length 1 bit (PO). Always 1 so that receivers do not misinterpret the message as a search.

REV: Length 10 bits (D1-D10). Alternating zeros and ones to bit synchronize the receiving station. Starts with zero and ends with one.

E: Length 1 bit (Dll). A separator that is always one and thus forms two consecutive ones with P10.

CNTR: Length 9 bits (D12-P20). Synchronous time counter with a resolution of 1/16 second. Indicates the time that has elapsed since the start of the synchronization and at the same time the isa.nom.an belonging to the recipient. After all, the recipient received the time information from the network controller in connection with the synchronization command. It takes 1/16 second to transmit 32 bits at 512 baud.

CRC: Length 10 bits (D21-D30). Checksum according to the POCSAG code standard.

EP: Length 1 bit (D31). Even parity from D0-D31.

The numbering in the following representation refers to the accompanying Figure 7.

1: The synchronizing base station switches on its own transmitter well in advance (3/16 seconds) before the start of the actual synchronization message transmission.

2: Before the power level of the transmitter may have stabilized, a half-frequency preamble pattern is already applied to the data line. Search receivers cannot misinterpret this for two reasons:

a) the so-called message flag is number one

(b) the checksum is incorrect

Also, the station to be synchronized cannot misinterpret this because it is looking for the correct frequency preamble pattern at this point. The purpose of this transmission is to ensure the operation of both the home and receiver synchronization receivers Rx so that the actual synchronization message is heard as correctly as possible.

3: The own receiver works properly at this stage at the latest, so the necessary transmitter-receiver pair delay measurement is then possible.

4: The actual sending of the synchronization message starts. A normal introductory pattern is transmitted for 1/16 second. In this connection, a measurement is also performed to determine the signal propagation time delay in the transmitter and receiver as a whole. This delay is described in the calculation example to be presented later by the term: ts + tap + tr + tag (cf. Figure 5). The delay is compensated together with a possible mobile delay (e.g. tn + tbd) by correcting the phase of the data to be transmitted by an amount corresponding to the phase. An exception to the above is the synchronization initiator, which does not explicitly perform delay measurement and compensation, as shown in the calculation example. The reason for this is that the clock running error thus generated during the synchronization interval can be obtained by measuring (step 10).

and 6: Sending a synchronization message (2).

7: When even the last bit (parity) has been transmitted, another 2/512 seconds are expected, after which the transmitter is switched off. The situation for the synchronizing station is now over.

8: The station to be synchronized is constantly heard. The incoming signal is not yet fully certain at this stage.

9: The synchronization receiver hears the incoming transmission, which at this stage is still a half-frequency preamble.

10: With the help of the introductory pattern, the receiver performs a phase comparison measurement in which the phase difference between its own transmission clock and the incoming data is measured. This measurement result is computed computationally to obtain the final result (run error during clock synchronization interval) and then to report the run error to the network controller.

The beginning of the introductory figure is discarded due to a phase correction made by the sender (cf. point 4 :). The preamble pattern continues in the actual synchronization message as described above, and the actual bit synchronization (exact synchronization) is performed at the change edge of the incoming data during the preamble pattern. The synchronization is still checked during the introduction pattern, and if the accuracy is found to be too poor, the bit synchronization is repeated. After synchronization, the incoming data is sampled in the middle of the bit. This is the 1st step of synchronization.

11: Synchronization messages are received. Message level synchronization is done using a separator (B). This is step 2 of the synchronization. The recipient has the distance received from the network controller in connection with the synchronization command, as well as the time (D12-D20 in the first sequence message) at which the synchronization message intended for it arrives. If the time in the incoming message matches and the message has been found to have been correctly received by another check, it is finally decided to synchronize it. This is the last step of synchronization, i.e. step 3.

If the time reading of the message did not match, the situation is checked during the second message, but the synchronization is not left valid because it was not intended for this station and the given distance correction would therefore not be correct. If synchronization is not achieved for any of the above reasons, the operation is resumed from the beginning according to step 8.

12: After reading the parity bit, which is the last bit of the message, 2/512 seconds is waited and the receiver is switched off.

13: After the receiver has been switched off for 2/16 seconds, the transmitter is switched on.

14: The synchronized station has thus itself become a sender and the situation is repeated in accordance with points 2: to 7 :. After performing this transmission, the base station controller turns off its transmitter and waits for the synchronization end command from the network controller and any further searches to be transmitted.

To retrieve isolated areas, proceed as follows:

The memory of the network controller has, for illustrative purposes, a matrix formed from the given output data. The matrix is ​​marked 1 for all Active stations that hear each other, while otherwise it is marked 0. In practice, the above is an input data table in the network controller memory, but is presented here as a matrix for ease of understanding.

E.g.

! 1 2 3 4 5 6 ! - 0 1 0 0 0

10-0101 3 ! 1 0 - 0 1 0 . 4 ! 0 1 0 - 0 1 ! 0 0 1 0 - 0

10 10 10The corresponding geographical location could be:

2 * *

6

The network controller generates a vector (also in the figurative sense) with as many elements as there are base stations in the system for retrieving areas. At this point, the value of each item is 0.

2 3 4 5 6 !0!0!0!0!0!0!

E.g.

Retrieval of isolated regions is accomplished by going through the above matrix and marking the vector with the existence of the connections shown in the matrix.

The search is started from the first base station by marking in the vector the base stations to which the connection exists. In the case of our example, the entry would thus be for elements 1 and 3 of the vector.

After this, the intention is to search for further connections that the already marked access points have. For this reason, the next position to be annotated is determined by the contents of the vector with the matrix acting as a source data table. In the case of our example, further connections are thus sought for station 3. As can be seen from the matrix, there are 3: 11a connections to 1 and 5, so the vector is marked as new contact element 5.

Subsequent connections are marked as described above. Thus, in the case of our example, for base station number 5. It is observed from the matrix that it has a connection to station 3 and thus no new contact information enters the vector.

At this point, it can be inferred from the contents of the vector that there are no more unprocessed downlink stations. This means that the isolated area has been determined. In the case of our example, the region is formed by positions 1, 3, and 5, according to the notations of the vector elements.

On the other hand, it can also be seen from the vector that not all base stations have been processed yet (items 2, 4 and 6 are unmarked), so the search still needs to be continued. The search for the next isolated area can in principle be started from any unmarked base station. In our example, the first unmarked is station number 2. In a manner similar to retrieving the first region, the matrix is ​​detected as access stations 4 and 6, so they are marked in the vector.

Now, in turn, the situation is such that there are no more unprocessed stations in the vector of the example, so there is no need to find out further connections. It can be seen that the second isolated region consists of positions 2, 4 and 6.

By way of example, the determination of isolated regions for a uniform region is performed as described above.

The geographical location of the stations could be, for example:

2 * *

6 * * *

The corresponding matrix is:

2 3 4 5 6 ί - 0 1 0 0 0

10-0101

110-110

10 11-01

10 0 10-0

10 10 10Start the search again from station 1, for which the access point according to the matrix is ​​3. The only possible way to find out connections is now station 3, for which the access points according to the matrix are 1, 4 and 5. According to the vector entries 5. Let us first consider even 4, for which the connection stations according to the matrix are stations 2, 3 and 6.

It is now observed that there are no more free elements in the vector so retrieval can be stopped.

As a result, all stations 1-6 form a single isolated region that can be fully synchronized.

The above examples also illustrate quite clearly why the determination of isolated areas is necessary at all and why they have to be synchronized separately.

Isolated areas are defined by the network controller whenever there is a change in the network situation between synchronizations (configuration change, a station deactivated, etc.). This need easily becomes apparent if we imagine that base station number 4 in our latter example is deactivated or deleted altogether. In this case, the system consists of two isolated areas instead of one whole. It should also be noted that, although the retrieval of areas is based on the determination of downlink connections, it does not as such determine the most optimal synchronization path.

To specify the initiator, proceed as follows

Once all the isolated areas of the network have been defined, an initiator must be defined for each area from which to synchronize the area. The criterion for selecting an initiator by the network controller is to do so in such a way that as few synchronization sequences as possible are needed to synchronize the isolated area.

From the first example presented above, we can easily see that for the isolated area formed by stations 1, 3 and 5, it is not advisable to choose 1 or 5 as the initiator, because the synchronization would proceed only one station at a time as if in a queue. After all, the whole area will be synchronized with only one synchronization sequence, when the number 3, which has a connection to both 1 and 5, is specifically chosen as the initiator.

The matrix shown in the example can be formed into a tree-like pattern in which the branches represent synchronous messages.

E.g.

»1 2 3 4 5 6! - 0 1 0 0 0! 0 - 0 1 0 1! 1 0 - 1 1 0! 0 1 1. - 0 1 ί 0 0 1 0 - 0! 0 10 10a)

2

<img file="FI71452C_D0001.tif" />

b)

<img file="FI71452C_D0002.tif" />

Different oxistor structures can be drawn for the structure according to the example, depending on the desired synchronization paths. The task of the network controller is simply to define a branch structure that is as branched as possible and the number of consecutive branches is minimized. As we can see from our example, the structure shown in Figure a) is more advantageous than that shown in Figure b) because only 2 sequences are needed for synchronization: 1 .: 3-4, 3-1, 3-5,

2.: 4-2, 4-6.

The most optimal structure is found as follows:

Mark station 1 the stations to which it is connected. This is the first sequence of synchronization. These stations, which are already synchronized (level 2), are then marked with access points. This is the second sequence of synchronization. The markings are continued from level to level as described above until the entire isolated area has been traversed. It is then memorized how many sequences are needed to synchronize the network if the initiator is station no. 1. The entire set of markings described above is repeated for each station in the area. The initiator is, of course, the one with the smallest number of synchronization sequences.

The synchronization route is already determined in the starter selection, as described above, level by layer. In the first sequence, the sender is the initiator alone. In the following already all those joi'la had a connection with the beginner. The number of transmitting stations generally increases as synchronization progresses. In this case, it is also possible that situations arise where one station hears two stations transmitting simultaneously. However, the information they send is, of course, exactly the same, because the content of the message is tied to the time of transmission.

E.g

In our example, we must assume that the other structure of the network is such that it causes synchronization to come from station no. 5 through.

Now that stations 2 and 3, after synchronizing, send a simultaneous message, 1 hears both. In order for 1 to be directly disturbed, the phase difference of the messages must be at least 488 μs and the same within the limits of intensities of 3 dB. This situation is practically impossible due to distance (3dB = 1.4 km = 4.7 us; 488 us = 146 km = 43 dB). In such a case, the network controller's internal synchronization plan selects the one with the smaller geographical distance to the person to be synchronized as the actual synchronizer, because it most likely corresponds to the practical situation. After synchronizing to the station, this manifests itself as distance information provided via the network.

Once the plan has been resolved, a time (1/16 sec. Resolution) is transmitted to each station, at which point the associated synchronization message arrives. The time is determined by the number of the synchronization sequence, i.e. the level (8/16 sec per sequence). In addition, that delay information already mentioned above is provided for delay compensation.

The following is a procedure in a synchronized network that guarantees the synchronous transmission of paging information regardless of the practical implementation of the communication network.

Once the network or isolated areas have been synchronized, this means in practice that the network is running exactly at the start-up time. The time of the initiator, in turn, is only to the nearest few seconds in the time of the network controller based on the time message transmitted over the network, as described above. This fact must be taken into account when sending the actual remote paging messages to ensure that all stations are sending the same paging at the same time.

The network controller tends to buffer the paging received from the terminal to some extent in order to obtain longer simultaneous transmission periods for the base stations. However, the buffering is such that, in a possible full load situation, it does not slow down the transmission of searches. When the network controller sends a paging message to the base stations, it also makes a plan for when the paging will be transmitted on the radio path. First, it is taken into account with a certain margin of safety that the message has had time to pass to all base stations. At this time, a certain margin of safety is added to compensate for the asynchronousness of the clocks (between the network controller and the base stations). The final transmission time is then obtained by adding this calculated time to the modern time of the network controller 's own clock. The time of transmission is marked in the search message with a resolution of 1/16 second. The network controller then calculates the time at which these searches were sent, when the start time is just marked in that search message.

In planning the time of transmission of subsequent searches, all the calculations described above are performed accordingly, and the result is compared with the time of completion of the current transmission. If it is determined that these new searches are due to be sent before the transmission of the previous searches ends (the data transmission network has a faster transmission capacity), the end of the previous ones is marked as the moment of transmission, in which case the transmission continues continuously. The new end time is also calculated accordingly. If, on the other hand, it is found that new searches do not arrive before the end of the previous transmissions - the transmission may be paused - a new transmission period is started by instructing the network message to send the 576-bit preamble specified by the POCSAG in such a situation before transmitting. The start time of the transmission is then set to the specifically calculated time. The share of the introduction (18/16 seconds) is also taken into account in the transmission end calculation.

The method described above allows the network controller PNU to provide synchronization information while the paging transmission of the base station is still in progress, because the PNU knows the exact situation at the base station. This has the advantage that the time taken to transmit the synchronization commands in the communication network does not reduce the transmission capacity of the retrieval information of the system.

Here are some problems with syncing and keeping it in sync.

The basic problems with synchronization are to perform it accurately enough and maintain synchronism. Based on the above, the analysis of the problems assumes that the communication network itself does not affect the synchronization and focuses on what happens to the synchronization information from the base station controller TSI24 and on the other hand the situation in the synchronized network.

The following specifications are given for the practical radio transmitter and synchronization receiver:

- to maintain synchronization, a clock signal of 6.4 MHz with a total operating accuracy of +/- 0.3 ppm is available from the radio section to the base station controller. The clock signal is divided by four at the base station controller, giving a basic clock frequency of 1.6 MHz. This in turn means that the basic synchronization resolution with the base station controller is 625 ns. The 512 baud transmission clock is formed from a 1.6 MHz clock by a programmable divider. The basic distribution constant is then 3125.

- the data signal from the base station controller TSI is delayed in the radio transmitter section measured from the antenna with a certain standard time +/- 40 μs inaccuracy.

- the data signal coming to the base station controller is synchronously delayed from the signal received by the antenna at the receiver, measured with a certain standard time +/- 40 ps inaccuracy.

The problem will be solved in the following light of a practical calculation example, which is illustrated with reference to Figure 5.

Denote: synchronizing transmitter: TCa synchronizable transmitter: TCb synchronizable transmitter: TCc transmitter (radio part) nominal delay: ts synkr is ivas taman11ime nominal delay: tr

Correspondingly, indicate:

TCa transmitter delay deviation: tap (+/- 40 ps max)

TCb: n: tbp (- -) ·

TCc's "": tcp (- -)

TCa receiver delay deviation: tag (+/- 40 ps max)

TCb: n: tbg (- -)

TCC: n tcg {- -)

714 5 2

The nominal distance between the stations is tn ~ 100 ps = 30 km. Since the actual distances vary between 20 km and 40 km, the travel time variation due to the ambiguity of the synchronization distance is obtained as tbd / tcd = +/- 33 ps max, if the travel time compensation were made on the basis of this nominal distance and not, as described above, on the basis of the actual distance.

Synchronization delays (TCa synchronizes TCb and TCb)

<td colspan="5">TCc: n):</td>

<td>TCb: Ile</td><td>come on</td><td>data:</td><td>(ts +1 ap) + (tn- ^ tbd)</td><td>+ (tr + 1 bq)</td>

<td>TCc: lie</td><td>come on</td><td>data:</td><td>(ts + tap) (tn + tcd)</td><td>+ (tr + tcq)</td>

<td>If.</td><td colspan="3">synchronization would be performed</td><td>only</td>

Corrections according to additional times are obtained as synchronization errors i:

TCb: tap + tbd + tbq

TCc: tap + tcd + tcq

Delays in transmission immediately after synchronization:

TCa: (ts + tap)

TCb: (tap + tbd + tbq) + (ts + tbp) TCc: (tap + tcd + tcq) + (ts + tcp)

If the receiver were in the common coverage area of ​​TCa and TCb, there would be a phase difference:

ts + tap - tap - tbd - tbq - ts - tbp = - (tbp + tbq) - tbd - 113 ps max.

If, on the other hand, the receiver were in the common coverage area of ​​TCb and TCc, there would be a phase difference: tap + tbd + tbq + ts + tbp - tap - tcd - tcq - ts - tcp = (tbp + tbq) - (tcp + tcq) + tbd - tcd = 226 ps max

From the above calculation examples, it can be seen that the error after synchronization is relatively large, so that due to the clock operation error, the need for synchronization is correspondingly more immediate. This in turn lowers the search capacity of the network.

The above synchronization problems are well isolated and can be solved as follows:

a) The characteristics of the radio transmitter / receiver station must be such that the receiver can listen to its own transmission. In this way, the transmitter matching unit can measure the total delay deviation of the transmitter and the receiver during synchronization. With this measurement result, the phase of the data going to the radio transmitter in normal operation is corrected in advance so that the signal is obtained in the correct phase for the paging receiver. It should be noted that the solution is irrelevant in terms of equipment cost.

(b) In connection with the system configuration, the contact and distance information already presented shall be provided as part of the presentation of each TSI. With this procedure, the error factors tbd and tcd can be substantially reduced from those shown in the calculation example.

As a result of the combined effect of the above corrections, the immediate phase error after synchronization drops to at least +/- 20 us in its worst case, leaving +/- 219 us for the clock running error. This means the synchronization interval converted to that clock stability specification is about 12 min.

This can be further illustrated by a practical calculation example:

For example, the synchronization path is selected: TCa synchronizes TCb and TCb synchronizes TCc, which makes the situation even more difficult than in the previous example.

Delays caused by TCa synchronization:

(ts + tap) + (tn + tbd) + (tr + tbg)

The TCb makes a correction as described above in connection with performing the synchronization (transmission step 4.);

- (tn + tbd) - (tr + tbq) - (ts + tbp)

In this case, the total error in the data line associated with the radio part from the base station controller is:

(ts + tap) - (ts + tbp)

Similarly, when TCb synchronizes TCc, the delays are:

(ts + tap) - (ts + tbp) + (ts + tbp) + (tn + tdd) + (tr + tcq) = (ts + tap) + (tn + tdd) + (tr + tcq)

A correction to this results in:

(ts + tap) - (ts + tcp)

Immediately after synchronization, there would be delays in leaving the antennas thus:

TCa: (ts + tap)

TCb: (ts + tap) - (ts + tbp) + (ts + tbp) = (ts + tap)

TCc: (ts + tap) - (ts + tcp) + (ts + tcp) = (ts + tap)

Thus, the transmission delay of the initiator remains in the system as if at the bottom, but it does not matter to the receiver, because it is the same for all stations. On the other hand, it can be argued that the transmitter and receiver delays of base stations do not need to be measurable separately.

In this connection, reference is made to the example of Fig. 6, in which it is assumed that the phase difference of the signal transmitted and received by the own transmission is measured to be 517 ps. From the given input data, the transit time delay has been calculated to be 49 ps. The phase of the transmitter T is advanced by a total delay of 566 ps before the signal is fed to the point T.