METHOD OF COMPACTION OF ANY MESSAGE RELATIVE
TO EXTREMELYWIDE SUBJECTS TO A FIXED SIZE AND
COMMUNICATION CHANNEL FOR COMPACTED MESSAGE TRANSFER
This invention relates to communication (telephony, television, telemetry, Internet) where the following things occur: a frequent transfer of large information contents to a subscriber or a frequent exchange of large messages through wire or wireless communication lines under conditions of limited transmission (receiving) or low capacity of a transmission line, or high ether noisiness, or the demand to exclude unauthorized access to the data transferred as well as various combinations of these factors. The invention enables to design telecommunication networks with the least number of lines (channel groups) and to increase a number of subscribers using a communication channel in already existing networks. It is achieved by means of message compaction (size reduction), i.e. reduction of the message transmission time.
The method of message size compaction whereby "data compaction is achieved through message coding with the aid of a hierarchical database as well as the communication channel for transfer of these messages are known (see V.I. Bodiakin "Where are you going, a man?". SINTEG, M., 1998). References (addresses) to combinations of characters (letters, words, fragments, images and the like) already existing in a database, to combinations of such combinations and so on are elements of hierarchical databases. Every the following database level contains references (addresses) to various combinations of the previous lower level combinations as to elements of a dictionary. The coded message as an element of this database is arranged at a higher level as compared with simpler combinations - elements of lower levels of the database - of which it is composed.
The message coded in that way by a sending subscriber is transferred through a communication line to a receiving subscriber where it is decoded by means of the same
database as a hierarchical database of the sending subscriber. While coding a transition to higher levels of a database take place; while decoding - to lower levels so long as the initial decoded form of the message sent will be restored at the lowest level. With the developed hierarchical database, data in the message transferred (coded) are compacted considerably as compared with the decoded form of the message to be transferred.
But on message coding such fragments or their new combinations may occur, in this case they are included in the database of the sending subscriber as new elements at various levels of his database, i.e. the database of the sending subscriber may be augmented.
The sending subscriber's message coded with the aid of his augmented database cannot be restored (decoded) no longer by means of the non-augmented database of the receiving subscriber. Therefore to restore the initial decoded form of the message the receiving subscriber's database needs to be augmented with the new elements of the sending subscriber's database, i.e. the databases of the sending subscriber and receiving subscriber should be made identical.
Message transfer is a sending of the element of a fixed length address from a certain level of the sending subscriber's database.
In case of coding of large combinations of elements by the characters agreed between the sending subscriber and the receiving subscriber, a size compaction of the message transferred also takes place, but it occurs due to limitation of transferable messages, i.e. owing to narrowing of subjects of the message transferred.
The method of a hierarchical database augmentation applied in archiving programs as well as the method of unarchiving of the requested message described in the above book of V.I. Bodiakin are the nearest prototypes.
The shortcoming of this method lies in the fact that the user's database (database of the receiving subscriber) is merged with the sending subscriber's database that eliminates the possibility to transfer the compacted message to the archive user.
The nearest prototype of the applied communication channel is the multilevel system for coding and decoding of a character string given in Appendix 3 of the above book of V.I. Bodiakin. This system at each level of a hierarchical database has a level dictionary not fixed and not limited beforehand. The coding is realized consecutively from a lower to an upper level, and decoding - from an upper to a lower one. The level dictionary is used in operation at each a database level.
In particular while coding at every current level the message for the next level is composed of references (figures, numbers) to combinations (numbers) in the coded message at the current level, which are available in the dictionary of that current level of the hierarchical database. If the required combination is not found, this dictionary of the level is augmented with this new not found combination with assignment of its own number to it. The decoding process is realized in a reversed order from an upper to the next lower level of the hierarchical database through expansion of every (number) of the decoded message with use of a dictionary of the next lower level, at which a certain combination of numbers corresponds every such number. The message for the next lower level is composed of such combinations.
The imperfection of this system for coding and decoding of a character string (i.e. message) is in the fact that the coding and decoding blocks are arranged within the limits of one (computing) device, and their common hierarchical database of message elements is arranged in the same place.
The technical problem of the applied method is augmentation of the sending subscriber's hierarchical database to achieve identity of the sending and receiving subscribers' databases.
The object of this invention is the method of compaction of any message relative to extremely wide subjects to a fixed size and maintenance of identity of the sending and receiving subscribers' databases mutually associated in this way under various conditions
of their interaction as well as the communication channel for transfer of compacted messages.
The method of identity maintenance described in this patent application* may be also applied in case of other sources of database augmentation.
The applied method has no imperfection of the nearest prototype as it duplicates the sending subscriber's database for the receiving subscriber at a "zero" spatial length of the communication channel between them. The duplication is realized by means of augmentation of the initial coincident (for instance, zero) database of the sending subscriber for the purpose to bring it to correspondence with the sending subscriber's database.
If the sending and receiving subscribers in the process of their interaction do not change their roles, then while message coding the sending subscriber's database is augmented.
The applied communication channel enables to solve several technical problems:
1) to find schemes of functional associations of blocks for message preparation and analysis as well as the structure of the transferred message itself; 2) to organize terminal devices at the ends of the communication line which may both receive and transfer messages; 3) to organize two-way transmission; 4) to connect terminal devices of the communication channel by means of real communication lines for messages transfer over one line in one direction and over the other line - in the opposite direction.
For solution of the stated problem the method of compaction of any message relative to extremely wide subjects to a fixed size by at least one sending subscriber with a hierarchical database includes provision of at least one receiving subscriber with the hierarchical database being identical the sending subscriber's database, and formation by this at least one subscriber pair as well as transfer of the message in the form of level number and address of the element in the sending subscriber's database to the receiving subscriber.
In the applied method a transfer of elements, which augment the sending subscriber's database over the communication channel, is realized to the same levels at the receiving subscriber's hierarchical database and to the same addresses which the augmentation elements have in the sending subscriber's database For proper arrangement of every such element in the receiving subscriber's database the sending subscriber should provide the element with data of the level of the database and its address at this level according to which this augmentation element is to be arranged, and the receiving subscnber should have a program for decoding of such data, i e a transfer protocol for this database pair should be available
While such a transfer of all elements for database augmentation to the receiving subscriber not one by one, but as a group, in general, a load time of the communication channel for that transfer reduces Thus it is rational to fix all augmentation elements appeared while coding of any message in the previously provided buffer of the sending subscriber and to transfer all fixed in the buffer elements as a group to the receiving subscriber after the coding completion
In a subscriber pair the transmitter and the receiver may from time to time change their roles If message coding, resulted in database augmentation, is realized by both parties in non-overlapping periods of time, the new elements of this base appeared during coding are transferred to the other hierarchical database of the subscriber pair
The above non-overlapping periods of time are determined by assignment of the sending role to one of the subscribers connected through a communication channel, that is realized in response to a signal initially transferred by one of these subscribers and confirmed by the other subscπber This role is transferred from one subscriber to the other one by means of the similar confirmed signal Thus the non-overlapping periods of time during which augmentation elements are transferred to any database of the subscriber pair over the communication channel with the purpose to maintain identity of the databases are determined completely
All the above does not exclude, of course, the transfer by the receiving subscriber of coded counter-messages not requiring augmentation of his database.
It should be noted that all rules of assignment and reassignment of the-sending and receiving roles are regulated by a transfer protocol for this subscriber pair. Of course, in these cases it is advantageous to transfer augmentation elements not one by one, but as a group fixed in a corresponding buffer. However it is not known before coding of any message whether in this case elements for database augmentation appear or not. Because of this with the purpose to save a database memory it is efficient to dedicate the buffer in a database only after appearance of any first augmentation element.
On the other hand, to simplify operation algorithm of a hierarchical database it is rational beforehand to provide for the buffer to fix augmentation elements which may appear while message coding.
If coding and message transfer from the both ends of the communication channel are made independently, i.e. generally speaking in overlapping periods of time, then augmentation elements appearing while coding cannot be included in his own database at once, as it was in the above described cases, since it may result in non-identity of databases of subscribers. To prevent this the augmentation elements appearing at each subscriber are firstly fixed in the buffer dedicated in his database and only then, upon determination of buffers sequence, the groups of elements fixed in buffers are transferred, for example, in this sequence at first to the database of the first subscriber and then to the database of the second subscriber or, for example, from one buffer to the first subscriber's database and to the second subscriber's database, and then from the other buffer to the first subscriber's database and to the second subscriber's database. The sequence of buffers and databases is specified in a transfer protocol which also regulates moments of the beginning and completion of this transfer. The moment of transfer beginning may, for example, be determined by the later moment of completion of own message coding by one of the two subscribers or by the fact of complete fill of a buffer.
To simplify the transfer protocol it is expedient to provide for availability of own buffer at each subscriber in a subscriber pair.
If any subscriber connected with several other different subscribers, then for communication with each of them an independent hierarchical augmentable database should be provided as though it represented the database of the second subscriber in a subscriber pair. Collection of sur database forms a subscriber station.
Some databases at such a subscriber station may contain the same elements. As a result total size of databases of a subscriber station may increase.
To reduce total size of subscriber station databases a common part - hierarchical non-augmentable database, is extracted from them. In this case the subscriber station database is composed of this non-augmentable database and a number of augmentable databases according to a number of connections of this subscriber station with other subscribers of a communication network. For saving of elements addressing in interconnected database pairs the same non-augmentable database is included in a hierarchical database of each subscriber or each subscriber station connected with this subscriber station. And since there is a potential possibility of connection of each subscriber with many other subscribers, the non-augmentable database being the same for all subscribers is additionally included in each subscriber station or a database of each individual subscriber not belonging to any subscriber station of a communication network of various subscribers.
The applied communication channel is intended to deliver ensemble N of various messages characterized by corresponding information contents I (byte, bit) to a receiving subscriber. The channel has a feature to deliver every such message by means of its transfer in a compacted (coded) form of less volume I being less than I, that is
therewith as number N of messages delivered to a subscriber increases, volume i< I decreases relatively, i.e. ratio i/l decreases and at the limit (N=>∞) achieves small quantities ε« 1, i.e.
lim| - \ = ε
N -» ∞
It is obtained through including of references c to previous messages (and to their combinations), which were previously transferred to a receiving subscriber and fixed by him in his hierarchical database (dictionary) and which again were included in a regular message I to be delivered to a subscriber.
In addition i<l includes new, previously not encountered elements represented in I and intended to continue augmentation of the receiving subscriber's hierarchical database. Besides that i<l includes previously not encountered combinations of elements of the subscriber's augmented database, which in their turn also augment his databases.
The detection of these new elements and new combinations occurs while coding of message I to be delivered to a subscriber. As a result of this both augmentation of the database of the coding block of a transmitter and formation of the structure of message i<l to be transferred to the subscriber. After reception of this message by a receiver a part of the message is used for corresponding augmentation of the hierarchical database of the receiving subscriber (receiver), and the other part is decoded. In doing so identity of databases of the transmitter coding block and the receiver decoding block is achieved, i.e. "adjustment" of the receiving subscriber's hierarchical database according to the sending subscriber's (transmitter) hierarchical database.
It is known that one or another coding or decoding of information transferred occurs in a communication channel. In this case there are channels in which transformation of message elements, i.e. change of a coding system and corresponding
adjustment of a decoding system, take place. The shortcoming of these channels is i<l, moreover often i/l«1.
In case of coding of large message elements (phrases) combinations by means of the characters agreed with a subscriber i/l<1 is achieved. But it occurs by means of limitation of a number of transferable messages, i.e. due to narrowing of subjects of the messages transferred.
The applied communication channel avoids the shortcomings described for the nearest prior art owing to use of the known wire or radio circuits for message (information) transfer at the required distance, through which after coding of the message to be transferred by the transmitter coding block with simultaneous augmentation of its hierarchical database the message preparation block generates the message of the structure agreed with the receiver and transfers it over a communication line to the receiver, where the message analysis block extracts the part of the message intended for augmentation of the hierarchical database levels of the receiver decoding block, and the other part - the coded message itself, decoded by the decoding block with use of level dictionaries augmented in this way of his database. As a result adjustment of hierarchical databases of the sending and receiving subscribers (transmitter and receiver), achievement of their identity as well as coding, transfer and decoding of the message take place.
While coding database levels are augmented in sequence beginning from the lower level, and while decoding augmented database levels are used beginning from the upper level. Hence the first part of the prepared message consists of newly appeared on coding augmentations to hierarchical databases beginning from the upper level (i.e. in the order convenient for operation of the decoding block), and the second part - of the decoded message.
It is expedient to provide for capability to transfer messages in both directions for the unidirectional communication channel formed in this way. To organize terminal
devices, which may receive and transfer messages, at the both ends of the communication line the transmitter should include the decoding block and the message analysis block connected by means of an augmentation circuit with a database and connected with each other, as well as the control unit with two gates controlled by it and installed: the first one in the augmentation circuit of the database from the message analysis block and the second one in the augmentation circuit of the database from the coding block. Besides that the control unit is connected both with the message analysis block and the message preparation block. Altogether they form one terminal device.
Correspondingly to organize the second terminal device the receiver includes the message preparation block and the coding block connected by means of an augmentation circuit with a database and connected with each other, as well as the control unit with two gates controlled by it and installed: the first one in the augmentation circuit of the database from the decoding block and the second one in the augmentation circuit of the database from the message analysis block. Besides that the control unit, as for the device considered above, is connected both with the message analysis block and the message preparation block.
As the both terminal devices look absolutely symmetric and similar, it is favorable to break such symmetry in order to prevent uncertain situation during operation of such a communication channel by means of assigning of one terminal device as a master device and the second one as a slave device. In this case the master device while transfer to the slave device of a signal with a directive of operation of the slave device in the only transfer or in the only receive mode confirmed by the slave device prevents cases of transfer of counter messages with database augmentation that may lead to non-identity of the master and slave device database. At the same time such regulation of the slave device operation does not limit the possibility to transfer in the only receive mode as well as the possibility to receive messages in the only transfer mode on coding and analysis of which
augmentation of its database does not take place. The same relates to the operation modes of the master terminal device.
For reduction of the probability of unauthorized access to a communication channel it is expedient to change the structure of the message transferred and coding parameters (number of elements in a combination) of the message to be transferred including the data of these changes in the transferred message itself. Thanks to it operation of the message preparation and message analysis blocks and the coding and decoding blocks is matched and thereby correct and hidden functioning of the whole communication channel is obtained.
On providing for simultaneous transfer of messages in both directions of the channel the amount of information transferred per a time unit increases. In case of organization of bi-directional transfer for the option of wire communication line it is sometimes rational to avoid duplication of a real communication line between the transmitter and receiver.
This problem is solved by means of separation of carrier frequencies used to transfer messages in direct and reverse directions. For that the transceiver of the transmitter contains the digital-analog converter for the first dedicated carrier frequency connected with the message preparation block and the digital-analog converter for the second dedicated carrier frequency connected with the message analysis block, and the transceiver of the receiver includes the digital-analog converter for the second dedicated carrier frequency connected with the message preparation block and the digital-analog converter for the first dedicated carrier frequency connected with the message analysis block.
The essence of the applied inventions is explained by the drawings. Fig. 1 shows the block diagram of a communication channel with one-way message transmission. Fig. 2 shows the block diagram of a terminal device of a communication channel with two-way message transmission.
The block diagram of the communication channel given on Fig. 1 contains a multilevel coding block 1 with the sending subscriber's hierarchical database 2, message preparation block 3 and transceiver 4 forming altogether the transmitter, as well as a transceiver 5, message analysis block 6, multilevel decoding block 7 with the receiving subscriber's hierarchical database 8 forming altogether the receiver, and also a wireless or wire communication line 9.
The device shown on Fig. 1 operates in the following way. On delivery of message A to be transferred to the coding block 1 , its coding (compaction) occurs and as a consequence of it the message coded in the form C is transferred to the message preparation block 3. While coding of message A codes Pnv which augment levels of the database 2 from the lower level to the upper level and at the same time they are delivered to the block 3 and as a result of this message B is prepared in which augmenters represented in a transposed way Pvn (i.e. from the upper level to the lower level). Message B through the transceiver 4, line 9 and device 5 comes to the message analysis block 6 where it is divided into Pvn and C parts. Pvn part is delivered to the block 8 for augmentation of the database levels beginning from the upper level, and C part - for decoding in the block 7 used augmented from Pvn part levels of the block 8. As a result at the output of the block 7 the transferred message A is prepared.
The block diagram of a terminal device of a communication channel with two-way message transmission shown on Fig. 2 includes the multilevel coding block 10, message preparation block 11 , hierarchical database 12, message analysis block 13, multilevel decoding block 17, control unit 15, gates 16 and 17 and transceiver 18.
The device shown on Fig. 2 operates in the following way. The device operates in two modes:
1st mode: gate 16 is opened and gate 17 is closed - transmission with database augmentation while coding in the block 10; 2nd mode: gate 16 is closed and gate 17 is opened
- receiving with database augmentation while analysis of the message received in the block 13.
In this case in the communication channel the following situation takes place: if one terminal device operates in the 1st mode, the other device - necessarily in the 2nd, in doing so the modes change each other. It is realized by the control unit 15 of the master device which through the block 11 sends signal Kd of a change of operation mode of terminal devices. The sent signal Kd comes to the block 13 of the slave device from which it comes to its control unit 15, which change positions of its gates 16 and 17 and through the block 11 of this device sends a confirmation of the received signal Kd to the master terminal device. This confirmed signal Kd received by the master device comes from the block 13 to the control unit 15 of the master device and as a consequence of it the unit 15 also changes positions of its gates 16 and 17.
The 1st mode. The received by the coding block 10 message A to be transferred is coded and augments, if necessary, the database 12 through the open gate 16. In this case in the block 11 the message is prepared and transferred to the other terminal device through the transceiver 18. Unlike the diagram shown on Fig. 1 , the terminal device given on Fig. 2 may at the same time not only transfer messages, but also receive them to the message analysis block 13. In this case if the message received does not contain a part for augmentation of the database 12, it may be decoded and given to a subscriber from the output of the block 14 as an outgoing message D. Otherwise taking into account that the gate 17 is closed, the message, as a rule, cannot be decoded.
The 2nd mode. The message received by the message analysis block 13 from the other terminal device augments, if necessary, the database 12 through the open gate 17 and is decoded by the block 14 and is given as an outgoing message in the D form. As during operation in the 1st mode, unlike the diagram shown on Fig. 1 , the terminal device given on Fig. 2 may at the same time not only receive messages, but also transfer them. Any message A to be transferred is decoded by the block 10, but without augmentation of
the database 12, as the gate 16 is closed. Then the message coded in this way and prepared by the block 11 is transferred through the transceiver 18 to the other terminal device of the communication channel, where, in case if there is no need in augmentation of the database 12 while its coding, it is necessarily decoded and is given to a subscriber as it was noted in description of the 1st mode. Otherwise it, as a rule, cannot be decoded.
Thus both the terminal devices of the communication channel have main, declared modes of operation (message transfer in the 1st mode and message receiving in the 2nd mode) and "auxiliary" ones (receiving in the 1st mode and transfer in the 2nd mode). The expression "as a rule" used on description of the "auxiliary" modes takes into account those possible cases when augmentation of the database 12 made on operation in main modes is sufficient for message decoding.