JP2018046392A - Communication device and communication program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a communication device and a communication program capable of losslessly compressing the amount of information to be transmitted and communicating with a smaller amount of information. A communication device of the present invention predicts one or more of the next transmitted or received information using information transmitted or received in the past, based on the characteristics of the communication procedure of each layer of the communication protocol. Each layer included in the transmission packet by referring to the prediction means that predicts the value for each layer and associates a code with each predicted value to create the predicted value correspondence information of each layer, and the predicted value correspondence information of each layer. It is provided with a coding means for replacing the information of the above with a corresponding code and a decoding means for replacing the code included in each layer of the received packet with the corresponding predicted value by referring to the predicted value correspondence information of each layer. It is a feature. [Selection diagram] Fig. 1

Description

  The present invention relates to a communication device and a communication program, and can be applied to a communication device and a communication program that perform reversible compression on the amount of information to be transmitted and communicate with a smaller amount of information.

  Patent Document 1 describes a signal compression technique. This technique is an improvement of entropy coding using Huffman codes. The basic concept is that the number of elements to be transmitted is a specific number N of elements, and arbitrary information (symbols) is given a short code in order from the data sequence with the highest appearance frequency. That is, the transmitting side sends a short code instead of the symbol to be transmitted, and the receiving side decodes the symbol for the short code to obtain received information.

  The purpose of the technology described in Patent Document 1 is to provide an effective encoding conversion table for a data series that does not have a statistically sufficient number of samples and has low appearance frequency information, or for a data series that does not follow a mathematical model. It is to prepare and perform compressed communication. Specifically, the transmission side and the reception side have a plurality of common encoding conversion tables, and the transmission side determines which encoding conversion table to use for the transmission data has the maximum compression effect. . Then, the transmission side transmits transmission information including the identification information of the encoding conversion table to be used and the compressed short code to the reception side.

  As another compression technique, there is a predictive coding technique. This is used in moving picture transmission technology. For example, when the transmitting side transmits a moving image, the frame in the moving image and the next frame are often approximated. Of course, the frame before and after may be greatly different, such as a scene change, but most of the time of the moving image is similar to the previous frame. Therefore, the transmission side refers to the previous frame and transmits difference information from the current frame, information according to the motion of the moving image, and the like to the reception side.

JP 2011-514438 A

  However, the compression technique described above has the following problems.

  The technique described in Patent Document 1 needs to include a plurality of encoding conversion tables, but each encoding conversion table may not always include an optimal code. In other words, by using each encoding conversion table, a relatively efficient code can be obtained, but it is not always an optimal code.

  Further, the predictive encoding technique described above has a problem of how to reduce the difference information from the previous frame, and it is desired to encode with a smaller amount of information.

  Therefore, there is a need for a communication device and a communication program that can reversibly compress the amount of information to be transmitted and communicate with a smaller amount of information.

  In order to solve such a problem, the communication apparatus according to the first aspect of the present invention (1) uses the information transmitted or received in the past based on the characteristics of the communication procedure of each layer of the communication protocol, and then transmits the information. Or prediction means for predicting one or a plurality of predicted values of received information for each layer, creating a predicted value correspondence information for each layer by associating a code with each predicted value, and (2) a predicted value for each layer Referring to the correspondence information, encoding means for replacing the information of each layer included in the transmission packet with the corresponding code, and (3) the prediction value correspondence information of each layer is referred to and included in each layer of the received packet Decoding means for replacing the generated code with the corresponding predicted value.

  The communication program according to the second aspect of the present invention is the following: (1) Information transmitted or received next using information transmitted or received in the past based on the characteristics of the communication procedure of each layer of the communication protocol. Predicting means for predicting one or a plurality of predicted values for each layer, creating a predicted value correspondence information for each layer by associating a code with each predicted value, and (2) referring to predicted value correspondence information for each layer Encoding means for replacing the information of each layer included in the transmission packet with the corresponding code, and (3) the code included in each layer of the received packet with reference to the predicted value correspondence information of each layer It is characterized by functioning as a decoding means that replaces the predicted value.

  According to the present invention, it is possible to communicate with a smaller amount of information by reversibly compressing the amount of information to be transmitted.

It is an internal block diagram which shows the internal structure of the communication apparatus which concerns on embodiment. It is a whole lineblock diagram showing the whole communication system composition concerning an embodiment. It is a flowchart which shows the communication process of the transmission side which concerns on embodiment. It is a flowchart which shows the communication process of the receiving side which concerns on embodiment. It is a flowchart which shows the operation processing in the communication apparatus which concerns on embodiment. It is a flowchart which shows the operation processing in the communication apparatus which concerns on embodiment. It is a sequence diagram which shows a TCP process when the receiving side receives all the sequence numbers normally. It is a sequence diagram which shows TCP processing when a packet is lost. It is a sequence diagram which shows TCP processing when a packet produces fluctuation.

(A) Basic Concept of the Present Invention The basic concept of the present invention will be described before describing embodiments of a communication device, a communication program, and a communication system according to the present invention.

  The present invention is a compression encoding method (decoding method) that reduces the amount of information to be transmitted / received using the characteristics of information to be transmitted / received according to a communication protocol.

  By sharing the predicted information between the transmission side and the reception side, even when the contents of the information are different, it is possible to exchange a code having the same code length with a short code length. In other words, even when the transmitting side and the receiving side exchange a short code having the same value, the information content obtained from the code is different, but the transmitting side and the receiving side can recognize the common information content.

  (A-1) For example, when communication is performed between the transmission side and the reception side, the transmission side transmits a packet with a destination address and a transmission source address to the reception side. When the reception side responds to the transmission side, the reception side changes the order of the destination address and the transmission source address included in the header of the received packet, and returns a response packet to the transmission side.

  That is, when a communication session is started, header information including two addresses as one set is used between the transmission side and the reception side. Then, between the start of the session and the end of the session, the transmission side and the reception side can predict that the frequency of sending and receiving packets using one set of address information will increase.

  Therefore, when a communication session is started, the set of addresses to be transmitted / received is replaced with a short code to communicate, and when the session ends, the code is released to obtain information on the packet header information. Can be reduced. Note that by releasing the short code at the end of the session, it is possible to assign a short code to a set of transmission / reception addresses of another new session.

  (A-2) The transmission side and the reception side recognize whether or not information can be correctly delivered to the reception side according to the characteristics of the communication protocol.

  For example, in the case of TCP, the transmission side transmits a packet including header information describing a sequence number to the reception side. The receiving side that has received the packet sets the value obtained by adding the MTU value (maximum data size value) to the sequence number that can be received as the ACK number (response confirmation number), and the ACK packet ( A response confirmation signal) is sent back to the transmission side.

  In the case of TCP, the ACK number of the ACK packet is different depending on whether the receiving side has received the correct sequence number or the packet has been lost and the sequence number has not been correctly received. If the MTU value is shared, an ACK number that can be received can be predicted.

  As described above, from the characteristics of the communication protocol, each of the transmission side and the reception side can commonly predict the information content exchanged according to the communication status.

  In addition, the predicted frequency of each predicted value varies depending on the communication status. Therefore, the amount of information can be reduced between the transmitting side and the receiving side by replacing a predicted value considered to be frequently generated with a code having a short code length.

  The above is an example, and there are many other information whose communication contents are easily predicted. The amount of communication can be reduced by encoding information predicted according to the characteristics of the communication protocol between the transmission side and the reception side.

(B) Main Embodiments Hereinafter, embodiments of a communication device, a communication program, and a communication system according to the present invention will be described in detail with reference to the drawings.

(B-1) Configuration of Embodiment FIG. 2 is an overall configuration diagram illustrating an overall configuration of a communication system according to the embodiment.

  In FIG. 2, the communication system 1 according to the embodiment includes a plurality of communication devices 2-1 and 2-2 that can be connected to a network 3.

  The network 3 is a communication network that communicates information using a predetermined communication protocol, and may be a wired line or a wireless line. In this embodiment, the case where the communication protocol of the network 3 is TCP / IP is illustrated.

  The communication devices 2-1 and 2-2 transmit and receive information to be transmitted via the network 3 according to TCP / IP. Each of the communication apparatuses 2-1 and 2-2 includes a transmission processing unit 10 that forms and transmits a packet including data to be transmitted, and a reception processing unit 20 that analyzes the received packet.

  FIG. 1 is an internal configuration diagram illustrating an internal configuration of a communication device 2 according to the embodiment.

  Note that the communication devices 2-1 and 2-2 have the same internal configuration, and are therefore referred to as the communication device 2 in FIG.

  In FIG. 1, the communication device 2 roughly includes a transmission processing unit 10 and a reception processing unit 20. The communication device 2 includes a transmission side input interface unit 31, a transmission side output interface unit 32, a reception side input interface unit 33, and a reception side output interface unit 34.

  The communication device 2 is connected to the terminal 4 and exchanges packets with the terminal 4. That is, the communication device 2 acquires data from the terminal 4 through the transmission side input interface unit 31 and outputs data to the terminal 4 through the reception side output interface unit 34.

  The communication device 2 is connected to the network 3, and exchanges packets including data with the network 3. That is, the communication device 2 receives the packet via the receiving side input interface unit 33 and transmits the packet including data to the network 3 via the transmitting side output interface unit 32.

  The transmission processing unit 10 encodes the data to be transmitted and the address data described in the header portion, adds a packet header including the encoded address data to the payload including the encoded data, forms a transmission packet, and transmits the packet To do.

  Although not shown, the hardware configuration of the transmission processing unit 10 includes, for example, a CPU, a ROM, a RAM, an EEPROM, an input / output interface unit, and the like. And each function of the transmission process part 10 is implement | achieved when CPU runs the processing program (communication program) stored in ROM. The processing program (communication program) may be installed in the communication device 2 so that each function can be constructed. In this case, the processing program functions as a block of the transmission processing unit 10 in FIG.

  As shown in FIG. 1, the transmission processing unit 10 includes a transmission packet header analysis / encoding unit 11, a protocol-specific encoding unit 12 (12-1 to 12-3), a payload encoding unit 13, and a transmission packet combining unit 14. Have

  The payload encoding unit 13 compresses and encodes data from the terminal 4 and supplies the encoded data to the transmission packet combining unit 14 as a payload. The payload encoding unit 13 compresses and encodes data, thereby reducing the data amount (information amount) of the payload portion. Note that various existing techniques can be applied to the data compression encoding process by the payload encoding unit 13, and thus detailed description thereof will be omitted.

  The transmission packet header analysis / encoding unit 11 encodes the header data described in the header part of the transmission packet and supplies the encoded header data to the transmission packet synthesis unit 14.

  Here, the transmission packet header analysis / encoding unit 11 is connected to the protocol-specific encoding units 12-1 to 12-3, and decomposes the header portion of the transmission packet into address data for each protocol. Is supplied to the corresponding protocol-specific encoding units 12-1 to 12-3. Further, the transmission packet header analysis / encoding unit 11 acquires data encoded for each protocol by the encoding units 12-1 to 12-3 for each protocol, combines these encoded data, and generates a header. Form data.

  The protocol-specific encoding units 12-1 to 12-3 encode address data for each protocol. Here, for each protocol, for example, each layer in the OSI reference model can be used. The address data for each protocol may be an Ether (registered trademark) header of a layer 2 protocol, an IP header of a layer 3 protocol, a TCP header of a layer 4 protocol, a UDP header, or the like. That is, the protocol-specific encoding units 12-1 to 12-3 are provided for each protocol, such as an Ether header, an IP header, a TCP header, or a UDP header, and encode address data for each protocol.

  Each protocol encoding unit 12-1 to 12-3 encodes address data of each protocol based on a preset method based on the characteristic information of each protocol. In addition, each protocol encoding unit 12-1 to 12-3 includes a next header prediction unit 51.

  The next header prediction unit 51 predicts predicted values of the next header data based on protocol characteristics, and creates conversion information (predicted value correspondence information) of the next header data by associating codes with these predicted values. It is. A detailed description of the processing of the next header prediction unit 51 will be described later.

  The transmission packet combining unit 14 adds the header data encoded from the transmission packet header analysis / encoding unit 11 to the data encoded from the payload encoding unit 13 and synthesizes and transmits the transmission packet. is there.

  The reception processing unit 20 decodes payload data and header data of a packet received from the network 3 and gives the decoded packet data to the terminal 4.

  The hardware configuration of the reception processing unit 20 includes, for example, a CPU, a ROM, a RAM, an EEPROM, an input / output interface unit, and the like, although not shown. Each function of the reception processing unit 20 is realized by the CPU executing a processing program (communication program) stored in the ROM. The processing program (communication program) may be installed in the communication device 2 so that each function can be constructed. In this case, the processing program functions as a block of the reception processing unit 20 in FIG.

  As illustrated in FIG. 1, the reception processing unit 20 includes a reception packet header analysis / decoding unit 21, a protocol-specific decoding unit 22 (22-1 to 22-3), a payload decoding unit 23, and a received packet combining unit 24.

  The payload decoding unit 23 decodes the data of the payload portion of the received packet and gives the decoded data to the received packet combining unit 24. As a decoding method of the payload decoding unit 23, a method corresponding to the encoding method of the payload encoding unit 13 can be applied, and various existing techniques can be applied, and thus detailed description of the processing is omitted here. .

  The received packet header analyzing / decoding unit 21 decodes the data of the header part of the received packet and supplies the decoded header data to the received packet combining unit 24.

  Here, the received packet header analysis / decoding unit 21 is connected to the protocol-specific decoding units 22-1 to 22-3, disassembles the header portion of the received packet for each protocol, and handles the data of each protocol. This is given to the protocol-specific decoding units 21-1 to 21-3. The received packet header analysis / decoding unit 21 acquires data decoded by each protocol by each protocol decoding unit 22-1 to 22-3, synthesizes these decoded data, and converts the original header data into the original header data. Form.

  The protocol decoding units 22-1 to 22-3 decode the data for each protocol to obtain original address data for each protocol. By protocol, as in the case of the encoding units 12-1 to 12-3 by protocol of the transmission processing unit 10, for example, for each layer in the OSI reference model. The protocol-specific decoding units 22-1 to 22-3 are provided for each protocol such as an Ether header, an IP header, a TCP header, or a UDP header, for example, and decode the address header for each protocol.

  Each of the protocol decoding units 22-1 to 22-3 decodes the original address data of each protocol based on the characteristic information of each protocol and based on a preset method. In addition, each of the protocol decoding units 22-1 to 22-3 includes a next header prediction unit 52.

  The next header prediction unit 52 predicts predicted values of the next header data based on the characteristics of the protocol, and creates conversion information of the next header data by associating codes with these predicted values. A detailed description of the processing of the next header prediction unit 52 will be described later.

  The received packet synthesizing unit 24 adds the header data decoded from the received packet header analysis / decoding unit 21 to the data of the payload portion decoded from the payload decoding unit 23, forms the original packet, and sends it to the terminal 4. To give.

(B-2) Operation of Embodiment Next, the operation of communication processing by the communication apparatus according to the embodiment will be described in detail with reference to the drawings.

  Here, the transmission packet header analysis / encoding unit 11, the protocol-specific encoding units 12-1 to 12-3, the next header prediction unit 51, the received packet header analysis / decoding unit 21, and the protocol-specific decoding unit 22-1 -22-3, the operation of the feature of the next header prediction unit 52 will be mainly described.

(B-2-1) Overall Operation In the following, first, the communication device 2-1 in FIG. 2 is the transmission side, and the communication device 2-2 is the reception side. A case where information is transmitted between the two will be described with reference to FIGS.

  In the communication device 2-1 on the transmission side, a transmission packet from the terminal 4 is given to the transmission processing unit 10. In the transmission processing unit 10, the transmission packet is given to the transmission packet header analysis / encoding unit 11 and the payload encoding unit 13.

  In the payload encoding unit 13, data included in the payload of the transmission packet is encoded, and the encoded data is given to the transmission packet combining unit 14.

  In the transmission packet header analysis / encoding unit 11, the header part of the transmission packet is decomposed for each protocol of the communication protocol, and the header data for each protocol is converted into the protocol-specific encoding units 12-1 to 12-3 for the corresponding protocol. Given to.

  Each protocol encoding unit 12-1 to 12-3 stores protocol characteristic information. Based on each protocol characteristic information, the header data is encoded by a method set in advance, and an encoded header is obtained. Data is given to the transmission packet combining unit 14. A detailed description of the header data encoding process for each protocol will be given later.

  The transmission packet synthesis unit 14 adds the encoded header data encoded by the transmission packet header analysis / encoding unit 11 to the encoded data encoded by the payload encoding unit 13 and transmits the encoded data to the network 3.

  In the communication device 2-2 on the receiving side, a packet from the network 3 is given to the reception processing unit 20. In the reception processing unit 20, the received packet is given to the received packet header analysis / decoding unit 21 and the payload decoding unit 23.

  In the payload decoding unit 23, data included in the payload of the received packet is decoded, and the original payload data is given to the received packet combining unit 24.

  In the received packet header analyzing / decoding unit 21, the header part of the received packet is decomposed for each protocol of the communication protocol, and the header data for each protocol is given to the corresponding protocol decoding units 22-1 to 22-3. It is done.

  Each protocol decoding unit 22-1 to 22-3 stores protocol characteristic information. Based on each protocol characteristic information, the header data is decoded by a preset method, and the decoded header data is stored. Is given to the received packet combining unit 24. A detailed description of the header data decoding process for each protocol will be given later.

  In the received packet synthesizing unit 24, the header data decoded by the received packet header analyzing / decoding unit 21 is added to the payload data decoded by the payload decoding unit 23, returned to the original packet, and given to the terminal 4.

  As described above, the communication amount can be reduced by replacing the frequently used data for the header data with a short code for communication.

(B-2-2) Address Data Encoding / Decoding Process by Protocol First, a process of compressing and encoding a set of a destination address and a transmission source address in the header data will be exemplified.

  The address information compression / encoding process (decoding process) is performed by the protocol-specific encoding units 12-1 to 12-3 and the protocol-specific decoding units 22-1 to 22 of the layer 2 protocol, the layer 3 protocol, and the layer 4 protocol. -3. That is, it can be used with the transmission / reception MAC address of the layer 2 header, the transmission / reception IP address of the layer 3 header, and the TCP / UDP transmission / reception port number of the layer 4 header.

  Here, for ease of explanation, processing in the layer-by-protocol encoding unit 12-1 and the protocol-decoding unit 22-1 will be exemplified.

  FIG. 3 is a flowchart illustrating communication processing on the transmission side according to the embodiment, and FIG. 4 is a flowchart illustrating communication processing on the reception side according to the embodiment.

  First, the protocol-side encoding unit 12-1 on the transmission side monitors the session status (S101), and determines the address information included in the Ether header. When the session between the communication device 2-1 and the communication device 2-2 is started and the combination of the destination address and the transmission source address is the first time, the protocol-specific encoding unit 12-1 displays the destination address and the transmission source address. Are stored in the storage unit 53-1, and control information regarding address encoding is set (S102).

  For example, assume that the destination address is “AAAA” and the source address is “BBBB” at the start of the session. As shown in FIG. 3, the protocol-specific encoding unit 12-1 uses the destination address “AAAA” as the first address, the source address “BBBB” as the second address, and encodes a set of “AAAA” and “BBBB”. Set “01”. Further, the protocol-specific encoding unit 12-1 reverses the address combination and sets the first address “BBBB” and the second address “AAA” as the code “02”.

  Usually, the destination address and the source address are inserted into the header for each protocol. However, as described above, by setting the control information related to address encoding, the pair of the destination address and the source address is replaced with the codes “01” and “02” which are 1-bit or several-bit data values. Therefore, the communication amount can be reduced.

  The address (destination address or source address) included in the header part is known at the start of communication and is dynamically changed. Therefore, as shown in FIG. 3, the data to be encoded (in this case, the first address and the second address) is called a variable condition section.

  As described above, the variable condition part is an encoding target in header data for each protocol, and the encoding target is dynamically changed according to a communication state. The variable condition part is set in association with the code at the start of the session, and is deleted after the session is ended in order to release the code.

  The encoding unit 12-1 on the transmission side uses a pair of a destination address and a destination address (that is, a pair of “AAAAA” and “BBBB” is a code “01”, and a pair of “AAAAA” and “BBBB”. Is included in the header and is given to the transmission packet header analysis / encoding unit 11 (S103).

  Next, the protocol-specific decoding unit 22-1 on the receiving side monitors the session status (S201), and determines the address information included in the Ether header.

  When the session is started, the combination of the destination address and the source address is the first time, and the conversion information related to the address encoding is included in the header part, the protocol-specific decoding unit 22-1 displays the conversion information related to the address encoding. The setting is made in the storage unit 53-2 (S202). Thereby, the conversion information regarding common address encoding can be shared between the transmission side and the reception side.

  Next, after the session starts, the address encoding process in the protocol-side encoding unit 12-1 on the transmission side will be described with reference to FIGS.

  The protocol-side encoding unit 12-1 on the transmission side is monitoring a session (S101), and is in communication between opposing devices.

  In this case, the protocol-specific encoding unit 12-1 refers to the conversion information regarding the address encoding in the storage unit 53-1, and replaces the pair of the destination address and the source address of the transmission packet with the corresponding code. Encoding is performed (S104), and the encoded address data is provided to the transmission packet header analysis / encoding unit 11 (S105).

  For example, when the communication device 2-2 returns a response signal to the communication device 2-1, the destination address is “BBBB” and the transmission source address is “AAAA”. In this case, the protocol-specific encoding unit 12-1 refers to the conversion information related to the address encoding and replaces it with the code “02” corresponding to the first address “BBBB” and the second address “AAA”.

  When the communication device 2-1 transmits a packet to the communication device 2-2, the destination address is “AAAA” and the transmission source address is “BBBB”. In this case, the protocol-specific encoding unit 12-1 refers to the conversion information related to the address encoding and replaces it with the code “01” corresponding to the first address “AAA” and the second address “BBBB”.

  On the other hand, the protocol-specific decoding unit 22-1 on the receiving side is monitoring the session (S201), and is in communication between opposing devices.

  In this case, the protocol-specific decoding unit 22-1 refers to the conversion information related to the address encoding in the storage unit 53-2, converts the code of the address data included in the header portion of the received packet, the destination address, and the transmission The packet is decoded into a set (S204), and the decoded destination address and transmission source address are given to the received packet header analysis / decoding unit 21 (S205).

  For example, the header of the packet received by the communication device 2-1 from the communication device 2-2 includes the address data code “02”. Therefore, the protocol-specific decoding unit 21-1 refers to the conversion information related to the address encoding, and decodes the destination address “BBBB” and the transmission source address “AAA” corresponding to the code “02”.

  The header of the packet received by the communication device 2-2 from the communication device 2-1 includes the address data code “01”. Therefore, the protocol-specific decoding unit 21-1 refers to the conversion information related to the address encoding, and decodes the destination address “AAAA” and the source address “BBBB” corresponding to the code “01”.

  Next, the operation at the end of the session will be described. When the session ends, the transmission-side protocol-specific coding unit 12-1 deletes the corresponding entry in the conversion information related to address coding based on the set of the destination address and the source address of the packet (S106). More specifically, the code “01” of the set of the first address “AAAA” and the second address “BBBB” is deleted, and the first address “BBBB” and the second address “AAAA”, which are the opposite combinations, are deleted. The code “02” in the set is also deleted.

  Similarly, the protocol-specific decoding unit 22-1 on the receiving side deletes the corresponding entry in the conversion information related to the address encoding based on the set of the destination address and the source address of the packet when the session ends ( S107).

  As a result, the code can be released and used for a set of addresses used in another session.

(B-2-3) Encoding / Decoding Processing of Header Data by Protocol Next, encoding / decoding of header data in, for example, the encoding unit 12-3 by protocol of the layer 4 protocol and the decoding unit 22-3 by protocol The process will be described with reference to the drawings.

  FIG. 5 is a flowchart showing an operation process in the communication device 2-1 according to the embodiment, and FIG. 6 is a flowchart showing an operation process in the communication device 2-2 according to the embodiment.

  First, it is mutually recognized that the algorithm which encodes an ACK number using a predicted value is implemented between the communication device 2-1 on the transmission side and the communication device 2-2 on the reception side. This can be realized, for example, when the communication device 2-1 and the communication device 2-2 exchange control information indicating the implementation of the algorithm at the start of a session.

  When executing an algorithm for encoding an ACK number, an MTU value used by the receiving side for calculating the ACK number can be applied to each predicted value for predicting the ACK number based on the characteristics of TCP. In the case of TCP communication, the receiving side adds an MTU value to the sequence number of the received data, calculates an ACK number, and transmits an ACK packet including the ACK number to the transmitting side. The transmission side can recognize the MTU value based on the sequence number transmitted to the reception side by the previous packet communication and the ACK number included in the ACK packet received from the reception side this time. Thereby, the transmission side and the reception side can share the MTU value.

  In FIG. 6, in the communication device 2-2, the protocol-specific encoding unit 12-3 monitors a session based on the header data of the layer 4 protocol of the transmission packet (S401). Then, the next header prediction unit 51 of the protocol-specific encoding unit 12-3 obtains a plurality of predicted ACK numbers as predicted values using the MTU value used for calculating the ACK number based on the characteristics of TCP. Then, the next header prediction unit 51 sets each prediction value in the variable setting unit and associates a code with each prediction value and sets conversion information of the next header data in the storage unit 53-2 (S402). The conversion information of the next header data is shared by the encoding unit 12-3 by protocol and the decoding unit 22-3 by protocol of the communication device 2-2.

  In FIG. 5, in the communication device 2-1, the protocol-specific coding unit 12-3 starts the session (S 301), and based on the TCP characteristic information, the ACK number included in the ACK packet received from the receiving side. A plurality of ACK numbers to be predicted next are obtained as predicted values using the MTU value determined based on the The conversion information of the next header data is set in the storage unit 53-1 by associating with the code (S302). The conversion information of the next header data is shared by the protocol encoding unit 12-3 and the protocol decoding unit 22-3 of the communication device 2-1.

  Here, the conversion information of the next header data sets an ACK number based on the TCP characteristics as a predicted value.

  More specifically, as will be described later, in the TCP communication employing the window control method, the packet arrival order is changed when the packet arrives normally, when the packet is lost, or when packet fluctuation occurs. Sometimes, the receiving side has different ACK numbers sent back to the transmitting side.

  In accordance with such a difference in communication status, the next predicted ACK number is set as a plurality of predicted values in the variable condition unit, and a code that replaces the ACK number as each predicted value is associated with each predicted value and the next Set header data conversion information. At this time, the code associated with each prediction value differs depending on the communication status, but an ACK number with a high predicted probability is associated with a code with a short code length, and an ACK number with a low predicted probability is code length. Is associated with a long code.

  Here, for ease of explanation, there is a high probability that packet communication is normally performed between the transmission side and the reception side, and there is a high probability that packet loss will occur next. Assume that the probability of occurrence is high.

  In this case, as shown in FIG. 5 and FIG. 6, since the first predicted value has a high probability of occurrence, the ACK number when received with the correct sequence number is set in the variable condition part, and the code “ 1 ”is associated. In the second predicted value, the ACK number at the time of packet loss is set to the variable code length, and the code “10” is associated. In the third predicted value, the ACK number at the time of fluctuation is set in the variable condition part, and the code “11” is associated.

  The variable condition part and code set in the conversion information of the next header data can change depending on the communication status between the transmission side and the reception side. Thus, the conversion information of the next header data may be changed.

  On the reception side, the MTU value is added to the sequence number received from the transmission side to obtain an ACK number, and an ACK packet including the ACK number is transmitted. This ACK packet is given to the communication device 2-2 as a transmission packet.

  In the communication apparatus 2-2, the next header prediction unit 51 of the protocol-specific coding unit 12-3 refers to the conversion information of the next header data, replaces the ACK number included in the ACK packet with the code, and codes ( S403). Then, the protocol-specific encoding unit 12-3 supplies the header data including the code to the transmission packet analysis / encoding unit 11 (S404). Thereby, header data in which the ACK number is encoded can be formed.

  In the communication device 2-1, the next header prediction unit 52 of the protocol decoding unit 22-3 refers to the conversion information of the next header data and converts the code included in the received ACK packet into an ACK number (S303). ). Then, the protocol-specific decoding unit 22-3 gives the header data including the decoded ACK number to the received packet header analysis / decoding unit 21 (S304). Thereby, header data including the decoded ACK number can be formed.

  Next, the operation at the end of the session will be described. In the communication device 2-1, when the session ends, the protocol-specific encoding unit 12-3 deletes the conversion information of the corresponding next header data (S405).

  Similarly, in the communication device 2-2, when the session ends, the protocol-specific decoding unit 22-3 deletes the conversion information of the corresponding next header data (S305).

  As a result, the code can be released and used for a set of addresses used in another session.

  Next, with reference to FIGS. 7 to 9, the characteristics of TCP communication employing the window control method will be described, and then the ACK number encoding / decoding processing according to the present invention will be specifically described.

  FIG. 7 is a sequence diagram showing TCP processing when the receiving side has normally received all sequence numbers. FIG. 8 is a sequence diagram illustrating a TCP process when a packet with a sequence number “2001 to 3000” is lost. FIG. 9 is a sequence diagram illustrating a TCP process when a packet with a sequence number “2001 to 3000” is fluctuated and the next sequence number “3001 to 4000” is received first.

  In the case of TCP communication, when the receiving side has received all the sequence numbers normally, the receiving side returns the ACK packet including this ACK number to the transmitting side as a value obtained by adding the MTU value to the received sequence number. To do. For example, if the receiving side normally receives the sequence number “0 to 1000”, the transmitting side transmits an ACK packet with “1001” obtained by adding the MTU value to the sequence number “1000” of the received data as the ACK number. Reply to As shown in FIG. 7, the same process is performed for other sequence numbers received normally.

  On the other hand, as shown in FIG. 8, when the packet with the sequence number “2001 to 3000” is lost, the receiving side has not received the packet with the sequence number “2001 to 3000”, and therefore cannot return the ACK packet. Next, when the receiving side receives the packet with the sequence number “3001 to 4000”, the receiving side requests the transmission side to retransmit the packet with the sequence number “2001 to 3000”. ACK packet including the ACK number “2001” is returned to the transmission side.

  Also, as shown in FIG. 9, when the packet with the sequence number “2001-3000” is fluctuated and the next sequence number “3001-4000” is received first, the packet with the sequence number “3001-4000” On the other hand, the receiving side returns an ACK packet including the ACK number “2001” to the transmitting side, and the receiving side receives the ACK number “4001” for the packet with the sequence number “2001 to 3000” received thereafter. An ACK packet is returned to the transmission side.

  As described above, the ACK number included in the ACK packet is different when the sequence number is normally received, when there is a packet loss, when the packet fluctuates, or the like.

  Therefore, if the transmission side and the reception side share the predicted value of the ACK number using the MTU value, the transmission side and the reception side can recognize the packet transmission / reception status in common. Can do.

  Also, because the reception status of the receiving side packet varies depending on the communication status, etc., a code with a short code length is assigned to a prediction value with a high occurrence frequency, and a code with a long code length is assigned to a prediction value with a low occurrence frequency. Can be used to reduce the amount of communication in the transmission section.

  For example, in FIGS. 7 to 9, a case where the transmission side transmits a packet having the sequence number “2001 to 3000” is illustrated.

  In the communication device 2-1 on the transmission side, the packet with the sequence number “2001 to 3000” is given to the transmission processing unit 10.

  This transmission packet does not include control information for executing an algorithm for encoding the next header data using the predicted value. That is, the transmission packet is a normal packet for TCP communication.

  Therefore, basically, the transmission processing unit 10 of the communication device 2-1 transmits the transmission packet as it is to the network 3 without performing the encoding process.

  However, in the communication device 2-1, in order to calculate the predicted value of the next ACK number, the transmission packet header analysis / encoding unit 11 decomposes the layer 4 header of the transmission packet and classifies the layer 4 header by protocol. This is given to the encoding unit 12-3.

  The next header prediction unit 51 of the protocol-specific encoding unit 12-3 obtains the predicted value of the ACK number to be received next based on the characteristics of the TCP, and sets the conversion information of the next header data in the storage unit 53-1. (S302).

  For example, when the receiving side normally receives the packet with the sequence number “2001 to 3000”, the next header prediction unit 51 adds the MTU value shared with the receiving side in advance to the sequence number “3000” “ 3001 "is predicted, and this is set as a first predicted value in the variable condition section. Further, since the frequency of normal reception is high, the code “1” having a short code length is set as the first predicted value “3001”.

  Further, for example, when a packet with the sequence number “2001 to 3000” is lost, the next header prediction unit 51 predicts that the ACK number “2001” is returned in order to request retransmission, and uses this as the second predicted value. Is set in the variable condition section. In this case, the code “10” is set to the second predicted value “2001”.

  Further, for example, when fluctuation occurs in the packet with the sequence number “2001 to 3000”, the next header prediction unit 51 predicts that the ACK number “4001” is returned, and uses this as the third predicted value as a variable setting unit. Set to. In this case, the code “11” is set to the third predicted value “4001”.

  As described above, the communication device 2-1 predicts the predicted value of the next ACK number and sets conversion information of the next header data (see FIG. 5).

  Next, processing in the communication device 2-2 will be described.

  The packet having the sequence number “2001 to 3000” is received by the reception processing unit 20 of the communication device 2-2 on the reception side.

  The transmission packet from the communication device 2-1 is a normal packet for TCP communication. Therefore, basically, the reception processing unit 20 of the communication device 2-2 does not perform the decoding process, and gives the received packet to the terminal 4 as it is.

  However, in order to calculate the predicted value of the next ACK number, the received packet header analysis / decoding unit 21 decomposes the layer 4 header of the received packet and provides the layer 4 header to the protocol-specific decoding unit 22-3.

  The next header prediction unit 52 of the protocol-specific decoding unit 22-3 obtains the predicted value of the ACK number to be returned next based on the characteristics of the TCP, and sets the conversion information of the next header data in the storage unit 53-1. (S402).

  For example, when the packet having the sequence number “2001 to 3000” is normally received, the next header prediction unit 52 adds the MTU value shared with the receiving side in advance to the sequence number “3000” “3001”. Is set as a first predicted value in the variable condition section. Further, since the frequency of normal reception is high, the code “1” having a short code length is set as the first predicted value “3001”.

  Also, for example, when a packet with the sequence number “2001 to 3000” is lost, the next header prediction unit 52 predicts that the ACK number “2001” will be returned in order to request retransmission, and uses this as the second predicted value. Is set in the variable condition section. In this case, the code “10” is set to the second predicted value “2001”.

  Further, for example, when the packet with the sequence number “2001 to 3000” fluctuates, the next header prediction unit 52 predicts that the ACK number “4001” is returned, and uses this as the third predicted value as a variable setting unit. Set to. In this case, the code “11” is set to the third predicted value “4001”.

  As described above, the communication device 2-2 predicts the predicted value of the next ACK number and sets the conversion information of the next header data (see FIG. 6).

  Thereafter, an ACK packet including an ACK number is given from the terminal 4 to the transmission processing unit 10 of the communication device 2-2.

  In the transmission processing unit 10 of the communication device 2-2, the transmission packet header analysis / encoding unit 11 decomposes the layer 4 header and provides the layer 4 header to the protocol-specific encoding unit 12-3.

  The protocol-specific encoding unit 12-3 refers to the conversion information of the next header data, encodes the ACK number (S403), and provides the header data including the code to the transmission packet header analysis / encoding unit 11 (S404). ).

  For example, when the terminal 4 normally receives a packet with the sequence number “2001 to 3000”, the ACK number is “3001”. Therefore, the protocol-specific encoding unit 12-3 refers to the conversion information of the next header data, replaces the ACK number “3001” with the code “1”, and converts the header data including the code “1” into the transmission packet header analysis / It gives to the encoding part 11 (refer FIG. 7).

  Also, for example, when a packet with the sequence number “2001 to 3000” is lost, the ACK number is “2001” because retransmission is requested. Therefore, the protocol-specific encoding unit 12-3 refers to the conversion information of the next header data, replaces the ACK number “2001” with the code “10”, and converts the header data including the code “10” into the transmission packet header analysis / It gives to the encoding part 11 (refer FIG. 8).

  Further, for example, when fluctuation occurs in the packet with the sequence number “2001 to 3000”, the ACK number is “4001”. Therefore, the protocol-specific encoding unit 12-3 refers to the conversion information of the next header data, replaces the ACK number “4001” with the code “11”, and converts the header data including the code “11” into the transmission packet header analysis / It gives to the encoding part 11 (refer FIG. 9).

  On the other hand, in the communication device 2-1, the ACK packet from the communication device 2-1 is received by the reception processing unit 20 via the network 3.

  In the reception processing unit 20 of the communication device 2-1, the received packet header analysis / decoding unit 21 decomposes the layer 4 header and provides the layer 4 header to the protocol-specific decoding unit 22-3.

  The protocol-specific decoding unit 22-3 refers to the conversion information of the next header data, decodes the ACK number (S303), and gives the header data including the decoded ACK number to the received packet header analysis / decoding unit 21 (S304). ).

  For example, when the code of the ACK number is “1”, the protocol-specific decoding unit 22-3 refers to the conversion information of the next header data and decodes the ACK number “3001” corresponding to the code “1”. Header data including the ACK number “3001” is given to the received packet header analysis / decoding unit 21. Thus, by giving the received packet including the header data to the terminal 4, the terminal 4 can recognize that the packet having the sequence number “2001 to 3000” has been normally received.

  For example, when the code of the ACK number is “10”, the protocol-specific decoding unit 22-3 refers to the conversion information of the next header data and decodes the ACK number “2001” corresponding to the code “10”. Header data including the ACK number “2001” is given to the received packet header analysis / decoding unit 21. Thus, by giving the received packet including the header data to the terminal 4, the terminal 4 can recognize that the packet having the sequence number “2001 to 3000” is lost, and retransmits the packet.

  Further, for example, when the code of the ACK number is “11”, the protocol-specific decoding unit 22-3 refers to the conversion information of the next header data and decodes the ACK number “4001” corresponding to the code “11”. Header data including the ACK number “4001” is given to the received packet header analysis / decoding unit 21. As a result, by giving the received packet including the header data to the terminal 4, the terminal 4 can recognize that the packet having the sequence number “2001 to 3000” has fluctuated.

(B-3) Effect of Embodiment As described above, according to this embodiment, the amount of header information necessary for communication can be reduced.

  The embodiment will explain advantages over the prior art. When encoding using conventional statistical information of appearance frequency, the code length remains long until the statistics change. For example, as soon as session start information is detected, the header information can be immediately replaced with a short code. .

  In addition, the conventional compression encoding method using a fixed conversion table has a plurality of conversions because the set of transmission / reception addresses changes one after another, and the ACK number also changes depending on the transmitted packet length. It is not enough to prepare a table in advance. On the other hand, as in this embodiment, since the next information is predicted based on the characteristics of each protocol, the probability that the predicted value is correct is much higher than that of the prior art, and a short code length. Even if it is replaced with, the success probability of communication is high.

(C) Other Embodiments Although various modified embodiments are mentioned in the above-described embodiments, the present invention can also be applied to the following embodiments.

  (C-1) The present invention can perform communication by reversibly compressing the amount of information to be transmitted and reducing the amount of header information.

  In particular, the present invention is effective in a scene where a small amount of traffic is desired, such as a wireless communication section, and a scene where a large number of terminals are expected to communicate with low power consumption, such as communication in an IoT section.

  This is because, for example, in communication in the IoT section, the data length included in the payload is very short, and therefore the data length of the header information accounts for a large proportion of all the data lengths included in the packet, thus reducing the data amount of the header information. This is because there is a strong demand. However, the present invention can be widely applied other than the above scenes.

  Further, it is preferable to mount the communication information in a place where the communication information can be rewritten freely, such as an edge of the SDN or a node on which the NFV function is mounted. Of course, it is not limited to these.

  (C-2) In the embodiment described above, the number of prediction values predicted by the next header prediction unit is not particularly limited. If the number of predicted values increases, the code length replaced with the predicted value becomes longer and the amount of information may increase. Therefore, the number of predicted values (number of algorithms) is set so that the amount of information is probabilistically the smallest. Is preferred.

  Further, it is desirable that information indicating that the header information is compression coding is added to the packet between the transmission side and the reception side. This is because it cannot be distinguished whether the packet includes header information that has been compression-encoded or is a normal packet.

  Furthermore, control information is exchanged between the transmission side and the reception side, and an algorithm based on compression encoding of header information is performed, or information indicating that compression encoding of header information is performed is included. Only in some cases may it be decided whether to implement the algorithm.

  Further, a technique in which the compression coding system of the present invention and the conventional compression coding system are mixed may be used. In other words, when there is a discrepancy in predicted values between the transmission side and the reception side due to the compression coding method of the present invention, the conventional compression coding method or the normal communication method may be performed.

DESCRIPTION OF SYMBOLS 1 ... Communication system, 2-1, 2-2 ... Communication apparatus, 3 ... Network 3 ... Terminal,
DESCRIPTION OF SYMBOLS 10 ... Transmission process part, 11 ... Transmission packet header analysis / encoding part, 12 (12-1 to 12-3) ... Coding part according to protocol, 13 ... Payload encoding part, 14 .... Transmission packet synthetic | combination part, 51 … Next header prediction part,
DESCRIPTION OF SYMBOLS 20 ... Reception processing part, 21 ... Received packet header analysis / decoding part 21 (22-1 to 22-3) ... Decoding part according to protocol, 23 ... Payload decoding part, 24 ... Received packet synthesis part, 52 ... Next header prediction part ,
31... Transmission side input interface unit, 32... Transmission side output interface unit, 33... Reception side input interface unit, 34.

Claims (4)

Based on the communication procedure characteristics of each layer of the communication protocol, one or more predicted values of information to be transmitted or received next are predicted for each layer using information transmitted or received in the past. Prediction means for associating a code with a value to create predicted value correspondence information for each layer;
An encoding unit that refers to the predicted value correspondence information of each layer and replaces information of each layer included in the transmission packet with a corresponding code;
A communication device comprising: decoding means that refers to the predicted value correspondence information of each layer and replaces a code included in each layer of the received packet with a corresponding predicted value.   The prediction means predicts a set of address information or a set of port numbers used in each layer, and associates a code with the set of address information or the set of port numbers. The communication apparatus according to claim 1.   The prediction means predicts a response confirmation number included in a response confirmation signal to be transmitted or received next, using a maximum data size included in a response confirmation number transmitted or received in the past. Or the communication apparatus of 2. Computer
Based on the communication procedure characteristics of each layer of the communication protocol, one or more predicted values of information to be transmitted or received next are predicted for each layer using information transmitted or received in the past. Prediction means for associating a code with a value to create predicted value correspondence information for each layer;
An encoding unit that refers to the predicted value correspondence information of each layer and replaces information of each layer included in the transmission packet with a corresponding code;
A communication program that functions as decoding means that refers to the predicted value correspondence information of each layer and replaces a code included in each layer of a received packet with a corresponding predicted value.

Images