JP2017022687A - Communication control device and communication control method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a communication control device capable of properly detecting a packet loss in a content directivity type network.SOLUTION: A communication control device 102 connected to a content directivity type network includes a transmitter 503 for transmitting a packet of interest, a receiver 502 for receiving a data packet corresponding to the packet of interest, a memory unit 104 for storing transmission PIT in which non-reception information representing a non-received data packet out of data packets corresponding to the packet of interest transmitted by the transmitter is recorded, and a packet processor 505 for detecting a packet loss of data packet corresponding to the packet of interest by using PIT information representing a transmission-uncompleted data packet and transmission PIT when the PIT information is contained in data packets received by a receiver.SELECTED DRAWING: Figure 5

Description

  The present invention relates to a communication control apparatus and a communication control method connected to a content-oriented network, and more particularly to a technique for detecting packet loss.

  In recent years, a content-oriented network represented by CCN (Content Centric Networking) has been proposed as a next-generation network that replaces an IP (Internet Protocol) -based network. For example, Non-Patent Document 1 discloses a technique for distributing content in CCN.

  A characteristic of communication in a content-oriented network is that, when content is acquired, content identification information (content name, etc.) is used instead of identification information (IP address, etc.) of the device storing the content. is there. In communication in general CCN, a device (client) on the receiving side transmits an interest packet (interest packet) related to content without specifying a destination device, and content data corresponding to the transmitted interest packet. A packet is received from any device on the CCN.

  Patent Document 1 discloses a method for detecting occurrence of packet loss in CCN. Specifically, each client measures the elapsed time since the packet of interest was transmitted, and if the data packet cannot be received before the time indicated by a predetermined threshold (timeout value) elapses, the packet Judge that a loss has occurred.

  Patent Document 2 discloses a method for determining that a packet loss has occurred when a receiver of a content data packet checks the sequence number of the received data packet and detects a discontinuity in the sequence number. ing.

Special table 2014-534495 gazette US Patent Application Publication No. 2014/0237085

Alexander Afanasyev, 22 others, “NFD Developer's Guide”, [online], NDN, Technical Report NDN-0021, [July 7, 2015 search], Internet <URL: http: // named-data. net / wp-content / uploads / 2014/07 / NFD-developer-guide.pdf>

  However, further improvements are needed to detect packet loss in CCN.

  One aspect of the present disclosure is a communication control device connected to a content-oriented network, in which a transmission unit that transmits an interest packet, a reception unit that receives a data packet corresponding to the interest packet, and the transmission unit transmits Among the data packets corresponding to the interested packets, a memory unit that stores transmission PIT in which unreceived information indicating data packets that have not been received is recorded, and transmission to the data packet received by the receiving unit is completed. A packet processing unit that detects a packet loss of the data packet corresponding to the packet of interest using the PIT information and the transmission PIT.

  According to the present disclosure, packet loss in a content-oriented network can be detected more appropriately.

It is a figure which shows an example of a structure of the communication control system using a CCN network. It is a figure which shows an example of a structure of PIT. It is a figure which shows an example of schematic structure of each apparatus connected to the CCN network. It is a sequence diagram which shows an example of the process which detects a packet loss in a communication control system. It is a figure which shows an example of a function structure of a router. It is a flowchart which shows an example of the operation | movement which a router processes a packet. It is a flowchart which shows an example of the operation | movement which performs the comparison process of PIT information. It is a flowchart which shows an example of the operation | movement of the process performed when a router receives packets other than a data packet. It is a flowchart which shows an example of the operation | movement of the process performed when a client receives a data packet. It is a figure which shows an example of the format of the packet used by CCN. It is a figure which shows an example of the condition where the interested packet and the data packet are transmitted / received between two routers in CCN.

(Knowledge that became the basis of the present invention)
Hereinafter, the knowledge that is the basis of the present invention will be described.

  FIG. 1 is a diagram illustrating an example of a configuration of a communication control system using a CCN network (hereinafter referred to as a CCN network). A client 100 that requests content is connected to a publisher 106 that provides the content via the CCN network 101. Content data is transmitted and received via a communication interface called a face. One or more routers 102 are connected to the CCN network 101.

  The router 102 includes a CS (Content Store) 103 that stores data of content distributed on the CCN network 101, a PIT (Pending Interest Table) 104 that records received and transferred interest packets, and a route for transferring the interest packets. FIB (Forwarding Information Base) 105 shown.

  In the CCN, in addition to performing communication using the name of the content as an identifier, the CS 103 provided in each router 102 on the CCN network 101 is utilized. Thus, more efficient data distribution can be realized when a plurality of clients 100 acquire the same content. The client 100 and the publisher 106 also have PITs 107 and 108, respectively. Since PITs 104, 107, and 108 have the same configuration, in the following description, PITs 102, 107, and 108 are collectively referred to as PIT 200.

  Here, a configuration of the PIT 200, a method for recording information on the PIT 200, and a method for erasing information recorded on the PIT 200 will be described.

  FIG. 2 is a diagram illustrating an example of the configuration of the PIT 200. The PIT 200 includes an attribute 201 indicating the direction of the interest packet (whether it has been received or transmitted), a content name 202 requested by the interest packet, and a face 203 (face information) indicating the face that received the interest packet. Combined (associated) entries are recorded.

  The PIT 200 includes a reception PIT 204 and a transmission PIT 205. In the reception PIT 204, an entry for the received interest packet, that is, an entry whose attribute 201 is “incoming” in FIG. 2 is recorded. In the transmission PIT 205, an entry about the transmitted interest packet, that is, an entry whose attribute 201 is “outgoing” in FIG. 2 is recorded.

  For example, it is assumed that the router 102 receives an interest packet requesting the content with the name “a” through the face “eth1” and then sends it out from the face “eth0”.

  In this case, as shown in FIG. 2, the router 102 records an entry combining the attribute 201 “Incoming”, the name 202 “a”, and the face 203 “eth1” in the reception PIT 204. Further, as illustrated in FIG. 2, the router 102 records an entry combining the attribute 201 “Outgoing”, the name 202 “a”, and the face 203 “eth0” in the transmission PIT 205.

  As another example, it is assumed that the router 102 is in the process of transferring the interest packet in the router 102 after receiving the interest packet requesting the content with the name “c” by the face “eth0”. That is, it is assumed that the interested packet is not transmitted in the router 102.

  In this case, as illustrated in FIG. 2, the router 102 records only the entry combining the attribute 201 “Incoming”, the name 202 “c”, and the face 203 “eth0” in the reception PIT 204, and records the entry in the transmission PIT 205. do not do.

  Since the router 102 may transfer (receive and transmit) an interest packet, the router 102 records an entry in both the reception PIT 204 and the transmission PIT 205. On the other hand, since the client 100 only transmits an interest packet, the client 100 records an entry only in the transmission PIT 205. Further, since the publisher 106 only receives an interest packet, it records an entry only in the reception PIT 204.

  Hereinafter, an entry stored in the reception PIT 204 is referred to as an entry of the reception PIT 204. An entry stored in the transmission PIT 205 is referred to as an entry of the transmission PIT 205.

  The entry of the transmission PIT 205 is deleted when a data packet corresponding to the entry is received. The data packet corresponding to the entry of the transmission PIT 205 indicates a data packet corresponding to the transmitted interest packet. The data packet corresponding to the interest packet indicates a data packet constituting the content requested by the interest packet.

  For example, as illustrated in FIG. 2, it is assumed that an entry combining the attribute 201 “Outgoing”, the name 202 “a”, and the face 203 “eth0” is recorded in the transmission PIT 205. In this case, it is assumed that the face “eth0” receives a data packet constituting the content with the name “a”. In this case, the router 102 deletes the entry (the combination of the attribute 201 “Outgoing”, the name 202 “a”, and the face 203 “eth0”) corresponding to the data packet.

  That is, by referring to the entry of the transmission PIT 205, it is possible to grasp information regarding data packets that have not been received among the data packets corresponding to the interest packets transmitted by the router 102 and the client 100, respectively.

  For example, by referring to the name 202 included in the entry of the transmission PIT 205, it is possible to grasp what kind of name the data packet that constitutes the content packet that has not received. Further, by referring to the face 203 included in the entry of the transmission PIT 205, it is possible to grasp which data packet that has not been received is a data packet corresponding to an interest packet transmitted by which face.

  As described above, the entry (unreceived information) of the transmission PIT 205 indicates a data packet that has not been received among the data packets corresponding to the transmitted interest packet.

  On the other hand, the entry of the reception PIT 204 is deleted when a data packet corresponding to the entry is transmitted. The data packet corresponding to the entry of the reception PIT 204 indicates an interest packet corresponding to the entry, that is, a data packet corresponding to the received interest packet.

  For example, as illustrated in FIG. 2, it is assumed that an entry combining the attribute 201 “Incoming”, the name 202 “c”, and the face 203 “eth0” is recorded in the reception PIT 204. In this case, it is assumed that the face “eth0” transmits a data packet constituting the content with the name “c”. In this case, the router 102 erases the entry (the combination of the attribute 201 “Incoming”, the name 202 “c”, and the face 203 “eth0”) corresponding to the data packet.

  That is, by referring to the entry of the reception PIT 204, it is possible to grasp information regarding data packets that have not been transmitted among data packets corresponding to the interest packets received by the router 102 and the publisher 106, respectively.

  For example, by referring to the name 202 included in the entry of the reception PIT 204, it is possible to grasp what kind of name the data packet that constitutes the content of the data packet whose transmission has not been completed. Further, by referring to the face 203 included in the entry of the reception PIT 204, it is possible to grasp which data packet whose transmission has not been completed is a data packet corresponding to an interest packet received by the face.

  As described above, the entry (untransmitted information) stored in the reception PIT 204 indicates a data packet that has not been transmitted among data packets corresponding to the received interest packet.

  As described above, the CCN uses the reception PIT 204 and the transmission PIT 205 to control transmission, reception, and transfer of interest packets and data packets.

  However, like conventional IP communication, CCN communication may be provided as a best-effort (no quality guarantee) service. That is, even in communication by CCN, packet loss may occur due to problems of network capacity and communication link quality. Therefore, in order to realize stable communication using CCN, a technique for realizing highly reliable communication on CCN is required. Specifically, packet loss detection and retransmission processing for recovering lost packets.

  In conventional IP communication, a router generally has no state regarding data transfer. For this reason, an end-to-end (E2E) approach, which is performed only by clients or servers located at both ends of the network, is generally used for packet loss detection and retransmission processing.

  On the other hand, the router 102 in the CCN holds the state related to the received interest packet and the transferred interest packet in the PIT 200 in order to perform routing using the name of the content. Therefore, in communication by CCN, the packet loss detection and retransmission processing can be executed by the router 102 on the data packet transfer path.

  As described above, in the communication by CCN, the router 102 located in the middle of the transfer path performs the retransmission process, so that the retransmission process is performed between the publisher 106 and the client 100 located at the end of the transfer path. In comparison, it is possible to quickly detect a packet loss and make a retransmission request. As a result, the influence of data arrival delay or the like due to packet loss can be further reduced.

  However, in order to implement efficient retransmission processing by the router 102 in the transfer path, it is necessary to appropriately detect packet loss. Therefore, in Patent Document 1, when each client measures the elapsed time since the packet of interest was transmitted and the elapsed time exceeds the time indicated by a certain threshold (timeout value), the data packet cannot be received. A method for determining that a packet loss has occurred is disclosed. Further, Patent Document 2 discloses a method of determining that a packet loss has occurred when a sequence number of a data packet received by a device that has requested content is confirmed and a sequence number skip (discontinuity) is detected. ing.

  However, even when the methods disclosed in Patent Document 1 and Patent Document 2 are used, it is difficult to appropriately detect packet loss in communication using CCN.

  In the method disclosed in Patent Document 1, it is very important to set the timeout value to an appropriate value in order to detect packet loss. In general, when determining a timeout value, it is necessary to constantly observe the time required to acquire content data and calculate an optimum timeout value based on the observed value. However, it is difficult to apply such a method in CCN communication because content data acquisition time fluctuates dynamically.

  Even when the content data of the same publisher 106 is continuously acquired, the content data depends on whether the content data exists only in the publisher 106 or is stored in the CS 103 of the router 102 on the CCN. The time required for data acquisition varies greatly. This result is observed as fluctuations in content data acquisition time.

  For example, when content data is acquired from the CS 103 of the router 102, the content data can be acquired in a shorter time than when data is acquired from the publisher 106. However, the client 100 cannot predict whether the requested content data is transmitted from the CS 103 of the router 102 or the publisher 106.

  For this reason, when the timeout value is set in accordance with the acquisition time of the data transmitted from the CS 103 of the router 102, the above-mentioned elapsed time is obtained even though the data is actually transmitted from the publisher 106 and no packet loss occurs. However, since the time indicated by the timeout value has been exceeded, there is a possibility that a packet loss is erroneously detected and a wasteful retransmission process is performed.

  On the other hand, if the timeout value is set in accordance with the acquisition time of the data transmitted from the publisher 106, the packet loss that occurs when the data is transmitted from the CS 103 of the router 102 cannot be detected quickly. Increase the delay in data arrival due to packet loss.

  On the other hand, the method disclosed in Patent Document 2 is a method generally used in real-time communication using IP. However, in communication by CCN, the arrival order of data packets is easily changed depending on whether the content is received from the publisher 106, the CS 103 of the router 102, the router 102, or the like. For this reason, even if packet loss does not occur, sequence number skipping (discontinuity) easily occurs. As a result, the occurrence of packet loss may be erroneously detected.

  An object of the present disclosure is to solve the above-described conventional problems, and to provide a technique capable of more appropriately detecting a packet loss in a content-oriented network.

  A communication control apparatus according to an aspect of the present disclosure is a communication control apparatus connected to a content-oriented network, wherein a transmission unit that transmits an interest packet, a reception unit that receives a data packet corresponding to the interest packet, Among the data packets corresponding to the interest packets transmitted by the transmission unit, a memory unit that stores transmission PIT in which unreceived information indicating data packets that have not been received is recorded, and the data packet received by the reception unit, A packet processing unit that detects packet loss of a data packet corresponding to the packet of interest using the PIT information and the transmission PIT, when PIT information indicating a data packet that has not been transmitted is included With.

  According to this configuration, using the unreceived information recorded in the transmission PIT stored in the memory unit, the data packet corresponding to the interest packet transmitted by the transmission unit is grasped. be able to. Further, when PIT information is included in a data packet received by the receiving unit, it is possible to grasp a data packet that has not been transmitted in the destination device of the packet of interest using the PIT information.

  For this reason, according to this configuration, a data packet that has not been received, which is grasped by using the transmission PIT, and a data packet, which has been grasped by using the PIT information, has not been completely transmitted in the destination device of the interest packet. , The packet loss of the data packet corresponding to the packet of interest can be detected appropriately.

  In addition, the receiving unit further receives the interest packet from another device, and the memory unit further includes data that has not been transmitted among data packets corresponding to the interest packet received by the receiving unit. A reception PIT in which untransmitted information indicating a packet is recorded is stored, the transmission unit further transmits a data packet corresponding to the interest packet to the other device, and the packet processing unit records in the reception PIT The transmitted untransmitted information may be included in the data packet transmitted by the transmitting unit as the PIT information.

  According to this configuration, untransmitted information recorded in the received PIT is included as PIT information in a data packet that the transmitting unit transmits to another apparatus. Accordingly, the other apparatus uses the untransmitted information included in the received data packet as the PIT information, and among the data packets corresponding to the interest packet transmitted by itself, the other apparatus transmits the apparatus to which the interest packet is transmitted. It is possible to grasp data packets that have not been completely transmitted.

  As a result, the other device responds to the interest packet when the grasped data packet and the data packet not received in the data packet corresponding to the interest packet transmitted by itself do not match. It is possible to appropriately detect that a data packet has been lost.

  The untransmitted information includes face information indicating a communication interface that has received an interest packet corresponding to the data packet indicated by the untransmitted information, and the packet processing unit receives from the reception PIT, Face information included in the untransmitted information indicating the data packet to be transmitted is extracted, and the untransmitted information including the same face information as the extracted face information recorded in the reception PIT is converted into the PIT information. In the data packet, the transmission unit may transmit the data packet through a communication interface indicated by the extracted face information.

  According to this configuration, the face information included in the untransmitted information indicating the data packet transmitted by the transmission unit is extracted from the reception PIT. Thereby, the communication interface that has received the interest packet corresponding to the data packet can be specified by the extracted face information. As a result, it is determined that the data packet needs to be returned by the specified communication interface, and the specified communication interface can be specified as a face for transmitting the data packet.

  According to this configuration, the untransmitted information including the same face information as the extracted face information recorded in the reception PIT is included in the data packet transmitted by the transmission unit as the PIT information. As a result, unsent information indicating a data packet that has not been transmitted yet needs to be transmitted through the same communication interface as the identified communication interface is included in the data packet transmitted by the transmission unit. Then, the data packet is sent out by the specified communication interface.

  For this reason, the other device that has received the data packet receives from the untransmitted information included in the data packet as PIT information, the data packet that has not been transmitted by the communication interface that received the interest packet transmitted by itself. I can grasp it.

  As a result, when the grasped data packet and the data packet that has not been received corresponding to the interest packet transmitted by the other device do not match, the other device has the data packet corresponding to the interest packet. It is possible to appropriately detect that there has been a packet loss.

  Further, the packet processing unit deletes the PIT information included in the data packet received by the receiving unit, and then uses the untransmitted information recorded in the received PIT as the PIT information. It may be included in

  According to this configuration, the PIT information included in the data packet is updated to the untransmitted information stored in the reception PIT held by itself, and the data packet including the updated PIT information is transmitted to the other device. The

  For this reason, the other apparatus that has received the data packet can grasp the data packet that the apparatus that transmitted the data packet has not completed transmission from the PIT information included in the data packet. Thereby, the other device can appropriately detect the packet loss in the transfer path between the device that transmitted the data packet and the other device. For this reason, if the communication control device of this configuration is connected to a content-oriented network, it is possible to detect packet loss individually for each data packet transfer path.

  Further, the packet processing unit compares the PIT information included in the data packet received by the receiving unit with the unreceived information recorded in the transmission PIT, and the data packet indicated by the PIT information When the data packet indicated by the unreceived information does not match, a packet loss of the data packet that does not match is detected, and an interest packet corresponding to the data packet that does not match is detected in the transmitter. It may be retransmitted.

  According to this configuration, the data packet that has not been transmitted, indicated by the PIT information, and the data packet that has not been received, indicated by the unreceived information recorded in the transmission PIT, do not match. Packet loss of data packets that do not match can be detected appropriately.

  Further, according to this configuration, when a packet loss of a data packet corresponding to an interest packet is detected, the interest packet corresponding to the data packet can be efficiently retransmitted.

  Further, in the comparison, the packet processing unit does not use the unreceived information whose elapsed time since recording is less than a predetermined threshold among the unreceived information recorded in the transmission PIT. Also good.

  For example, after the first device of this configuration transmits the first interest packet and before the second device receives the first interest packet, the second device corresponding to the second interest packet received in the past is received. Assume that a data packet is transmitted. In this case, since the second device has not received the first interest packet, the second data packet indicates that the transmission of the first data packet corresponding to the first interest packet is not completed. The indicated PIT information cannot be included.

  Thereafter, in the first device, the receiving unit receives the second data packet before the elapsed time from when the unreceived information indicating the first data packet is recorded in the transmission PIT reaches a predetermined threshold. To do.

  In such a case, the first apparatus of this configuration performs the comparison without using unreceived information indicating the first data packet recorded in the transmission PIT. Therefore, the data packet indicated by the PIT information included in the second data packet is prevented from erroneously detecting the packet loss of the first data packet by not including the first data packet. can do.

  Further, the packet processing unit further sets a timeout value for the unreceived information recorded in the transmission PIT, and the unreceived information is recorded after the unreceived information is recorded in the transmission PIT. When the time until the receiving unit receives the data packet indicated exceeds the timeout value, the packet loss of the data packet is detected, and the interested packet corresponding to the data packet is retransmitted to the transmitting unit. It is good as well.

  According to this configuration, every time the elapsed time since the unreceived information corresponding to the data packet is recorded in the transmission PIT exceeds the timeout value, the packet loss of the data packet is detected, and the data packet corresponding to the data packet is detected. The interest packet is retransmitted.

  For this reason, even if a packet loss of a data packet corresponding to an interest packet occurs, an appropriate retransmission process of the data packet can be performed. As a result, the data packet can be acquired without impairing reliability.

  In addition, when the transmission unit transmits each of a plurality of data packets constituting one content, the packet processing unit indicates the data packet indicating the data packet constituting the one content recorded in the reception PIT. Only transmission information may be included in the data packet transmitted by the transmission unit as the PIT information.

  Assume that the transmission unit continuously and intensively transmits a plurality of data packets constituting one content for a certain period. According to the present configuration, in such a case, the untransmitted information indicating the data packet constituting the content other than the content recorded in the received PIT is included in the transmitted data packet as the PIT information. It can be avoided. Thereby, the size of the PIT information included in the data packet can be reduced during the period. As a result, the communication load of the content-oriented network can be reduced.

  In addition, when the reception unit receives a plurality of data packets constituting one content, the packet processing unit is recorded in the PIT information included in the received data packet and the transmission PIT. The detection may be performed using the unreceived information indicating the data packet constituting the one content.

  It is assumed that the receiving unit continuously and intensively receives each of a plurality of data packets constituting one content for a certain period. According to the present configuration, in such a case, the PIT information included in the received data packet and the unreceived information indicating the data packet constituting the content recorded in the transmission PIT during the period. It is possible to centrally detect only packet loss of data packets constituting the content.

  Moreover, this indication discloses not only a communication control apparatus provided with the process part for performing the above characteristic processes but the communication control method which performs the said characteristic process in the said communication control apparatus.

  Note that each of the embodiments described below shows a specific example of the present disclosure. Numerical values, shapes, components, steps, order of steps, and the like shown in the following embodiments are merely examples, and are not intended to limit the present disclosure. In addition, among the constituent elements in the following embodiments, constituent elements that are not described in the independent claims indicating the highest concept are described as optional constituent elements. In all the embodiments, the contents can be combined.

(Embodiment 1)
In the communication control system using the CCN network (hereinafter referred to as a CCN network) in the first embodiment, the publisher 106 and the router 102 send information indicated by the entry of the reception PIT 204 of each device to the lower router 102 and the client 100. And transmitting. Further, the lower router 102 and the client 100 detect the packet loss using the information indicated by the received entry of the received PIT 204 of the other apparatus.

  Here, the higher-level device refers to a device (a device to which the interest packet is transmitted) located in the direction in which the interest packet is transmitted as viewed from a certain device on the CCN network 101. The device on the CCN network 101 indicates a device connected to the CCN network 101, that is, the client 100, the router 102, or the publisher 106. The low-order device indicates a device (data packet transmission destination device) located in the direction in which the data packet corresponding to the interest packet is transmitted.

  Specifically, each device on the CCN network 101 extracts a face 203 included in an entry corresponding to the data packet from its reception PIT 204 when transmitting the data packet to a lower device.

  Thereby, each apparatus specifies the face indicated by the extracted face 203 as the face that has received the interest packet corresponding to the data packet. Each device determines that the data packet needs to be returned from the identified face, and identifies the face indicated by the extracted face 203 as a face to which the data packet is transmitted.

  Therefore, each device extracts the name 202 from the entry including the same face 203 as the extracted face 203 among the entries recorded in its own reception PIT 204. Each device includes the extracted name 202 in the data packet as information indicating a data packet that has not been transmitted (hereinafter referred to as PIT information). Thereby, each device needs to transmit by the same face as the identified face, but includes information indicating a data packet that has not been transmitted in the data packet.

  Then, each device causes the data packet to be transmitted to the face indicated by the extracted face 203.

  On the other hand, the lower-level device that has received the data packet includes the PIT information obtained from the entry of the higher-level device's reception PIT 204 and the entry of its own transmission PIT 205 that are included in the received data packet. The face 203 indicating the face that received the data packet is compared with an entry including the same face 203.

  As a result of the comparison, if the name 202 indicated by the PIT information matches the name 202 included in the entry of the transmission PIT 205, the device does not detect a packet loss. On the other hand, if they do not match, the device detects the packet loss of the data packet that constitutes the content of the name 202 that does not match. Then, according to the entry of its own transmission PIT 205, the apparatus retransmits the interest packet corresponding to the data packet constituting the content of the name 202 that does not match.

  Also, when transferring each data packet, each device deletes the PIT information included in the received data packet, and then includes the information of the entry of its own received PIT 204 as the PIT information in the data packet. To the device. That is, each device updates the PIT information included in the received data packet to the information indicated by the entry of its own reception PIT 204. As a result, it is possible to individually detect the packet loss and control retransmission of the lost data packet for each transfer path of the data packet.

  Hereinafter, it explains in detail using a drawing. FIG. 3 is a diagram illustrating an example of a schematic configuration of each device connected to the CCN network 101 (content-oriented network). The client 100 is connected to the publisher 106 via the router A 102A and the router B 102B corresponding to the router 102 described above. The client 100, the router A 102A, the router B 102B, and the publisher 106 include a PIT 107, a PIT 104A, a PIT 104B, and a PIT 108, respectively.

  Reference numeral 307 denotes a face “eth0” included in the client 100. Reference numerals 308 and 309 denote faces “eth0” and “eth1” included in the router A 102A. Reference numerals 310 and 311 denote faces “eth0” and “eth1” included in the router B 102B. Reference numeral 312 denotes a face “eth0” included in the publisher 106.

  Hereinafter, in the communication control system shown in FIG. 3, when the client 100 requests two contents, a loss of one data packet occurs in the data packet transfer path (link) between the router A 102A and the router B 102B. A process for detecting a packet loss will be described using an example.

  FIG. 4 is a sequence diagram illustrating an example of processing for detecting packet loss in the communication control system. Hereinafter, the client 100 requests two contents with names “a” and “b”, and data packets constituting the contents with the name “a” are lost when the data is transmitted from the router B 102B to the router A 102A. Will be described.

  Hereinafter, the content with the name “a” is referred to as content “a”. An interest packet requesting the content “a” is described as “I: a”. A data packet constituting the content “b” is described as “D: b”. When the data packet “D: a” includes the content name “a” as PIT information, “D: a: PIT [a]” is described. The PIT information is referred to as PIT information “PIT [a]”.

  An entry of the reception PIT 204 that combines the attribute “incoming”, the name “a”, and the face “eth0” is described as an entry “i [a (eth0)]” of the reception PIT 204. An entry of the transmission PIT 205 that combines the attribute “outgoing”, the name “a”, and the face “eth1” is described as an entry “o [a (eth1)]” of the transmission PIT205.

  In the PIT 200 (107, 104A, 104B, 108), the entries “i [a (eth0)]”, “i [b (eth0)]” of the two reception PITs 204, and the entries “o [a ( eth1)] ”and“ o [b (eth1)] ”are described as“ PIT [i [a, b (eth0)] o [a, b (eth1)]] ”.

  In the initial state, it is assumed that no data packets are stored in the CS 103 of the router A 102A and the router B 102B.

  Hereinafter, the sequence shown in FIG. 4 will be described in order. First, in process 404, the client 100 transmits interest packets “I: a” and “I: b” requesting the contents “a” and “b”. As a result of the transmission, the client 100 records an entry indicating that the interest packet “I: a” “I: b” is transmitted by the face “eth0” in the transmission PIT 205. As a result, the state of the PIT 107 becomes a state 407 (“PIT [o [a, b (eth0)]]”).

  In processing 408 and processing 409, the router A 102A and the router B 102B transfer the interest packets “I: a” and “I: b” transmitted in processing 404 to the publisher 106. As a result of the transfer, the router A 102 A records the entry “i [a, b (eth 0)]” indicating that the face 308 “eth 0” has received the interest packet “I: a” “I: b” in the reception PIT 204. To do. Further, the router A 102A records an entry “o [a, b (eth1)]” indicating that the face packet 309 “eth1” has transmitted the interest packets “I: a” and “I: b” in the transmission PIT 205.

  As a result, the state of the PIT 104A of the router A 102A is changed from the initial state 405 “PIT []” in which no entry is recorded to the state 411 “PIT [i [a, b (eth0)] o [a, b (eth1)”. ]] ”. Similarly, the PIT 104B of the router B 102B is also updated from the initial state 406 “PIT []” to the state 412 “PIT [i [a, b (eth0)] o [a, b (eth1)]]”.

  When the face 312 “eth0” receives the interest packet “I: a” “I: b” transmitted in the process 409, the publisher 106 records an entry indicating the reception in the reception PIT 204. As a result, the state of the PIT 108 is updated from an initial state “PIT []” (not shown) to a state 410 “PIT [i [a, b (eth0)]]”.

  Next, in process 414, the publisher 106 performs a process of transmitting a data packet “D: a” that constitutes the content “a” owned by the publisher 106. Specifically, when the publisher 106 starts the process 414, first, in the entries “i [a (eth0)]” and “i [b (eth0)]” recorded in its reception PIT 204, The face 203 “eth0” included in the entry “i [a (eth0)]” corresponding to the data packet “D: a” is extracted.

  Thereby, the publisher 106 specifies the face 312 “eth0” indicated by the extracted face 203 “eth0” as the face that has received the interest packet corresponding to the data packet “D: a”. Then, the publisher 106 determines that the data packet “D: a” needs to be returned from the specified face 312 “eth0”, and sends the specified face 312 “eth0” to the data packet “D: a”. ”Is specified as a face to be transmitted.

  Next, the publisher 106 selects the above-described extracted face 203 “eth0” in the entries “i [a (eth0)]” and “i [b (eth0)]” recorded in its reception PIT 204. The names 202 “a” and “b” are extracted from the entries “i [a (eth0)]” and “i [b (eth0)]” including the same face 203 “eth0”. Then, the publisher 106 includes the extracted names 202 “a” and “b” as PIT information in the data packet “D: a”.

  Thus, the entry corresponding to the data packet “D: a” or “D: b” that needs to be transmitted by the same face 312 “eth0” as the identified face 312 “eth0” but has not been transmitted yet. Information of “i [a (eth0)]” and “i [b (eth0)]” (names 202 “a” and “b”) is included in the data packet “D: a” as PIT information.

  Then, the publisher 106 transmits the data packet “D: a: PIT [a, b]” to the identified face 312 “eth0”, and ends the processing 414.

  When the publisher 106 completes the transmission of the data packet “D: a: PIT [a, b]” in the process 414, the publisher 106 receives the received data packet “D: a: PIT [a, b]”. The entry “i [a (eth0)]” is deleted. Thus, the publisher 106 updates the state of the PIT 108 from the state 410 “PIT [i [a, b (eth0)]]” to the state 420 “PIT [i [b (eth0)]]”.

  On the other hand, the router B 102B receives the data packet “D: a: PIT [a, b]” through the face 311 “eth1”. Then, the router B 102B transmits the content names “a” and “b” indicated by the PIT information “PIT [a, b]” included in the received data packet and the transmission PIT 205 of its own in the state 412. Among the entries, included in the entries “o [a (eth1)]” and “o [b (eth1)]” including the same face 311 “eth1” as the face 311 “eth1” that received the data packet The names 202 “a” and “b” are compared.

  In this case, the content names “a” and “b” indicated by the PIT information “PIT [a, b]” and the content names 202 “a” and “b” obtained from the entry of its own transmission PIT 205 are identical. I'm doing it. That is, the transmission is not completed by the higher-level publisher 106 indicated by the PIT information, the two data packets constituting the contents “a” and “b”, and the entry of the own transmission PIT 205 indicate that it is not received by itself. , The two data packets constituting the contents “a” and “b” match. Therefore, the router B 102B determines that no packet loss has occurred in the data packet transfer path between the upper publisher 106 and itself.

  Therefore, the router B 102B performs a process of transferring the data packet “D: a: PIT [a, b]” received from the publisher 106 in process 413. Specifically, when starting the processing 413, the router B 102B first deletes the PIT information from the received data packet “D: a: PIT [a, b]”. Then, similarly to the process 414, the router B 102B deletes the PIT information “PIT [a, b]” obtained from the entry recorded in its own reception PIT 204 after deleting the PIT information at the start of the process 413. It is included in the data packet “D: a”. Then, similarly to the process 414, the router B 102B transmits the data packet “D: a: PIT [a, b]” by the face 310 “eth0” specified by using the entry recorded in the reception PIT 204 of the router B 102B. Then, the process 413 is terminated.

  Then, the router B 102B changes the state of its own PIT 104B from the state 412 “PIT [i [a, b (eth0)] o [a, b (eth1)]]” to the state 415 “PIT [i [b (eth0)”. ]] O [b (eth1)]] ”. Also, the router B 102B deletes the PIT information included in the data packet “D: a: PIT [a, b]” sent in the process 413, and deletes the data packet “D: a” after the deletion. Store in CS103.

  Then, it is assumed that the data packet “D: a: PIT [a, b]” transmitted in the process 413 is lost. At this time, the state of the PIT 104B of the router B 102B is the state 415 “PIT [i [b (eth0)] o [b (eth1)]]”. On the other hand, the router A 102A has not received the data packet “D: a: PIT [a, b]” transmitted in the process 413. Therefore, the state of the PIT 104A of the router A 102A remains unchanged as the state 411 “PIT [i [a, b (eth0)] o [a, b (eth1)]]”, and the data packet “D: The entry “o [a (eth1)]” corresponding to “a” remains. The packet loss is detected when the router A 102A receives the data packet of the content “b”.

  Thereafter, in the process 417, the publisher 106 transmits the data packet “D: b: PIT [b]” including the PIT information as in the process 414. When the router B 102B receives the data packet “D: b: PIT [b]”, the router B 102B performs the same comparison process as described above. As a result, the content name “b” indicated by the PIT information “PIT [b]” included in the data packet “D: b: PIT [b]” and the content obtained from the entry of the transmission PIT 205 in the PIT 104B in the state 415 Therefore, the router B 102B determines that no packet loss has occurred. Therefore, the router B 102B transfers the data packet “D: b: PIT [b]” including the PIT information in the process 416 similarly to the process 413.

  When the router A 102A receives the data packet “D: b: PIT [b]” transmitted in the process 416, the router A 102A performs the same comparison process as described above. That is, the router A 102A is in the state 411 with the content name “b” indicated by the PIT information “PIT [b]” included in the received data packet “D: b: PIT [b]”. Compares the names 202 “a” and “b” included in the entry including the same face 309 “eth1” as the face 309 “eth1” that received the data packet in the entry of its own transmission PIT 205 To do.

  In this case, the content name “a” in the content name “b” indicated by the PIT information “PIT [b]” and the content name 202 “a” “b” obtained from the entry of its own transmission PIT 205. Does not match. In other words, the data packet constituting the content “b” that is not completely transmitted by the upper publisher 106 indicated by the PIT information and the content “a” “not received by itself that is indicated by the entry of the own transmission PIT 205. In the two data packets constituting “b”, the data packets constituting the content “a” do not match. For this reason, in the process 418, the router A 102A detects that a packet loss of the data packet constituting the content “a” has occurred in the data packet transfer path between the upper router B 102B and itself.

  When the router A 102A detects a packet loss of the data packet constituting the content with the name “a”, the router A 102A executes the process 422. Specifically, in the process 422, the router A 102A follows the entry “o [a (eth1)]” corresponding to the data packet “D: a” of its transmission PIT 205 in the state 411 in accordance with the data packet. The interest packet “I: a” corresponding to is transmitted (retransmitted) again by the face “eth1”.

  In the process 421, the router A 102A transfers the data packet “D: b: PIT [a, b]” including the PIT information “PIT [a, b]” in the same manner as the processes 413 and 416. That is, in this data packet transfer, the router A 102A uses the PIT information “PIT [a, b]” obtained from the entry “i [a, b (eth0)” of its own received PIT 204 in the state 411 as the data packet. Is included.

  Upon receiving the data packet “D: b: PIT [a, b]”, the client 100 performs the same comparison process as described above. In this case, the content names 202 “a” and “b” included in the entry “o [a, b (eth0)” of the transmission PIT 205 in the state 407 and the PIT information “PIT” included in the data packet. The content names “a” and “b” indicated by “[a, b]” match. Therefore, the client 100 determines that no packet loss has occurred. That is, the client 100 does not detect a packet loss and does not retransmit the interest packet.

  When the router B 102B receives the interest packet “I: a” retransmitted in the process 422, the router B 102B acquires the data packet “D: a” stored in the CS 103 on the router B 102B after the process 413 (process 425). Then, the router B 102B performs a process 427 similar to the processes 413 and 416, and the router A 102A performs a process 426 similar to the processes 413 and 416. Thereby, the client 100 receives the data packet “D: a” corresponding to the interest packet “I: a” retransmitted in the process 422.

  In the state where the transfer of all the data packets requested by the client 100 is completed, the entries are recorded in the states 429, 430, 431, and 432 of the PITs 107, 104A, 104B, and 108 of the devices 100, 102A, 102B, and 106, respectively. The initial state is not performed.

  As described above, in the communication control system according to the first embodiment, the packet loss can be detected more appropriately by transmitting and receiving the data packet including the information indicated by the entry of the PIT 200, and efficient retransmission control is performed. It becomes possible.

  In the first embodiment, CCN is taken up as a representative example of a content-oriented network. However, the content-oriented network is not limited to the CCN network 101 but may be another content-oriented network such as NDN (Named Data Networking).

  Further, in the first embodiment, the communication control device according to the present disclosure is described in detail as the router 102 (102A and 102B). The communication control device according to the present disclosure is a client 100 connected to the CCN network 101, and receives a data packet including PIT information indicating a data packet that has not been transmitted, which is indicated by an entry of a reception PIT 204 of a higher-level device. The packet loss detection and retransmission control are performed using the PIT information included in the received data packet. Further, the communication control device according to the present disclosure is a publisher 106 in the CCN, receives an interest packet, and transmission is completed in a data packet corresponding to the received interest packet indicated by an entry of its own reception PIT 204. A data packet including PIT information indicating a data packet that has not been transmitted is transmitted.

  Next, the configuration of the router 102 will be described in detail. FIG. 5 is a diagram illustrating an example of a functional configuration of the router 102. The router 102 includes a CPU (not shown), a memory such as a RAM and a ROM, a communication interface (face), and the like. Each processing unit 502 to 506 shown in FIG. 5 is realized by causing the CPU to execute a program stored in the memory and functioning as each processing unit 502 to 506. Note that the processing units 502 to 506 illustrated in FIG. 5 may be realized by combining dedicated circuits that function as the processing units 502 to 506. The memory unit 510 is realized by a storage area of the memory.

  The packet reception unit 502 (reception unit) and the packet transmission unit 503 (transmission unit) are connected to the CCN network 101 via the communication interface (face), and transmit and receive interest packets and data packets. The packets transmitted and received by the packet reception unit 502 and the packet transmission unit 503 with the CCN network 101 are not limited to the interest packet and the data packet. The packet to be transmitted / received may be, for example, a control packet for exchanging routing information, or a NACK (Negative Acknowledgement) packet indicating that no content exists or the network is congested.

  The packet processing unit 504 is connected to the packet receiving unit 502, the packet transmitting unit 503, and the memory unit 510, and is transmitted and received by the packet receiving unit 502 and the packet transmitting unit 503 using information stored in the memory unit 510. The processing method of the packet to be processed. The packet processing unit 504 further operates as a PIT information comparison unit 505 (packet processing unit) described later and a PIT information addition unit 506 (packet processing unit) described later.

  The memory unit 510 stores the above-described CS 103, PIT 104, and FIB 105.

  Hereinafter, in the communication control system according to the first embodiment, an operation when the router 102 or the client 100 receives a packet will be described in detail with reference to FIGS. In the following, description of the same processing as that described in detail with reference to FIG. 4 is simplified.

  FIG. 6 is a flowchart illustrating an example of an operation in which the router 102 processes a packet. First, in FIG. 6, the processing flow is activated by reception of a packet by the packet receiving unit 502 (S600). The packet processing unit 504 determines whether or not the packet received by the packet receiving unit 502 in step S600 is a data packet (S601).

  Specifically, in step S601, the packet processing unit 504 refers to information indicating the packet type stored in the header area of the packet used in the CCN. Then, the packet processing unit 504 determines whether or not the received packet is a data packet depending on whether or not the referred information indicates a data packet. A packet format used in the CCN will be described later.

  When the packet processing unit 504 determines that the received packet is not a data packet (S601; NO), the packet processing unit 504 performs step S800 described later (S602). If the packet processing unit 504 determines that the received packet is a data packet (S601; YES), the packet processing unit 504 calls a PIT information comparison process (S700) described later (S603).

  Subsequently, as described above with reference to FIG. 2, the packet processing unit 504 deletes the entry corresponding to the data packet received in step S600 from the transmission PIT 205 in the PIT 104 (S604). Then, the packet processing unit 504 determines whether or not the entry corresponding to the data packet received in step S600 is recorded in the reception PIT 204 of the router 102 itself (S605).

  If the packet processing unit 504 determines that the data is not recorded in step S605 (S605; NO), the packet processing unit 504 stores the data packet in the CS 103 without transferring the data packet received in step S600. As a result, when an interest packet requesting the content that the data packet constitutes is received thereafter, the data packet can be acquired from the CS 103 and transferred. If the data packet received in step S600 is already stored in the CS 103, the data packet is discarded (S606). Then, the packet processing unit 504 ends the packet transfer process (S611).

  On the other hand, it is assumed that the packet processing unit 504 determines that the data is recorded in step S605 (S605; YES). In this case, the PIT information adding unit 506 updates the PIT information included in the data packet as preparation for transferring the data packet received in step S600 to the lower router 102 (S607).

  Specifically, in step S607, the PIT information adding unit 506 first deletes the PIT information included in the data packet. Then, the PIT information adding unit 506 extracts the face 203 included in the entry corresponding to the data packet, which is recorded in the reception PIT 204 of the router 102 itself, similarly to the process 413 of FIG. As a result, the face indicated by the extracted face 203 is specified as the face to which the data packet is transmitted. Then, the PIT information adding unit 506 includes the name 202 included in the entry including the same face 203 as the extracted face 203 recorded in the received PIT 204 of the router 102 itself as the PIT information in the data packet. Then, the PIT information adding unit 506 ends Step S607.

  Next, the packet processing unit 504 causes the packet transmission unit 503 to transfer the data packet received in step S600 to the lower router 102 (S608). Specifically, in step S608, the packet processing unit 504 instructs the packet transmission unit 503 to send out the data packet received in step S600 by the face specified in step S607. Thereby, the packet transmission unit 503 transmits the data packet to the face specified in step S607. As a result, the data packet is transferred to the lower router 102.

  Then, the packet processing unit 504 deletes the entry corresponding to the data packet transferred in step S608 from the reception PIT 204 of the router 102 itself in the same manner as after performing the processing 413 in FIG. 4 (S609). Then, the packet processing unit 504 stores the same data packet as the data packet transferred in step S608 in the CS 103 (S610). Then, the packet processing unit 504 ends the packet transfer process (S611).

  Next, details of the PIT information comparison process (S700) performed in step S603 will be described with reference to FIG. FIG. 7 is a flowchart of the PIT information comparison process (S700).

  When the PIT information comparison process is called by the packet processing unit 504 (S700), the PIT information comparison unit 505 determines whether or not the PIT information is included in the data packet received in step S600 (S701). .

  The data packet received in step S600 may be a data packet transmitted from a publisher 106 or the like of a communication control system other than the communication control system of the first embodiment, and PIT information is included in the data packet. It may not be. Therefore, if the PIT information comparison unit 505 determines in step S701 that the PIT information is not included in the data packet (S701; NO), the PIT information comparison unit 505 determines that the data packet is transmitted from another communication control system device. Then, the PIT information comparison process ends. As a result, the process returns to the caller process (S705).

  When step S607 is performed after returning to the caller process, the PIT information is included in the data packet, and in step S608, the data packet including the PIT information is transferred to the lower device. As a result, the lower apparatus can detect a packet loss.

  On the other hand, it is assumed that the PIT information comparison unit 505 determines in step S701 that the data packet includes PIT information (S701; YES). In this case, the PIT information comparison unit 505 extracts the name 202 from the entry including the same face 203 as the face 203 indicating the face that has received the data packet in the entry of the transmission PIT 205 of the router 102 itself. The PIT information comparison unit 505 compares the name 202 indicated by the PIT information included in the data packet with the name 202 extracted from the transmission PIT 205 of the router 102 itself (S702).

  If the comparison result in step S702 shows that the name 202 extracted from the transmission PIT 205 and the name 202 indicated by the PIT information included in the data packet do not match (S703; YES), the PIT information comparison unit 505 It detects that a packet loss of a data packet constituting the content of the name 202 that does not match occurs in the packet transfer path between itself and the host device. In this case (S703; YES), the packet processing unit 504 performs an interest packet corresponding to the data packet that constitutes the content of the name 202 that does not match, according to the entry of the transmission PIT 205, in the same manner as the processing 422 of FIG. The packet transmission unit 503 retransmits the data, and the PIT information comparison process ends (S704). As a result, the process returns to the caller process (S705).

  On the other hand, if the comparison result in step S702 shows that the name 202 extracted from the transmission PIT 205 matches the name 202 indicated by the PIT information included in the data packet (S703; NO), the PIT information comparing unit 505 Is not generated, and the PIT information comparison process is terminated. As a result, the process returns to the caller process (S705).

  Next, details of the processing performed in step S602 when the router 102 receives a packet other than the data packet corresponding to the interest packet will be described with reference to FIG. FIG. 8 is a flowchart showing an operation of processing performed when the router 102 receives a packet other than the data packet corresponding to the interest packet.

  When the packet processing unit 504 starts the process in step S602 (S800), the packet processing unit 504 determines whether the packet received by the packet receiving unit 502 in step S600 is an interest packet (S801).

  Specifically, in step S801, the packet processing unit 504 refers to information indicating the type of packet included in the header area of the packet used in CCN. Then, the packet processing unit 504 determines whether or not the received packet is an interest packet depending on whether or not the referred information indicates an interest packet. A packet format used in the CCN will be described later.

  If the packet processing unit 504 determines that the received packet is not an interest packet (S801; NO), the packet processing unit 504 performs a predetermined process defined for the received packet, and ends the process (S802).

  On the other hand, when the packet processing unit 504 determines that the received packet is an interest packet (S801; YES), the packet processing unit 504 receives an entry indicating that the interest packet has been received in the same manner as the processing 408 and 409 in FIG. It adds to PIT (S803). Next, the packet processing unit 504 searches for a data packet corresponding to the interest packet in the CS 103 (S804).

  When the data packet corresponding to the interested packet is not stored in the CS 103 (S805; NO), the packet processing unit 504 instructs the packet transmission unit 503 to perform the same as the processing 408 and 409 in FIG. The face used for communication with the host device is instructed to transfer the interested packet to the host device (S806). As a result, the interested packet is transferred to the host device. The packet processing unit 504 records an entry indicating that the interested packet has been transmitted in the transmission PIT 205 (S807), and ends the process (S811).

  On the other hand, when the packet processing unit 504 determines that the data packet corresponding to the interest packet is stored in the CS 103 (S805; YES), the PIT information adding unit 506 acquires the data packet from the CS 103, and step S607 is performed. Similarly, the PIT information obtained from the entry recorded in its own reception PIT 204 is included in the data packet (S808). In step S808, the PIT information adding unit 506 identifies the face that transmits the data packet in the same manner as in step S607.

  Then, the packet processing unit 504 causes the packet transmission unit 503 to transfer the data packet to the lower device (S809). Specifically, in step S809, the packet processing unit 504 instructs the packet transmission unit 503 to transmit the data packet including the PIT information in step S808 by the face specified in step S808. As a result, the packet transmission unit 503 transmits the data packet to the face specified in step S808. As a result, the data packet is transferred to the lower router 102.

  Then, the packet processing unit 504 deletes the entry of its own reception PIT 204 corresponding to the transferred data packet (S810), and ends the processing (S811).

  The processing flow in the router 102 has been described above. When the packet loss detection method in the router 102 is applied to the client 100 and the publisher 106, the client 100 and the publisher 106 may be configured as follows.

  Specifically, the client 100 and the publisher 106 may include processing units 502 to 506 and a memory unit 510 similar to those of the router 102 illustrated in FIG.

  In the client 100, after the packet transmission unit 503 (hereinafter referred to as the packet transmission unit 503a) included in the client 100 transmits the interest packet, the packet processing unit 504 (hereinafter referred to as packet processing) included in the client 100 is transmitted in the same manner as in step S807. The unit 504a) may add an entry indicating that the interested packet has been sent to its own transmission PIT 205.

  On the other hand, the operation in which the client 100 processes the data packet may be as follows. FIG. 9 is a flowchart showing an operation of processing performed when the client 100 receives a data packet.

  When the packet receiving unit 502 (hereinafter referred to as the packet receiving unit 502a) included in the client 100 receives the data packet (S900), the packet processing unit 504a calls the PIT information comparison processing (S700) (S901). Accordingly, the PIT information comparison unit 505 included in the client 100 performs the PIT information comparison process according to the processing flow illustrated in FIG. As a result, the packet loss can be detected also in the client 100, and when the packet loss is detected, the retransmission process in step S704 can be performed.

  Next, the packet processing unit 504a determines whether an entry corresponding to the data packet received in step S900 is recorded in the transmission PIT 205 of the client 100 itself (S902).

  If the packet processing unit 504a determines that it is not recorded in step S902 (S902; NO), the packet processing unit 504a completes the reception process as it is (S904). On the other hand, if the packet processing unit 504a determines that it is recorded in step S902 (S902; YES), as described above with reference to FIG. 2, the entry corresponding to the data packet received in step S900 from the transmission PIT 205 is used. (S903), the reception process is completed (S904).

  On the other hand, in the publisher 106, after the packet receiving unit 502 provided in the publisher 106 receives the interest packet, the packet processing unit 504 provided in the publisher 106 (hereinafter referred to as the packet processing unit 504b) May be added to its own reception PIT 204.

  When the packet transmission unit 503 (hereinafter referred to as the packet transmission unit 503b) included in the publisher 106 transmits a data packet, the PIT information addition unit 506 included in the publisher 106 is used by the publisher 106 in the same manner as in steps S808 and S809. After the PIT information obtained from the entry of the received PIT 204 is included in the data packet, the packet processing unit 504b may cause the packet transmission unit 503b to transmit the data packet. Then, after sending the data packet, the packet processing unit 504b may delete the entry corresponding to the data packet from its own reception PIT 204 in the same manner as in step S810.

  Next, the format of a packet used in CCN will be described. FIG. 10 is a diagram illustrating an example of a format of a packet used in CCN. As shown in FIG. 10, the data packet includes a header area (1000 to 1006) and a message area (1007 to 1012). The header area (1000 to 1006) is common to all packets used in the CCN, and the message area differs depending on the packet type.

  Information indicating the version of the packet is stored in an area 1000 (hereinafter, version area 1000). Information indicating a packet type is stored in an area 1001 (hereinafter, type area 1001). For example, the type area 1001 stores information indicating whether the packet is an interest packet or a data packet. That is, in the above-described steps S601 and S801, the type area 1001 is referred to. In the area 1002, the packet length is stored. The area 1003 stores information specific to the packet type, and the area 1004 stores the header length. In the area 1005, information for identifying information and data stored in the message area is stored. A message length is stored in an area 1006.

  For example, when the packet is the data packet described above, an identifier for identifying the data packet is stored in the type area 1001. In the area 1005, the name of the content is stored. The message area (1007 to 1012) is classified into two areas: an area of 1007 to 1009 and an area of 1010 to 1012. In the areas 1007 to 1009, content data included in the data packet and information related to the data are stored. Further, the areas 1010 to 1012 store PIT information and data indicating the PIT information.

  Specifically, an identifier indicating content data is stored in the area 1007, the data length of the content is stored in the area 1008, and data constituting the content is stored in the area 1009. The An area 1010 stores an identifier indicating PIT information, an area 1011 stores the data length of the PIT information, and an area 1012 stores PIT information.

  On the other hand, when the packet is the above-described interest packet, the type area 1001 stores an identifier for identifying the interest packet. In the area 1005, the name of the content requested by the interested packet is stored. The message area (1007 to 1012) is not used or does not exist.

(Embodiment 2: Compression of PIT information to be transmitted)
In the first embodiment, when PIT information is included in a data packet to be transmitted by each device, the face 203 included in the entry corresponding to the data packet is extracted from its own reception PIT 204, and then the entry of its own reception PIT 204 The name 202 is extracted from all entries including the same face 203 as the extracted face 203, and this is used as PIT information. However, it is not always necessary to include the names 202 extracted from all the entries as PIT information in all data packets to be transmitted. As shown below, the size of the PIT information included in the data packet to be transmitted may be reduced.

  For example, content whose name is “a” is further divided into content having names of sequential child hierarchies such as “a / 1”, “a / 2”, “a / 3”,. There may be. Therefore, for example, the content with the name “a” (content “a”) has three contents (contents) with the names “a / 1”, “a / 2”, and “a / 3” as described above. “A / 1”, content “a / 2”, content “a / 3”). In addition, the content with the name “b” (content “b”) also has three contents with the names “b / 1”, “b / 2”, and “b / 3” (content “b”) as described above. / 1 ”, content“ b / 2 ”, content“ b / 3 ”).

  Then, in order for the client 100 to acquire the contents “a” and “b”, the three contents “a / 1”, “a / 2”, “a / 3”, and the contents “ It is assumed that a total of six interest packets requesting three contents “b / 1”, “b / 2”, and “b / 3” constituting “b” are transmitted.

  As a result, the entry of the reception PIT 204 of the router 102 becomes “i [a / 1, a / 2, a / 3, b / 1, b / 2, b / 3 (eth0)]”. Then, it is assumed that the router 102 transfers the data packet “D: a / 1” constituting the content “a / 1” by the face “eth0” among the three data packets constituting the content “a”. .

  In this case, the packet processing unit 504 of the router 102 includes the data packet (contents “a / 1”, “a”, and the contents “a” in the entry including the face “eth0” stored in the reception PIT 204. / 2 ”and“ a / 3 ”), only the entry“ i [a / 1, a / 2, a / 3 (eth0)] ”corresponding to the name 202“ a / 1 ”,“ “a / 2” and “a / 3” may be extracted, and the extracted name 202 may be included in the data packet “D: a / 1” as PIT information.

  In other words, the content “a” configured by the data packet “D: a / 1” to be transferred in the entry including the face “eth0” stored in the reception PIT 204 by the packet processing unit 504 of the router 102 is different from the content “a”. Entry “i [b / 1, b / 2, data packet constituting content“ b / 1 ”,“ b / 2 ”,“ b / 3 ”) corresponding to content“ b ” The names 202 “b / 1”, “b / 2”, and “b / 3” may not be extracted from “b / 3 (eth0)]”. Thereby, the size of the PIT information may be reduced.

  As described above, when the packet processing unit 504 transmits a plurality of data packets constituting one content, the entry corresponding to (related to) the data packet constituting the content recorded in the reception PIT 204 is recorded. Only the name 202 included may be included in the data packet to be transmitted as PIT information.

  Thereby, when a plurality of data packets constituting one content are continuously and intensively transmitted for a certain period of time, the data constituting the other content is recorded in the reception PIT 204. It may be avoided that the name 202 included in the entry corresponding to the packet is included in the data packet to be transmitted as PIT information. In this way, during the period, the size of the PIT information included in the data packet may be reduced, and the communication load on the network may be reduced.

  However, when this method is used, it is necessary that the device on the data packet receiving side can identify that the PIT information included in the data packet is reduced. Also, in step S702, the name 202 extracted from the entry of the transmission PIT 205 to be compared with the name indicated by the PIT information included in the received data packet needs to be reduced in the same manner as the PIT information. This may be realized as follows.

  For example, when the packet processing unit 504 includes the PIT information in the data packet, the identifier indicating the PIT information stored in the area 1010 of the packet illustrated in FIG. 10 includes the normal PIT information. What is necessary is just to make it differ by the time of including the PIT information reduced as mentioned above. Alternatively, information peculiar to the data packet stored in the area 1003 of the packet shown in FIG. 10 is differentiated between when the normal PIT information is included and when the reduced PIT information is included as described above. do it.

  In step S702, the PIT information comparison unit 505 may determine whether or not the PIT information included in the received data packet is reduced PIT information. If the PIT information comparison unit 505 determines that the PIT information is reduced, the data packet constituting the same content as the content constituted by the received data packet in the entry of its own transmission PIT 205. The name 202 may be extracted from only the entry corresponding to. Then, the extracted name 202 and the name indicated by the reduced PIT information may be compared to detect packet loss.

  For example, when the entry of the transmission PIT 205 of the router 102 is “o [a / 1, a / 2, a / 3, b / 1, b / 2, b / 3 (eth0)]”, the face “eth0” , The data packet “D: a / 1, PIT [a / 1, a / 2, a / 3]” is received. Further, it is assumed that the PIT information “PIT [a / 1, a / 2, a / 3]” included in the data packet is reduced PIT information.

  In this case, in step S702, the PIT information comparison unit 505 determines whether the PIT information included in the received data packet is reduced PIT information from the area 1010 or the area 1003 of the received data packet. May be determined.

  If it is determined that the PIT information has been reduced, the PIT information comparison unit 505 may grasp the content that the received data packet constitutes. In this specific example, it can be understood from the prefix of the content name “a / 1” that the received data packet constitutes that the data packet is a data packet that constitutes the content “a”.

  Then, the PIT information comparison unit 505 includes the entry “o [a / 1, a / 2, a / 3 (eth0)]” corresponding to the data packet constituting the content “a” among the entries of its own transmission PIT 205. Only the names 202 “a / 1”, “a / 2”, and “a / 3” may be extracted. The PIT information comparison unit 505 then extracts the extracted names 202 “a / 1”, “a / 2”, “a / 3” and the reduced PIT information “PIT [a / 1, a / 2, a / 3] ”may be compared with the names“ a / 1 ”,“ a / 2 ”, and“ a / 3 ”.

  In this case, when a plurality of data packets constituting one content are continuously and intensively received for a certain period, the reduced PIT information included in the received data packet during the period, Only the packet loss of the data packet constituting the content can be detected intensively by using the entry corresponding to the data packet constituting the content recorded in the transmission PIT 205.

  When the PIT information included in the data packet “D: a” is “PIT [a, b]”, the content name “a” indicated by the PIT information overlaps with the name of the content that the data packet constitutes. ing. Therefore, when the name of the content that the data packet constitutes is included in the name of the content indicated by the PIT information, the name of the content that constitutes the data packet is not included in the PIT information. Also good. That is, in the above example, the PIT information included in the data packet “D: a” may be “PIT [b]” instead of “PIT [a, b]”. In this way, the PIT information included in the data packet may be reduced.

  However, even when this method is used, the PIT information included in the data packet is reduced by the device on the data packet receiving side using the 1010 area or 1003 area of the data packet as described above. Need to be identified. In step S702, it is determined whether or not the PIT information is reduced. If it is determined that the PIT information is reduced, the data packet is configured with the name indicated by the PIT information included in the received data packet. You need to add the name of the content that you want.

  However, for example, when the PIT information included in the data packet “D: a” is “PIT [a]”, the PIT information included in the data packet “D: a” does not exist. As a result, in step S701, the PIT information comparison unit 505 determines that the PIT information is not included in the data packet, and step S702 is not performed. Therefore, in order to avoid this problem, in step S701, it is determined whether or not the PIT information is reduced, it is determined that the PIT information is reduced, and the PIT information is not included. In such a case, the process in step S702 may be performed.

  Furthermore, the name indicated by the PIT information included in the data packet is not limited to a character string such as “a” and “b”, and for example, a hash value of the name indicated by the PIT information may be calculated. However, in this case, in the apparatus that has received the data packet, in step S702, the hash value indicated by the PIT information included in the received data packet and the hash value of the name 202 extracted from the entry of its own transmission PIT 205 are obtained. It is necessary to compare.

  Further, the hash value indicated by the PIT information may be configured as a Bloom filter corresponding to the name 202 extracted from the entry of the transmission PIT 205. In this case, in the apparatus that has received the data packet including the PIT information, in step S702, the name 202 extracted from the entry of the own transmission PIT 205 using the Bloom filter indicated by the PIT information included in the received data packet is It may be possible to confirm only that it is not included in the name corresponding to the PIT information.

(Embodiment 3: Combination of PIT information and timeout)
In the first and second embodiments, the method using comparison of information extracted from the entry of the reception PIT 204 of the higher-level device and the entry of its own transmission PIT 205 in order to detect the packet loss has been described.

  However, the packet processing unit 504 further sets a timeout value for the entry of its own transmission PIT 205, and the time from when the entry is recorded in the transmission PIT 205 until reception of the data packet corresponding to the entry times out. When the value is exceeded, the packet loss of the data packet may be detected. When the packet processing unit 504 detects the packet loss, the packet transmission unit 503 may retransmit the interest packet corresponding to the data packet.

  Specifically, the transmission PIT 205 may be associated with each entry so that the date and time when the entry was recorded and the timeout value of the entry can be recorded. When the packet processing unit 504 records the entry in the transmission PIT 205 in step S807, the packet processing unit 504 may record the timeout value of the entry and the date and time when the entry was recorded in association with the entry. .

  The packet processing unit 504 periodically refers to the entry of the transmission PIT 205, and the time-out value associated with the entry from the date and time when the entry associated with the referenced entry is recorded. When the time exceeding the period has elapsed, the packet loss of the data packet corresponding to the entry may be detected, and the interest packet corresponding to the data packet may be retransmitted as in step S704.

  In this case, the packet processing unit 504 needs to be configured to update the date and time associated with the referenced entry to the date and time when the interest packet is retransmitted so that the interest packet is not retransmitted.

  With this configuration, for example, even when a packet loss of a data packet corresponding to the transmitted interest packet occurs in a burst manner and the data packet cannot be received for a certain period, an entry corresponding to the data packet is transmitted. The interest packet is retransmitted every time the elapsed time after recording in the PIT 205 exceeds the timeout value.

  Thereby, even if a packet loss of a data packet corresponding to the interested packet occurs, an appropriate retransmission process of the data packet can be performed. As a result, the data packet can be acquired without impairing reliability.

(Embodiment 4: Response to transmission delay)
FIG. 11 is a diagram illustrating an example of a situation in which an interest packet 1104 and a data packet 1105 are transmitted and received between two routers 102A and 102B in a CCN. FIG. 11 shows a situation in which the transmission of the interest packet 1104 “I: c” by the router A 102A and the transfer of the data packet 1105 “D: a: PIT [a, b]” by the router B 102B are performed almost simultaneously. ing.

  At this time, the entry of the transmission PIT 205 of the router A 102A is “o [a, b, c]”. On the other hand, since the interest packet 1104 has not arrived at the router B 102B when the router B 102B transmits the data packet 1105, the PIT information included in the data packet 1105 is “PIT [a, b]”. Therefore, when the router A 102A receives the data packet 1105, there is a possibility that the packet loss of the data packet corresponding to the interest packet 1104 is detected even though no packet loss actually occurs.

  In order to solve this problem, the PIT information comparison unit 505 does not use the name 202 extracted from the entry in which the elapsed time after recording is equal to or less than a predetermined threshold in the entry of the transmission PIT 205 in the comparison in step S702. You may do it.

  In order to realize this, the transmission PIT 205 may be associated with each entry so that the date and time when the entry was recorded can be recorded. Then, when the packet processing unit 504 records an entry in the transmission PIT 205 in step S807, the date and time when the entry is recorded may be recorded in association with the entry. The threshold may be set by, for example, measuring a time required when two adjacent routers 102 transmit and receive control signals, and reciprocate a data packet, and set the measured time.

  If comprised in this way, the effect shown below will be acquired. For example, after the upper router 102 transmits a data packet, the lower router 102 transmits an interest packet. Then, the lower router 102 receives the data packet before the elapsed time after recording the entry of the transmission PIT 205 indicating that the interest packet has been transmitted reaches a predetermined threshold, and step S702 is performed. Suppose that

  In this case, in step S702, the name 202 extracted from the entry of the transmission PIT 205 indicating that the interest packet has been transmitted is not compared with the name indicated by the PIT information. Thereby, it is possible to prevent erroneous detection of a packet loss of a data packet corresponding to the interest packet.

(Embodiment 5: Reduction in the number of times of transmission of PIT information)
In the above-described embodiment, each device always transmits the data packet including the PIT information when transmitting the data packet. However, for example, each device may count the number of transmitted data packets, and each time a predetermined number of data packets without PIT information are transmitted, a data packet including PIT information may be transmitted. . Alternatively, each device transmits a data packet not including PIT information until a predetermined time elapses after transmitting a data packet including PIT information, and includes the PIT information after the predetermined time elapses. A data packet may be transmitted.

  In addition, each device may transmit a data packet that includes only PIT information and does not include content data. Specifically, when a lower-level device transmits an interest packet for a data packet that includes only PIT information and does not include content data, the device that has received the interest packet extracts the entry from its own received PIT 204. A data packet including only the PIT information may be returned. Alternatively, each device voluntarily and periodically transmits a data packet including only the PIT information extracted from the entry of its own reception PIT 204 by the face included in the entry from which the PIT information is extracted. It may be.

  That is, by using the above method, the number of times of transmitting data packets including PIT information can be reduced, the size of data packets transmitted and received on the CCN network 101 can be reduced, and the communication load of the network can be reduced. Good.

  However, even when the above method is used, it is necessary for the apparatus on the data packet receiving side to be able to identify what information each data packet contains. For example, when each device that transmits a data packet transmits a data packet, information stored in the area 1010 or 1003 of the data packet shown in FIG. 10 is included, and only the PIT information is included in the data packet. This can be realized by differentiating between the case where only PIT information is included and the case where PIT information and content are included.

  The present disclosure can be used for a communication control device in a content-oriented network, and in particular, can be used for a router device that relays data, a client device that receives data, a publisher that provides data, and the like.

100 client (communication control device)
101 CCN network (content-oriented network)
102, 102A, 102B Router (communication control device)
103 CS
104, 107, 108 PIT
105 FIB
106 Publisher (communication control device)
204 Receive PIT
205 Send PIT
502 Packet receiver (receiver)
503 Packet transmitter (transmitter)
504 Packet processing unit 505 PIT information comparison unit (packet processing unit)
506 PIT information addition unit (packet processing unit)
510 Memory unit

Claims (10)

A communication control device connected to a content-oriented network,
A transmission unit for transmitting an interest packet;
A receiving unit for receiving a data packet corresponding to the packet of interest;
A memory unit that stores a transmission PIT in which unreceived information indicating a data packet that has not been received is recorded among data packets corresponding to an interest packet transmitted by the transmission unit;
If the data packet received by the receiving unit includes PIT information indicating a data packet that has not been transmitted, the data corresponding to the packet of interest using the PIT information and the transmission PIT And a packet processing unit that detects packet loss of the packet. The receiver further receives the interest packet from another device,
The memory unit further stores a received PIT in which untransmitted information indicating a data packet that has not been transmitted is recorded in a data packet corresponding to the interest packet received by the receiving unit,
The transmitter further transmits a data packet corresponding to the interest packet to the other device,
The communication control apparatus according to claim 1, wherein the packet processing unit includes the untransmitted information recorded in the reception PIT as the PIT information in a data packet transmitted by the transmission unit. The untransmitted information includes face information indicating a communication interface that has received an interest packet corresponding to the data packet indicated by the untransmitted information,
The packet processing unit extracts face information included in the untransmitted information indicating a data packet transmitted by the transmitting unit from the received PIT, and the extracted face information recorded in the received PIT Including the untransmitted information including the same face information as the PIT information in the data packet,
The communication control apparatus according to claim 2, wherein the transmission unit causes the data packet to be transmitted by a communication interface indicated by the extracted face information. The packet processing unit deletes the PIT information included in the data packet received by the receiving unit, and then includes the untransmitted information recorded in the received PIT as the PIT information in the data packet The communication control apparatus according to claim 2. The packet processing unit compares the PIT information included in the data packet received by the receiving unit with the unreceived information recorded in the transmission PIT, and compares the data packet indicated by the PIT information with the unreceived data packet. When the data packet indicated by the received information does not match, the packet loss of the data packet that does not match is detected, and the interested packet corresponding to the data packet that does not match is retransmitted to the transmitter. The communication control apparatus according to claim 1. The packet processing unit does not use, in the comparison, the unreceived information whose elapsed time since recording is less than a predetermined threshold among the unreceived information recorded in the transmission PIT. The communication control device according to claim 5. The packet processing unit further sets a timeout value for the unreceived information recorded in the transmission PIT, and the data indicated by the unreceived information after the unreceived information is recorded in the transmission PIT. When the time until the reception unit receives the packet exceeds the timeout value, the packet loss of the data packet is detected, and further the interest packet corresponding to the data packet is retransmitted to the transmission unit. The communication control device according to claim 5, wherein: The packet processing unit, when the transmission unit transmits a plurality of data packets constituting one content, respectively, the untransmitted information indicating the data packet constituting the one content recorded in the reception PIT The communication control apparatus according to claim 2, wherein only the PIT information is included in a data packet transmitted by the transmission unit. The packet processing unit is recorded in the PIT information included in the received data packet and the transmission PIT when the receiving unit receives each of a plurality of data packets constituting one content. The communication control apparatus according to claim 1, wherein the detection is performed using the unreceived information indicating a data packet constituting the one content. A communication control method in a communication control device connected to a content-oriented network,
The communication control device includes a memory unit that stores a transmission PIT in which unreceived information indicating a data packet that has not been received is recorded among data packets corresponding to the transmitted interest packet,
Sending the packet of interest;
Receiving a data packet corresponding to the packet of interest;
When the received data packet includes PIT information indicating a data packet whose transmission has not been completed, the packet of the data packet corresponding to the interested packet using the PIT information and the transmission PIT A communication control method characterized by detecting a loss.

Images