JP2013038660A - Middle layer module, recording media and data transfer method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a protocol module and a data transfer method, etc. which helps to spread an NDN architecture on a full scale as a candidate for the next generation Internet without modifying the functionality of the Ethernet (R).SOLUTION: A sub-layer 21 is provided between an NDN layer 22 and Ethernet (R) 11, etc., the sub-layer 21 being used to execute a segmentation/reassembling process. A network interface (Face module) constituting the NDN layer 22 conducts a data transfer between adjacent NDN nodes, and also holds an MTU in a link. Therefore, the Face module may be considered to be the one which finally invokes a data transmission instruction (to the Ethernet (R) 11, etc., for example). It is appropriate that the segmentation/reassembling process be executed immediately before this data transmission instruction. The NDN Face module is extended to implement a data transfer method and a middle layer module of the present invention.

Description

  The present invention provides a communication protocol including a middle layer defined on a network interface layer or the like in the TCP / IP model, and a named data networking (NDN) layer defined on the middle layer. The present invention relates to a module (program) of the middle layer in the protocol stack to be implemented.

  Conventionally, an Open Systems Interconnection (OSI) reference model established by the International Organization for Standardization (ISO) is known. FIG. 8A shows seven layers (layers) based on the OSI reference model. In FIG. 8A, reference numeral 1 is a physical layer of the first layer that defines physical connection (number of pins and shape of connector) between networks, and 2 is data between communication devices that are directly connected. The data link layer of the second layer responsible for transmission, 3 the network layer of the third layer responsible for network route selection / relay work, 4 the end-to-end of the network such as retransmission, congestion, etc. 4) Transport layer for performing communication management of the end) 5 is a session layer of 5th layer that defines a procedure from the start to the end of communication between communication programs, and 6 is a layer that defines a data expression method and the like. Six presentation layers, and seven an application layer for providing specific communication services such as file transfer.

  The OSI reference model described above has not been widely used for reasons such as implementation. On the other hand, TCP / IP (Transmission Control Protocol / Internet Protocol) developed from ARPAnet (Advanced Research Projects Agency Network) developed under the US Department of Defense developed rapidly with the spread of the Internet. FIG. 8B shows a TCP / IP hierarchical model. The layer of TCP / IP shown in FIG. 8B and the layer of the OSI reference model shown in FIG. 8A are associated in the horizontal direction, but which layer of both is associated with which layer. There are different points depending on the person, and the correspondence between FIGS. 8A and 8B is an example (see Non-Patent Document 1, for example). In FIG. 8B, reference numeral 10 denotes a hardware layer of the 0th layer corresponding to the physical layer 1, and an implementation example includes a UTP (Unshield Twisted Pair) cable. Reference numeral 11 denotes a first network interface layer corresponding to the data link layer 2, and examples of implementation include Ethernet (registered trademark), PPP (Point to Point Protocol), etc. (hereinafter referred to as "Ethernet (registered trademark) 11". "" Reference numeral 12 denotes the Internet of the second layer corresponding to the network layer 3, and examples of the implementation include IP (hereinafter sometimes referred to as “IP12”). Reference numeral 13 denotes a transport in the third layer corresponding to the transport layer 4, and an example of mounting is TCP or the like (hereinafter also referred to as “TCP13”). Reference numeral 14 denotes a fourth layer application corresponding to the session layer 5, the presentation layer 6, and the application layer 7, and examples of the implementation include HTTP (Hyper Text Transfer Protocol). In TCP / IP, an IP address is assigned to a node, and a packet is transferred by looking at the IP address in IP12.

  In each of the seven layers based on the OSI reference model shown in FIG. 8A and the TCP / IP layer shown in FIG. 8B, the size of data that can be easily handled in each layer is different. For example, the maximum transmission size (Maximum Transmission Unit: MTU) that can be handled by the Ethernet (registered trademark) 11 is 1 KB or 1.5 KB. For this reason, the upper layer IP 12 has a segmentation function for dividing data into 1 KB or 1.5 KB and passing it to the lower layer Ethernet (registered trademark) 11. Conversely, the IP 12 also has a reassembly function that assembles a plurality of packets received from the lower layer Ethernet (registered trademark) 11 and passes them to the upper layer TCP 13. The Ethernet 11 does not have a segmentation / reassembly function.

  Next, Named Data Networking (NDN) will be described. NDN is said to be a candidate for the next-generation Internet, and is a model in which transmission / reception is controlled by assigning a name to content held by a node existing in the network. In FIG. 8B, this is a model in which TCP 13 and IP 12 are replaced with an NDN layer. In TCP / IP, IP addresses are assigned to nodes and packets are transferred by looking at the IP address in IP12. However, in NDN, names are given to contents existing in nodes in the network, and the names are seen in the NDN layer. And transfer the content.

  FIG. 9 shows an NDN packet in the NDN architecture (see FIG. 9 and the source of FIG. 10 to be described later refer to Non-Patent Document 2. Hereinafter, description will be given with reference to Non-Patent Document 2). Communication in NDN is driven by the receiving end, the data consumer. FIG. 9A shows an interest packet 100. Interest packet 100 is a packet sent by a data consumer to receive data and carries a name identifying the desired data. As shown in FIG. 9A, the parameters of the interest packet 100 are a content name 101 that identifies data that the data consumer wants to receive and a selector 102 that identifies the issuer. When the interest packet 100 arrives at the node having the requested data, a data packet is sent back. FIG. 9B shows the data packet 150. As shown in FIG. 9B, the parameters of the data packet 150 are a content name 101, a signature 151 with a data producer's key, information 152 related thereto, and data 153.

  FIG. 10 shows a transfer process in the NDN node 200. 10, reference numerals 210, 211, and 212 are examples of the external communication environment of the NDN node 200, 204 is an interface (Face 0) to the external communication environment 210, 205 is an interface (Face 1) to the external communication environment 211, 206 Is an interface (Face 2) to the external communication environment 212. As shown in FIG. 10, a pointer to the content store 201 is recorded in the ptr column of the index 207, and “C” is recorded in the corresponding type column. Similarly, a pointer to a pending interest table (PIT) 202 is recorded in the other ptr column of the index 207, and “P” is recorded in the corresponding type column. Similarly, a pointer to a forwarding information base (FIB) 203 is recorded in another ptr column of the index 207, and “F” is recorded in the corresponding type column. For example, it is assumed that the data consumer transmits an interest packet 100 requesting /parc/videos/WidgetA.mpg. The router of the NDN node 200 records the interface 204 or the like that received the interest packet 100 (“0” for Face 0, “1” for Face 1) in the Face List column of the FIB 203, and the transfer is the name (Prefix) in the FIB 203. ). Data packet 150 is returned to the data consumer by following the reverse of the path created by interest packet 100. Neither the interest packet 100 nor the data packet 150 carries an address such as an IP address. The interest packet 100 is routed towards the data creator based on the name carried by the interest packet 100. The data packet 150 is returned based on the information set at each router hop by the interest packet 100. That is, the data packet 150 is returned following the same path as the path through which the interest packet 100 has passed. One data packet 150 fills one interest packet 100 while crossing hops (in the path) (hop-by-hop principle).

  The router of the NDN node 200 holds the interest packet 100 and the data packet 150 in the PIT 202 for a certain period. If multiple interest packets 100 for the same data are received from the downstream, only the first interest packet 100 is sent upstream toward the data source. The router records the interest packet 100 in the PIT 202. The PIT 202 records the content name of the interest packet 100 in the prefix field and the set of interfaces that have received the plurality of interest packets 100 in the Requesting Face (s) field. When the data packet 150 arrives, the router finds the corresponding PIT 202 entry and forwards the data packet 150 to all the interfaces recorded in the entry. Thereafter, the router deletes the entry of the corresponding PIT 202 and caches the data packet 150 in the content store 201. The name of the data packet 150 is recorded in the Name column of the content store 201, and the contents of the data 153 are recorded in the Data column. The content store 201 basically uses the buffer storage of the router. By using the content store 201, NDN can automatically support content distribution when many users, such as Youtube®, request the same data simultaneously.

  As described above, the NDN architecture is a model in which the TCP 13 and the IP 12 in FIG. 8B are replaced with an NDN layer. However, since it is difficult to replace NDN with TCP13 and IP12 suddenly, NDN currently operates on TCP13 and IP12. FIG. 11 shows the current NDN implementation. In FIG. 11, portions denoted by the same reference numerals as those in FIG. As shown in FIG. 11, the current NDN layer 22 uses TCP 13 and IP 12 as a pseudo third layer or second layer, and uses the segmentation / reassembly function of TCP 13 or IP 12 to perform segmentation / reassembly. Processing. That is, the current NDN layer does not have the segmentation function and the reassembly function that IP12 and the like have.

  As described above, Ethernet (registered trademark) 11 has a limitation of MTU, but it does not have a segmentation / reassembly function. However, since Ethernet (registered trademark) 11 is widely used worldwide, there is a problem that it is not a good idea to modify the function of Ethernet (registered trademark) 11. As described above, the NDN architecture is a candidate for the next-generation Internet, and it is necessary to directly mount the NDN layer on the Ethernet (registered trademark) 11 in order to spread in earnest. However, there is a problem that the current NDN layer does not have the segmentation function and reassembly function that IP12 and the like have. Accordingly, an object of the present invention is to solve the above-mentioned problem, and it is necessary to fully disseminate the NDN architecture as a candidate for the next generation Internet without modifying the function of the Ethernet (registered trademark) 11. It is an object to provide a protocol module and a data transfer method that can be made possible.

  The middle layer module of the present invention includes a lower layer corresponding to the network interface layer in the TCP / IP model or the data link layer in the OSI reference model, a middle layer defined on the lower layer, and a named layer defined on the middle layer. -The middle layer in a protocol stack that implements a communication protocol comprising a data networking (NDN) layer that controls the transmission and reception by naming the contents of nodes existing in the network. And when the size of the input byte sequence passed from the NDN layer is larger than the maximum transfer size of the lower layer, the input byte sequence is a plurality of byte sequences less than or equal to the maximum transfer size. The order that shows the order in each byte sequence A segmentation means that adds control data including information and subsequent information indicating the presence / absence of the subsequent byte sequence and sequentially passes to the lower layer, and the subsequent byte sequence is determined by the subsequent information in the control data included in the byte sequence passed from the lower layer. If it is determined that there is, record the data byte sequence excluding the control data from the byte sequence passed from the lower layer in the order indicated by the sequence information in the control data, wait for the next byte sequence, and follow by the subsequent information When it is determined that there is no byte sequence, reassembly is performed by adding the data byte sequence excluding control data from the byte sequence passed from the lower layer to the recorded data byte sequence and transferring the collected byte sequence to the NDN layer This is a middle-layer module for functioning as a storage means.

  Here, in the middle layer module of the present invention, the segmentation means is the first byte sequence in the input byte sequence when the size of the input byte sequence passed from the NDN layer is equal to or smaller than the maximum transfer size of the lower layer. Further comprising means for attaching control data including order information indicating that there is no subsequent byte sequence to the lower layer and passing the control data to the lower layer, wherein the reassembly means includes a byte sequence passed from the lower layer. If it is determined by the order information in the control data included in the first byte sequence and the subsequent information that there is no subsequent byte sequence, the data byte sequence excluding the control data from the byte sequence passed from the lower layer is A means for passing to the NDN layer may be further provided.

  Here, in the middle layer module of the present invention, the control data attached to the byte sequence includes a bit indicating whether the subsequent byte sequence is present and the total size of the byte sequence passed to the lower layer to a predetermined value. A bit indicating the byte length and a bit indicating the order of the byte string can be included.

  The recording medium of the present invention is a computer-readable recording medium on which any of the middle layer modules of the present invention is recorded.

  The data transfer method of the present invention includes a lower layer corresponding to the network interface layer in the TCP / IP model or the data link layer in the OSI reference model, a middle layer defined on the lower layer, and a named layer defined on the middle layer. -The middle layer in a protocol stack that implements a communication protocol comprising a data networking (NDN) layer that controls the transmission and reception by naming the contents of nodes existing in the network. In the data transfer method executed by the module (program), when the size of the input byte sequence passed from the NDN layer is larger than the maximum transfer size of the lower layer, the input byte sequence is a plurality of data smaller than the maximum transfer size. The byte sequence is divided into each byte sequence. A segmentation step to which the control data including the order information and the subsequent information indicating the presence or absence of the subsequent byte sequence is attached and sequentially passed to the lower layer, and the subsequent information in the control data included in the byte sequence passed from the lower layer If it is determined that there is a byte string, record the data byte string excluding the control data from the byte string passed from the lower layer in the order indicated by the order information in the control data, wait for the next byte string, and follow If it is determined by the information that there is no subsequent byte sequence, the data byte sequence excluding control data from the byte sequence passed from the lower layer is added to the recorded data byte sequence, and the combined byte sequence is sent to the NDN layer. And a reassembling step for passing.

  Here, in the data transfer method of the present invention, the segmentation step is the first byte sequence in the input byte sequence when the size of the input byte sequence passed from the NDN layer is equal to or smaller than the maximum transfer size of the lower layer. Further including a step of passing control data including order information indicating that there is no subsequent byte sequence to the lower layer and passing the control data to the lower layer, wherein the reassembly step includes a byte sequence passed from the lower layer If it is determined by the order information in the control data included in the first byte sequence and the subsequent information that there is no subsequent byte sequence, the data byte sequence excluding the control data from the byte sequence passed from the lower layer is The method may further comprise passing to the NDN layer.

  Here, in the data transfer method according to the present invention, the control data attached to the byte string includes a padding for aligning a bit indicating the presence / absence of the subsequent byte string and a total size of the byte string to be passed to the lower layer to a predetermined value. A bit indicating the byte length and a bit indicating the order of the byte string can be included.

  According to the middle layer module and the data transfer method of the present invention, a sublayer is provided between the NDN layer and Ethernet (registered trademark), and the sublayer performs segmentation / reassembly processing. As a result, it is possible to provide a protocol module, a data transfer method, and the like that can fully spread the NDN architecture as a candidate for the next-generation Internet without modifying the Ethernet (registered trademark) function. effective. In addition, the upper layer application can use an NDN packet having an optimal size without being aware of the lower layer MTU. Since data transfer can be performed in accordance with the MTU which is a characteristic of the link, there is an effect that the affinity with the hop-by-hop transfer principle in NDN can be increased.

It is a figure for demonstrating the process of the data transfer method of this invention for mounting an NDN layer on Ethernet (trademark) 11. FIG. It is a figure which shows the structure of control data. It is a figure which shows a series of n byte sequence 32-1 -32-n which the sublayer 21t passes to Ethernet (trademark) 11t. FIG. 3 is a diagram showing the structure of a frame 34. FIG. 4 is a diagram illustrating a bi-string 30 when the size of an input byte string 30 is equal to or smaller than an MTU of Ethernet (registered trademark) 11t. It is a figure which shows the operating environment and functional block of the module (program) of the sublayers 21t and 21r of this invention. It is a block diagram which shows the internal circuit 80 of the computer 50, such as the NDN node 20t which performs the data transfer method and middle layer module of this invention. It is a figure which shows the hierarchy of an OSI reference model, and the hierarchy of TCP / IP. It is a figure which shows the NDN packet in a NDN architecture. FIG. 4 is a diagram showing a transfer process in an NDN node 200. It is a figure which shows the present NDN mounting.

  Hereinafter, embodiments will be described in detail with reference to the drawings. The codes of each layer such as the OSI reference model and the TCP / IP model shown in FIGS. 7 and 10 described in the background art are appropriately used.

  The data transfer method of the present invention is defined on the Ethernet (registered trademark) 11 (lower layer) corresponding to the network interface layer 11 in the TCP / IP model or the data link layer 2 in the OSI reference model, and the Ethernet (registered trademark) 11. A data transfer method executed by a sub-layer module (program) in a protocol stack for implementing a communication protocol including a sub-layer (middle layer) and an NDN layer 22 defined on the sub-layer is there. The protocol stack is software that implements the network protocol as described above. Since the network protocol is hierarchically defined as the network protocol is hierarchically defined, the “stack” Is used. In recent computers, since the OS itself has a network function, the protocol stack is usually implemented in the kernel. The data transfer method (middle layer module) of the present invention can also be installed inside the OS of an NDN node (computer such as router or PC).

  FIG. 1 is a diagram for explaining processing of a data transfer method according to the present invention for mounting an NDN layer on an Ethernet (registered trademark) 11. In FIG. 1, the NDN node that transmits the packet of the NDN architecture is distinguished by attaching “t” and “r” to the sign (numerical part), such as 20t for the NDN node that transmits the packet and 20r for the receiving side. The reference numerals (numeric parts) have the same meanings as in FIGS. 7 and 10 except that the reference numerals assigned to the layers of the NDN nodes 20t and 20r are similarly distinguished by attaching “t” and “r”. Therefore, the description is omitted.

  With reference to FIG. 1, processing in the case of transmitting NDN packets 30 such as interest packet 100 and data packet 150 will be described. In the application layer 14t, for example, a situation can be assumed in which a user who is a data consumer requests a site such as Youtube (registered trademark) to distribute a moving image.

1. The NDN layer 22t creates the NDN packet 30. Since the NDN layer 22t knows the structure of the NDN packet 30 (see FIG. 9), it knows in which position (byte offset from the head) each parameter 101 and the like should be placed.

2. The NDN layer 22t passes the NDN packet 30 to the sub-layer (middle layer) as a group of byte strings (step S10).

3. When the size of the input byte sequence 30 passed from the NDN layer is larger than the MTU of the Ethernet (registered trademark) 11t (lower layer), the sub layer 21t divides the input byte sequence into a plurality of byte sequences having a size equal to or smaller than the MTU. Then, each byte sequence is sequentially passed to the Ethernet (registered trademark) 11 as a byte sequence 32 to which control data including order information indicating the order and subsequent information indicating the presence or absence of the subsequent byte sequence is attached (step S20, segmentation step). FIG. 2 shows the structure of the control data. As shown in FIG. 2A, the byte string 32 is composed of a byte string 32d (transfer data), control data 32c, and padding bytes 32p, which will be described later. As shown in FIG. 2B, the control data 32c includes more bits (following information) indicating the presence / absence of a succeeding byte sequence and a total size of the byte sequence 32 passed to the Ethernet (registered trademark) 11 with a predetermined value (for example, Pad bits indicating the length of the padding bytes 32p when being aligned to a multiple of 32 bits) and seq bits (order information) indicating the order of the byte string 32 divided as described above. As shown in FIG. 2B, the size of each bit is 1 bit for more bits (when “1”, there is a subsequent byte sequence, and when “0”, there is no subsequent byte sequence), pad It is preferable that the bit is 2 bits, the seq bit is 5 bits, and the control data 32c is 1 byte. In recent data communication, in order to facilitate hardware processing, the data size is generally aligned to a multiple of 32 bits (4 bytes). In the data transfer method and the like of the present invention, when adding 1-byte control data 32c to the end of transfer data 32d, the meaningless data ( Padding bytes 32p) are embedded in between. As a result, the total length of the transfer data 32d + padding bytes 32p + control data 32c is an integral multiple of 4 bytes. Depending on the size of the original transfer data 32d, since the length of the padding byte 32p is changed from 0 byte to 3 bytes, the pad bit length may be 2 bits. More generally, when the length of the byte string 32 is made a multiple of n bytes, the length of the padding bytes 32p is 0 to n-1 bytes, so the length of the pad bits may be log ₂ n bits. . For example, when n = 8 bytes, the length of pad bits is 3 bits. In this case, in order to store the control data 32c in 1 byte, the seq bit may be 4 bits, or the control data 32c may be 2 bytes. FIG. 3 shows a series of divided n byte strings 32-1 to 32-n that the sublayer 21t passes to the Ethernet (registered trademark) 11t. As shown in FIG. 3, since the byte string 32-1 is the first byte string, the more bit = 1 and the seq bit = 1 of the control data 32c-1, and the byte string 32-2 is the second byte string 32-2. Therefore, the more bit = 1 of the control data 32c-2 and the seq bit = 2. Since the byte string 32-n is the last byte string, the more bit = 0 and the seq bit = n of the control data 32c-n.

4). The Ethernet (registered trademark) 11t creates a frame 34 from the byte string 32 and sequentially passes it to the hardware layer 10t (step S30). FIG. 4 shows the structure of the frame 34. As shown in FIG. 4, the frame 34 itself has a normal MAC frame structure, a preamble that is a bit pattern for synchronization when receiving the MAC frame, a destination address for identifying a transmission partner, a transmission source A source address for identifying the user data, a type indicating the length of user data, and a byte string 32 (not shown, but control data 32c and necessary padding bytes as shown in FIG. 2A). 32p) and a frame check sequence (FCS) in which a checksum result is recorded.

5). The hardware layer 10t transmits the frame 34 to the NDN node 20r side (step S40).

6). The hardware layer 10r receives the frame 34 and passes it to the Ethernet (registered trademark) 11r (step S50).

7). The Ethernet (registered trademark) 11r extracts the user data 32 from the received frame 34 and sequentially passes it to the sublayer 21r (step S60).

8). When the sub layer 21r determines that there is a subsequent byte sequence based on the more bit in the control data 32c included in the byte sequence 32 passed from the Ethernet (registered trademark) 11r, the sub layer 21r indicates the seq bit in the control data 32. Data byte sequence 32d excluding the control data 32c from the byte sequence 32 passed from the Ethernet (registered trademark) 11r in this order (if the pad bit is not 0, the data byte excluding the padding byte 32p of the length indicated by the pad bit) Record column 32d (same below) in memory and wait for the next byte sequence. If it is determined by the more bit of the next byte string 32 that there is no subsequent byte string, the control data is transferred from the byte string 32 last passed from the Ethernet (registered trademark) 11r to the data byte string 32d recorded in the memory. The data byte sequence 32d excluding 32c and the like is added and the combined byte sequence 30 is passed to the NDN layer 22r (step S70, reassembly step). The byte sequence 30 passed to the NDN layer 22r is one NDN packet. The sublayer 21r is not conscious of the NDN packet 30, but is handled as a simple byte string 30.

9. Since the NDN layer 22r can interpret the structure of the NDN packet 30, the NDN layer 22r extracts necessary parameters 101 and processes them appropriately.

  When the size of the input byte sequence 30 passed from the NDN layer 22t in the above-described process 3 is equal to or smaller than the MTU of the Ethernet (registered trademark) 11t, the sub layer 21t determines that the input byte sequence 30 is the first byte sequence. It passes to the Ethernet (registered trademark) 11t as the byte sequence 30 with the control data 32c including the seq bit to indicate and the more bit (= “0”) indicating that there is no subsequent byte sequence (the padding byte has been described above) Since it is the same as the contents, it is omitted.) FIG. 5 shows the byte string 30 when the size of the input byte string 30 is equal to or smaller than the MTU of the Ethernet (registered trademark) 11t. As shown in FIG. 5, the control data 32 c has more bits = 0 and seq bits = 1, and the transfer data is an input byte string 30. Subsequent processing of each layer is the same as described above. When the sub-layer 21r determines that the first byte string is based on the seq bit in the control data 32c included in the byte string 32 passed from the Ethernet (registered trademark) 11r and there is no subsequent byte string based on the more bit, A data byte sequence 30 excluding control data 32c and the like from the byte sequence 32 delivered from the Ethernet (registered trademark) 11r is delivered to the NDN layer 22r.

  When the NDN packet 30 is the interest packet 100, it is considered that the size is equal to or smaller than the MTU of the Ethernet (registered trademark) 11t, and therefore the sublayer 11t processes as a byte string 32 as shown in FIG. On the other hand, when the NDN packet 30 is the data packet 150, the size is large (for example, about 1 GB for a moving image), so the sublayer 11t processes the byte sequences 32-1 to 32-n as shown in FIG. .

  FIG. 6 shows the operating environment and functional blocks of the modules (programs) of the sub-layers 21t and 21r of the present invention. In FIG. 6, reference numeral 50 denotes a computer on which the modules of the sub-layers 21t and 21r of the invention are mounted, and 70 denotes a functional block of the module. When the size of the input byte sequence 30 passed from the NDN layer 22t is larger than the MTU of the Ethernet (registered trademark) 11t, the segmentation unit 71 (segmentation means) shown in FIG. The control data 32c including the seq bit indicating the order and the more bit indicating the presence / absence of the succeeding byte string is attached to each byte string, and sequentially passed to the Ethernet (registered trademark) 11t. When the size of the input byte string 30 passed from the NDN layer 22t is equal to or less than the MTU of the Ethernet (registered trademark) 11t, the segmentation unit 71 includes a seq bit indicating the first byte string in the input byte string 30 and a subsequent byte string. The control data 32c including the more bit indicating that there is no is attached to the Ethernet (registered trademark) 11t.

  The reassembly unit 72 (reassembly means) shown in FIG. 6 determines that there is a subsequent byte sequence based on the more bit in the control data 32c included in the byte sequence 32 passed from the Ethernet (registered trademark) 11r. In this case, the data byte string 32d excluding the control data 32c and the like from the byte string 32 passed from the Ethernet (registered trademark) 11r in the order indicated by the seq bits in the control data 32c is stored in the data byte string recording DB 60 of the computer 50. Record and wait for the next byte sequence. When it is determined by the more bit that there is no subsequent byte sequence, the reassembly unit 72 adds the data byte sequence recorded in the data byte sequence record DB 60 from the byte sequence 32 passed from the Ethernet (registered trademark) 11r. The data byte string 32d excluding the control data 32c is added and the combined byte string 30 is passed to the NDN layer 22r. The reassembling unit 72 determines that the first byte string is based on the seq bit in the control data 32c included in the byte string 32 passed from the Ethernet (registered trademark) 11r, and that the subsequent byte string is absent according to the more bit. In this case, the data byte sequence 32d excluding the control data 2 and the like from the byte sequence 32 delivered from the Ethernet (registered trademark) 11r is delivered to the NDN layer 22r. The data byte string record DB 60 is a reception buffer, and can be continuously reassembled by setting it to an appropriate size.

  The NDN layers 22t and 22r described above are composed of the network interface (Face 0, etc.), the index 207, the content store 201, the PIT 202, and the FIB 203 described in the background art. According to Non-Patent Document 2, a network interface (Face module) manages data transfer between adjacent NDN nodes, and also holds an MTU in a link (for example, Ethernet (registered trademark) 11t). Therefore, it is considered that the Face module finally calls a data transmission command (for example, to Ethernet (registered trademark) 11 or the like). Since it is appropriate to perform segmentation / reassembly processing immediately before this data transmission instruction, the data transfer method and middle layer module of the present invention can be easily implemented by extending the NDN Face module. .

  As described above, according to the first embodiment of the present invention, the sub-layer 21 is provided between the NDN layer 22 and the Ethernet (registered trademark) 11 and the like, and the sub-layer 21 performs the segmentation / reassembly process. As a result, it is possible to provide a protocol module, a data transfer method, and the like that can fully spread the NDN architecture as a candidate for the next-generation Internet without modifying the function of the Ethernet (registered trademark) 11. . In addition, the upper layer application can use an NDN packet having an optimal size without being aware of the lower layer MTU. Since data transfer can be performed in accordance with the MTU which is the characteristic of the link, the affinity with the hop-by-hop transfer principle in NDN is high.

  FIG. 7 is a block diagram showing an internal circuit 80 of the computer 50 such as the NDN node 20t for executing the data transfer method and the middle layer module of the present invention. As shown in FIG. 7, the CPU 81, ROM 82, RAM 83, image control unit 86, controller 87, input control unit 90, and external I / F unit 92 are connected to a bus 93. In FIG. 7, the above-described middle-layer module of the present invention is recorded on a recording medium (including a removable recording medium) such as a ROM 82, a disk 88, a DVD, or a CD-ROM 89. The disk 88 can store the data byte string recording DB 60 described above. The middle layer module of the present invention is loaded into the RAM 83 from the ROM 82 via the bus 93 or from the recording medium such as the disk 88 or DVD or CD-ROM 89 via the controller 87 via the bus 93. The image control unit 86 sends image data of various screens to the VRAM 85. A display device (display) 84 displays the above data sent from the VRAM 85. The VRAM 85 is an image memory having a capacity corresponding to the data capacity of one screen of the display device 84. The input operation unit 91 is an input device such as a mouse or a keyboard for inputting to the computer 50, and the input control unit 90 is connected to the input operation unit 91 and performs input control or the like. The external I / F unit 92 has an interface function when connecting to the NDN node 20r or the like outside the computer 50.

  As described above, the computer (CPU) 81 executes the middle layer module of the present invention, thereby achieving the object of the present invention. The middle layer module can be supplied to the computer (CPU) 81 in the form of a recording medium such as a DVD or a CD-ROM 89 as described above, and a recording medium such as a DVD or a CD-ROM 89 on which the middle layer module is recorded is also the same. This constitutes the present invention. In addition to the recording medium described above, for example, a memory card, a memory stick, an optical disk, or the like can be used as the recording medium on which the middle layer module is recorded.

  As an application example of the present invention, the present invention can be applied to a segmentation / reassembly process between the Ethernet (registered trademark) 11 and the NDN layer 22.

  1 physical layer, 2 data link layer, 3 network layer, 4 transport layer, 5 session layer, 6 presentation layer, 7 application layer, 10, 10t, 10r hardware, 11, 11t, 11r network interface, 12 internet, 13 Transport, 14 applications, 20t, 20r, 200 NDN nodes, 21t, 21r sub-layer, 22, 22t, 22r NDN layer, 30 NDN packets, 32, 32-1, 32-2, 32-n, 32d, 32d- 1, 32d-2, 32d-n byte sequence, 32c, 32c-1, 32c-2, 32c-n control data, 32p padding bytes, 34 frames, 50 computers, 60 data byte sequence record DB, 7 0 functional block, 71 segmentation unit, 72 reassembly unit, 81 CPU, 82 ROM, 83 RAM, 84 display device, 85 VRAM, 86 image control unit, 87 controller, 88 disk, 89 CD-ROM, 80 input control unit 91 Input operation unit, 92 External I / F unit, 93 Bus, 100 Interest packet, 101, 102, 151, 152 Parameter, 150 Data packet, 153 Data, 201 Content store, 202 PIT, 203 FIB, 204, 205 , 206 Face, 207 Index, 210, 211, 212 External communication environment.

"TCP / IP and OSI reference model", [online], 68user's page, [Search June 17, 2011], Internet, <URL: http://x68000.qed.net/~68user/net/osi7. html>. Lixia Zhang, Deborah Estrin, Jeffrey Burke, etal., “Named Data Networking (NDN) Project”, NDN-0001, October 31, 2010,, [online], [Search June 17, 2011], Internet, < URL: http://www.named-data.net/ndn-proj.pdf>.

Claims (7)

A lower layer corresponding to the network interface layer in the TCP / IP model or the data link layer in the OSI reference model, a middle layer defined on the lower layer, and named data networking defined on the middle layer (Named Data Networking) NDN) module (program) in a protocol stack that implements a communication protocol that includes a layer that controls the transmission and reception by naming content held by nodes existing in the network. Computer
When the size of the input byte sequence passed from the NDN layer is larger than the maximum transfer size of the lower layer, the input byte sequence is divided into a plurality of byte sequences equal to or less than the maximum transfer size, and order information indicating the order of each byte sequence And segmentation means for passing control data including subsequent information indicating the presence or absence of the subsequent byte sequence and sequentially passing to the lower layer,
When it is determined that there is a subsequent byte sequence based on the subsequent information in the control data included in the byte sequence passed from the lower layer, from the byte sequence passed from the lower layer in the order indicated by the order information in the control data Record the data byte string excluding the control data, wait for the next byte string, and if it is determined by the subsequent information that there is no subsequent byte string, control is performed from the byte string passed from the lower layer to the recorded data byte string A middle-layer module for functioning as a reassembling means for transferring a byte sequence that is a collection of data byte sequences excluding data to the NDN layer. The middle layer module according to claim 1,
When the size of the input byte sequence passed from the NDN layer is equal to or less than the maximum transfer size of the lower layer, the segmentation means has no order information and subsequent byte sequence indicating that the input byte sequence is the first byte sequence Further comprising means for attaching control data including subsequent information indicating that the data is to be passed to the lower layer,
When it is determined that the reassembling means is the first byte string based on the order information in the control data included in the byte string passed from the lower layer and the subsequent byte string is absent according to the subsequent information, from the lower layer An intermediate layer module, further comprising means for passing a data byte sequence excluding control data from the transferred byte sequence to the NDN layer.   3. The middle layer module according to claim 1 or 2, wherein the control data attached to the byte string is used to align a bit indicating the presence / absence of a subsequent byte string and a total size of the byte string to be passed to the lower layer to a predetermined value. A middle layer module comprising a bit indicating a padding byte length and a bit indicating the order of the byte string.   A computer-readable recording medium on which the middle layer module according to any one of claims 1 to 3 is recorded. A lower layer corresponding to the network interface layer in the TCP / IP model or the data link layer in the OSI reference model, a middle layer defined on the lower layer, and named data networking defined on the middle layer (Named Data Networking) : NDN) module (program) in a protocol stack that implements a communication protocol with a name that controls the transmission / reception by naming the contents of nodes existing in the network A data transfer method comprising:
When the size of the input byte sequence passed from the NDN layer is larger than the maximum transfer size of the lower layer, the input byte sequence is divided into a plurality of byte sequences equal to or less than the maximum transfer size, and order information indicating the order of each byte sequence And a segmentation step that passes control data including the subsequent information indicating the presence or absence of the subsequent byte sequence and sequentially passes to the lower layer,
When it is determined that there is a subsequent byte sequence based on the subsequent information in the control data included in the byte sequence passed from the lower layer, from the byte sequence passed from the lower layer in the order indicated by the order information in the control data Record the data byte string excluding the control data, wait for the next byte string, and if it is determined by the subsequent information that there is no subsequent byte string, control is performed from the byte string passed from the lower layer to the recorded data byte string A data transfer method comprising: a reassembling step of adding a data byte sequence excluding data and transferring the combined byte sequence to the NDN layer. The data transfer method according to claim 5, wherein
In the segmentation step, if the size of the input byte sequence passed from the NDN layer is equal to or smaller than the maximum transfer size of the lower layer, there is no order information indicating that the input byte sequence is the first byte sequence and the subsequent byte sequence. And further including a step of passing control data including subsequent information indicating that the
When the reassembly step is determined to be the first byte sequence by the order information in the control data included in the byte sequence passed from the lower layer and the subsequent byte sequence is absent by the subsequent information, from the lower layer A data transfer method, further comprising a step of transferring a data byte sequence excluding control data from the transferred byte sequence to the NDN layer. The data transfer method according to claim 5 or 6, wherein the control data attached to the byte string is used when a bit indicating the presence / absence of a subsequent byte string and a total size of the byte string to be passed to the lower layer are aligned to a predetermined value. A data transfer method comprising: a bit indicating a padding byte length; and a bit indicating the order of the byte string.

Images