JP3069842B2 - High-speed frame transfer method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to a high-speed frame transfer method in a LAN for interconnecting a plurality of networks and transferring frames.

[0002]

2. Description of the Related Art Generally, in a network node that performs frame transfer processing on an optimal (shortest) route, a so-called router device, as shown in FIG. 18, a routing table created based on routing information exchanged between nodes in advance. Is performed on the basis of. As a specific process at a node, first, a destination table of a frame received from a terminal is addressed to itself, and a routing table (optimum route information is stored) using a destination network address stored in the frame as a key. Is shown), and the physical address of the router through which the frame passes next and the output port to the network (transmission path) where the router exists are obtained.

[0003] Then, a frame is reassembled to the obtained next router (the destination physical address is set as the physical address of the next router), and the frame is transmitted to the obtained output port. By repeating these operations at each router, the router that finally accommodates the destination terminal reassembles the frame to the terminal (the destination physical address is used as the physical address of the terminal), and the obtained output port The communication between the terminals is provided by transmitting the frame. Further, by performing frame transfer based on a routing search result based on such a destination network address, a frame can always be transferred along an optimal route.

[0004]

However, in frame transfer at a node by such a router operation, it is necessary to always perform frame reassembly processing at the time of transmission to an output network port. There is a problem that it takes time for the transfer processing in the inside. Therefore, in the future, with an increase in transmission line speed and the spread of multimedia communication, it is necessary to reduce the delay of internal processing at the node and increase the frame transfer capability.
Therefore, an object of the present invention is to reduce the delay of internal processing at a node and improve the frame transfer capability.

[0005]

SUMMARY OF THE INVENTION In order to solve the above-mentioned problems, the present invention provides a frame communication between an originating terminal and a destination terminal by means of each node connecting a plurality of physical transmission lines and a physical transmission line. When the first node accommodating the destination terminal receives a frame addressed to the destination terminal in the network that provides the destination terminal, the first node that stores the destination terminal's network address and the physical address of the own node stored in the received frame. First information for notifying the second node accommodating the originating terminal that transmitted the frame (AR frame), and relaying the information notified by the first means. The third node has second means for extracting and holding the correspondence between the physical address in the information and the port receiving the information, and the second node has Tsu network address, in which to have a third means for holding and extracting the relationship between the reception port of the physical address and the information. Therefore, the second node uses the network address of the destination terminal stored in the frame received from the calling terminal as a search key and compares the physical address of the first node and the output port held in the third means with each other. The third node that has obtained and obtained the physical address as the destination physical address of the received frame and reassembled the frame and transmitted the output port to the output port obtained by reassembling the frame. By transmitting the frame to the output port obtained by acquiring the output port by the second means, the third node reassembles the frame with respect to the received frame in the normal frame transfer (replacement of the destination physical address and network Address header processing), and as a result, high-speed frame transfer can be performed. .

In addition, the information to be notified is a physical address of a destination terminal instead of the physical address of the first node,
Fourth means for holding the correspondence between the physical address and the port accommodating the destination terminal is provided in the first node,
An output port for outputting a frame from the physical address of the destination terminal is obtained. As a result, not only the omission of frame assembly at the third node, but also the output port can be obtained from the destination physical address at the first node, and the frame reassembly processing here can be omitted. Can provide transfer. Further, an identifier for identifying a communication flow between terminals is provided as the information to be notified, and fifth means for holding a correspondence including the identifier is provided in the second node. As a result, by reassembling the obtained physical address as the destination physical address only for the frame in which the destination network address and the flow identifier are constant, the third node and the first node can be used only for the frame of the specific communication flow. , The frame reassembly process can be omitted, and high-speed transfer limited to frames of a specific communication flow can be performed. Further, the first node periodically notifies the information. As a result, even when the communication path is changed due to a failure in the transmission path, information notified after the change is notified to the most upstream node (second node) according to the new path, and the notified information is used. Updating the correspondence enables high-speed transfer via a new relay node (third node) adapted to the route change. Further, the third node, which has not received the information regularly, deletes the correspondence relation regarding the information held in the second means. As a result, high-speed transfer adapted to route change becomes possible. Further, when the first node that periodically transmits the information does not receive the frame addressed to the destination related to the information for a certain period of time, the first node deletes the corresponding relationship related to the information. Send information to each node. As a result, the third and second nodes always maintain an appropriate correspondence, and can perform high-speed transfer while avoiding unnecessary searches.

[0007]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below with reference to the drawings. First, the outline of the present invention will be described with reference to FIGS. FIG. 1 is a diagram showing the principle of the invention corresponding to claim 1 of the present invention. In FIG. 1, when a router (network node; node) Rd accommodating the destination terminal Td receives the first frame addressed to the destination terminal Td, the router Rd stores the network of the destination terminal Td stored in the received frame. Address report information (AR frame) including the address z2 and its own physical address f
Is notified to the router Rs accommodating the calling terminal Ts.

[0008] This AR frame is transferred to the node Rs along a route according to the routing table in the same manner as a normal frame. Also, it is shown that the relay router Rc that has received this AR frame has means 2 for extracting and holding the correspondence between the physical address f stored in this frame and the port number P1 that has received this frame. ing. Further, in the most upstream router Rs accommodating the calling terminal Ts, the correspondence between the network address z2 and the physical address f stored in the AR frame and the port number P1 that has received the frame is extracted and held. This shows that the device has means 3.

By providing each of these means 1 to 3, first, regarding the frame received from the calling terminal Ts by the router Rs, the network address z2 of the destination terminal Td stored in the frame is used as a key by the above means 3 Obtain the physical address f of the router Rd containing the destination terminal Td and the output port P1 for the router,
The frame is reassembled using the obtained physical address f as the destination physical address of the received frame, and the frame is transmitted to the obtained output port P1. In the relay router Rc receiving this frame, the output port P1 is acquired by the means 2 using the destination physical address f of the frame as a search key, and the frame is transmitted to the obtained output port P1.

As a result, in the relay router Rc,
The frame transfer using only the destination physical address of the frame becomes possible, the routing table search and the frame reassembly process in the conventional router can be bypassed, and the same optimal path (the same as the path obtained in the routing table) AR frames are forwarded along a routing table).

Next, FIG. 2 is a diagram showing the principle of the invention corresponding to claim 2 of the present invention. The difference from FIG. 1 is that a means 1A for notifying the physical address h of the destination terminal as an AR frame instead of its own physical address f is held. Similarly to the relay router Rc, the router Rd has means 2A for holding the correspondence between the physical address h and the port P1 accommodating the destination terminal Td. By these means, it becomes possible to perform the frame transfer using only the destination physical address not only in the relay router Rc but also in the router Rd accommodating the destination terminal Td.

Next, FIG. 3 is a diagram showing the principle of the invention corresponding to claim 3 of the present invention. The difference from FIG. 1 is that an identifier Q indicating an upper flow type specified in the received frame is further added as an AR frame, and the router Rs only applies to the relay router Rc for a frame corresponding to the flow identifier Q. Frame transfer using only the destination physical address. This is intended to provide high-speed frame transfer for communication in which the communication time between terminals is long or a communication flow that requires real-time processing.

FIG. 4 is a diagram showing the principle of the invention corresponding to claims 4 and 5 of the present invention. When the router Rd periodically transmits the AR frame, the routing table is changed due to a failure or the like, and even when the optimum path is changed, the frame is transferred to the most upstream router Rs along the changed path. . Therefore, by updating the content notified by the AR frame, it is possible to always perform the transfer using the physical address according to the optimum route. In addition, when the AR frame is not received for a certain period of time, by deleting the correspondence, the optimum correspondence can be always maintained.

FIG. 5 is a diagram showing the principle of the invention corresponding to claim 6 of the present invention. When the node Rd that periodically transmits the AR frame does not receive the frame addressed to the destination terminal Td corresponding to the AR frame for a certain period of time, the node Rd notifies the node Rd of deleting the correspondence. This shows that means 1D for transmitting the address report AR deletion frame is provided. By this means 1D, the relay router Rc and the most upstream router Rs can always hold an appropriate correspondence.

Hereinafter, the main configuration of the present invention will be described in detail with reference to FIGS. FIG. 6 is a block diagram showing a configuration of a node (router device). Node 10
IP routing table 11, direct routing table 12, MAC routing table 13
, A frame processing unit 14 and a received frame holding unit 15
And a switching unit 16 and line interface units 17 and 18.

Here, in the node 10, the frame received from the network is added with the port number received by the corresponding line interface unit, and is stored in the received frame holding unit 15 via the switching unit 16. The frame processing unit 14 extracts the frames stored in the received frame holding unit 15 one by one and performs a processing operation according to the flowcharts shown in the subsequent drawings. And
Based on the obtained output port information, the frame is transmitted from the corresponding line interface unit to the network via the switching unit 16.

FIG. 7 is a diagram showing a format of a frame transferred in the network. This frame format uses M bits represented by 48 bits as a physical address.
The format of an IP address represented by 32 bits as an AC address and a network address is shown. FIG. 8 shows the contents of various types of information set in the address report (AR) frame. Next, FIG.
FIG. 4 is a diagram showing the contents of a P routing table 11; This table 11 is created in advance using a routing protocol (RIP, OSPF) and an address resolution protocol (ARP), which are conventional techniques, and is received according to the processing flow shown in each of FIGS. The IP frame is used for table search.

FIG. 10 is a diagram showing the contents of a direct routing table 12, which is an example of an embodiment corresponding to the means 3 and 3B of the present invention. This table is stored in accordance with the processing flow shown in each of FIGS.
The received IP frame created using the R frame is used for table search. FIG. 11 is a diagram showing the contents of a MAC routing table 13 showing an example of an embodiment corresponding to the means 2 and means 2A of the present invention. This table is created using an AR frame in accordance with the processing flow shown in the following FIGS. 12 to 17 and is used for searching the received IP frame for a table.

The detailed operation of the node 10 will be described below with reference to the flowcharts of FIGS. FIG. 12 is a flowchart showing the main processing of the frame processing unit 14 in the node 10. The frame processing unit 14 processes a frame received by the own node and stored in the received frame holding unit 15 as described above. That is, the destination MAC address in the received frame held in the received frame holding unit 15 is addressed to the own node (in the case of coincidence in step S1), and this received frame is
If the frame is not a frame (in the case of a data frame in step S2), and is not registered in the direct routing table 12 (in the case of "N" in step S3), a normal IP routing table 11 search process is performed (step S4). ).

If the node 10 is not a node accommodating the destination terminal Td ("N" in step S5), normal IP frame relay processing is performed (steps S6, S7, S8). , Acts as a relay node Rc. On the other hand, if this node is the node Rd accommodating the destination terminal Td, the node relays the frame and starts the AR frame processing shown in FIG. 17 (step S12).

FIG. 17 is a flowchart showing an example of an embodiment corresponding to the means 1, 1A, 1B, 1C, 1D, and 2A of the present invention. The AR frame of the node Rd accommodating the destination terminal Td is shown in FIG. The operation of the processing will be described. Node Rd
According to this processing flow, while periodically transmitting an AR frame to register an AR frame with each node,
A indicating deletion when no frames are received for a certain period
An R frame is sent to each node. That is, if the AR frame has not been transmitted, the destination IP in the received frame
The address is read and set as the notification IP address of the AR frame (step S62), and its own physical address is set as the notification MAC address of the AR frame.
The frame is set (step S63). Then, the destination IP address is set (step S64), and the source IP address is set in the AR frame (step S64).
65).

Thereafter, in the case of the node Rd having the means 2A for holding the correspondence between the physical address and the port accommodating the destination terminal Td, the relationship between the notified MAC address and the output port is stored in the MAC routing table. 13 (step S65A). Next, the transmission timer is started (step S66), and the IP routing table 11 is searched using the destination IP address as a key (step S67).
The output port is read (step S68), and the destination MAC
The address is updated (step S69). Then, this frame is output to the output port read in step S68 (step S70), and a predetermined transmission interval (step S70).
After step 71), the flow returns to step S70 to transmit the frame to the output port again. If no frame arrives for a certain period, the transmission timer times out and the process goes to step S72.
Is determined to be "Y", in this case, a deletion notification flag is set and the deletion notification processing of the AR frame to be transmitted is started (step S73).

FIG. 14 is a flow chart showing an example of an embodiment relating to the means 2, 3, 3B of the present invention. The AR in the so-called frame relay node Rc and the most upstream node Rs accommodating the calling terminal Ts is shown. It shows a processing operation when a frame (registration) is received. When the relay node Rc receives the AR frame, the relay node Rc performs registration / update processing in the MAC routing table 13, and the most upstream node Rs executes the direct routing table 12.
Performs registration / update processing for.

That is, when an AR frame is received, A
After updating the R frame reception timer (step S30), the IP routing table 11 is searched using the destination IP address as a key (step S31). Then, it judges whether or not it is the accommodating node Rs of the calling terminal Ts (step S32). If it is not the node Rs (that is, in the case of the relay node Rc), the notification MAC address and the reception port number in the AR frame are It reads out and performs registration / update processing in the MAC routing table 13 (step S3).
3). Then, from the IP routing table 11, MA
Read the C address and output port (step S3
4), the destination MAC address is updated (step S)
35) Thereafter, a process of outputting a frame to an output port is performed (step S36). If the node itself is the most upstream node Rs accommodating the calling terminal Ts, the notification MAC address, the notification IP address, and the reception port number in the AR frame are read and the direct routing table 12 is registered / updated (step). S37). In addition, if the reception timer for the AR frame is monitored and this timer times out, the corresponding entry is set to M
It is deleted from the AC routing table 13 or the direct routing table 12.

FIG. 15 is a flow chart showing an example of an embodiment relating to the means 2D and 3D of the present invention. The processing operation when an AR frame (deletion) is received at the so-called frame relay node Rc and the most upstream node Rs. Is shown. That is, when an AR frame is received, the destination I
The IP routing table 11 is searched using the P address as a key (step S41). Then, oneself is the calling terminal Ts
It is determined whether or not the node is the accommodation node Rs (step S42),
In the case of the relay node Rc, the notification MAC address reception port number in the AR frame is read and the MAC routing table 13 is deleted (step S43). Thereafter, the MAC address and output port are read from the IP routing table 11 (step S44), and the destination M
After updating the AC address (step S45), a process of outputting a frame to an output port is performed (step S4).
6). If the node itself is the most upstream node Rs accommodating the calling terminal Ts, the notification MAC address, the notification IP address, and the reception port number in the AR frame are read and the direct routing table 12 is deleted (step S47). .

FIG. 16 shows a detection process for the direct routing table 12 registered in the process flow of FIG. In this processing flow, the most upstream node R
In s, a frame is assembled using the MAC address notified in the AR frame as the destination MAC address and transmitted to the network. That is, the direct routing table 12 is searched using the destination IP address as a key (step S51). If there is a corresponding IP address (step S5
If “Y” in step 2), the MAC address and output port are read from the table 12 (step S53). Then, the destination MAC address is updated (step S5).
4) The frame is transmitted to the read output port (step S55). If the IP address cannot be searched, the process proceeds to step S4 in FIG. 12, and the IP routing table 11 is searched based on the IP address.

FIG. 13 shows a search process for the MAC routing table 13 registered in the processing flow of FIG. 14. In the relay node Rc, the MAC address notified in the AR frame is searched as a destination MAC address and transmitted to the network. I do. That is, the MAC routing table 13 is searched by the destination MAC address (step S21). Then, the MAC routing table 1
3 is read out (step S23), and a frame is transmitted to the output port (step S23).

As described above, in the terminal-to-terminal communication via the router, the frame is transferred by the normal router operation, but once the physical address of the destination terminal accommodation router Rd or the physical address of the destination terminal is notified, In the router Rs accommodating the calling terminal Ts, the physical address is added instead of the physical address of the adjacent router, and in all the relay routers Rc through which the frame passes, not the router operation but the router operation is performed. The transfer process can be performed by a so-called bridge operation of searching for an output port by a destination physical address. Therefore, the transfer processing capability can be improved as compared with the router operation, and the frame can be transferred along the optimal path obtained by the router operation.

[0029]

As described above, according to the present invention, when a frame addressed to a destination terminal is received at the first node accommodating the destination terminal, the destination node stored in the received frame is transmitted to the first node. A first means for notifying the second node accommodating the calling terminal that transmitted the frame of information (AR frame) comprising the network address and the physical address of the own node; A third node that relays the information notified by a second node that extracts and holds a correspondence between a physical address in the information and a port that has received the information, the second node Then, the third node extracts and holds the relationship between the network address, the physical address, and the reception port of the information in the information, so that the third node uses a normal frame. Can omit the reassembly processing of the frame for the received frame in the arm forward, this result,
High-speed frame transfer becomes possible as a system.

Further, the information to be notified is a physical address of a destination terminal instead of the physical address of the first node,
Fourth means for holding the correspondence between the physical address and the port accommodating the destination terminal is provided in the first node,
Since the output port for outputting the frame is obtained from the physical address of the destination terminal, not only the omission of frame assembly at the third node can be omitted, but also the output port can be obtained from the destination physical address at the first node. Therefore, since the frame reassembly process can be omitted, higher-speed frame transfer can be provided. Also, an identifier for identifying a communication flow between terminals is provided as the information to be notified, and a fifth means for holding a correspondence relationship including the identifier is provided in the second node. By reassembling the obtained physical address as the destination physical address only for the frame having a constant value, the frame reassembly process at the third node and the first node can be omitted for only the frame of the specific communication flow. Thus, high-speed transfer limited to frames of a specific communication flow can be performed. In addition, since the first node periodically notifies the information, even if the communication path is changed due to a transmission path failure or the like, the information notified after the change is changed according to the new path. By updating the correspondence with the notified information, high-speed transfer via a new relay node adapted to the route change becomes possible. Further, the third node, which does not receive the information periodically, deletes the correspondence relation of the information held in the second means, so that high-speed transfer adapted to route change can be performed. Become. Further, when the first node which periodically transmits the information does not receive a frame addressed to the destination regarding the information, the first node transmits deletion information for notifying that the correspondence relation regarding the information is to be deleted. Since the information is transmitted to each node, the third and second nodes always maintain an appropriate correspondence, and high-speed transfer without unnecessary search becomes possible.

[Brief description of the drawings]

FIG. 1 is an explanatory diagram illustrating a main part of a first invention.

FIG. 2 is an explanatory diagram illustrating a main part of a second invention.

FIG. 3 is an explanatory diagram illustrating a main part of a third invention.

FIG. 4 is an explanatory diagram illustrating a main part of a fourth invention.

FIG. 5 is an explanatory diagram illustrating a main part of a fifth invention.

FIG. 6 is a block diagram illustrating a configuration of a node to which the present invention has been applied.

FIG. 7 is a diagram showing a format of a frame transferred in a network.

FIG. 8 is a diagram showing various information set in an address report (AR) frame.

FIG. 9 is a diagram showing contents of an IP routing table.

FIG. 10 is a diagram showing contents of a direct routing table.

FIG. 11 is a diagram showing the contents of a MAC routing table.

FIG. 12 is a flowchart illustrating an operation of a frame processing unit.

FIG. 13 is a flowchart illustrating a search process for a MAC routing table.

FIG. 14 is a flowchart illustrating a processing operation when an AR frame (registration) is received.

FIG. 15 is a flowchart showing a processing operation when an AR frame (deletion) is received.

FIG. 16 is a flowchart illustrating a search process for a direct routing table.

FIG. 17 is a flowchart illustrating an operation of an AR frame process.

FIG. 18 is an explanatory diagram illustrating an operation of a conventional system.

[Explanation of symbols]

10, Rs, Rc, Rd: node (router), 11: I
P routing table, 12 direct routing table, 13 MAC routing table, 14
... Frame processing unit, 15 ... Reception frame holding unit, 16 ...
Switching units 17, 18, line interface unit, Ts originating terminal, Td destination terminal.

Continuation of front page (56) References JP-A-10-164119 (JP, A) JP-A-10-136016 (JP, A) JP-A 8-79294 (JP, A) JP-A 8-32597 (JP) , A) JP-A-7-202908 (JP, A) JP-A-6-46062 (JP, A) Proceedings of the 1993 Fall Meeting of the Institute of Electronics, Information and Communication Engineers, 3 Communication, B-390, Nakamura Nakanaga et al. -Storage method of connectionless communication in LAN ”IEICE Technical Report, Vol. 93 No. 257, SSE93-75, Kazuo Sakakawa et al., "A storage method for connectionless communication in ATM-LAN", pages. 69-74 1995 IEICE General Conference, Transactions 2, SB-5-9, Shigehiko Ushijima et al. "Application of ATM connectionless communication to multimedia applications", pages. 666-667 (58) Field surveyed (Int.Cl. ⁷ , DB name) H04L 12/46 H04L 12/28 H04L 12/66

Claims (6)

(57) [Claims] A first node accommodating a destination terminal in a network that provides frame communication between an originating terminal and a destination terminal by each node connecting a plurality of physical transmission paths and the physical transmission path. When a frame addressed to the destination terminal is received, information including the network address of the destination terminal stored in the received frame and the physical address of the own node is stored in the second frame containing the calling terminal that transmitted the frame. And a third node that relays the information notified by the first means. The third node relays the physical address in the information and a port that has received the information. A second node for extracting and holding the correspondence, wherein the second node extracts the relationship between the network address, the physical address and the reception port of the information in the information. Frame fast transfer system characterized by having a third means for holding the. 2. The method according to claim 1, wherein the information to be notified is the physical address of the destination terminal instead of the physical address of the first node, and the physical address of the first node and the port accommodating the destination terminal. 4. A high-speed frame transfer method, comprising: providing a fourth means for holding the correspondence of (1) to (3) to obtain an output port for outputting a frame from the physical address of the destination terminal. 3. The method according to claim 1, wherein an identifier for identifying a communication flow between terminals is provided as the information to be notified, and a second node holds a correspondence including the identifier. A high-speed frame transfer method, characterized in that a means is provided. 4. The high-speed frame transfer method according to claim 1, wherein the first node periodically notifies the information. 5. The method according to claim 4, wherein the third node that has stopped receiving the information periodically includes:
A high-speed frame transfer method, wherein a correspondence relation regarding information held in a second means is deleted. 6. The method according to claim 1, wherein the first node that periodically transmits the information does not receive a frame addressed to a destination related to the information. Transmitting a deletion information to each node for notifying that the correspondence relation regarding the information is deleted, to a high-speed frame transfer method.