JP2003324463A - Communication path switching apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a communication path switching apparatus in which a transmission destination of a packet is detoured before a packet wherein a fault occurs arrives by detecting the fault in a layer 1, and a time for detecting the fault on a path where the fault occurs is shortened. <P>SOLUTION: A PHY part 11 detects a fault in a physical layer, and an LF/RF detecting part 12 changes a port status table 21 to be referred to by a layer 2 switch 30 based upon a status of fault occurrence. When the fault is a local fault, in order to report the fault in an up communication direction as well, an RF transfer part 13 prepares an RF packet and outputs it by inserting a remote fault/RF/ between a terminate/T/ and an idle/I/ of data generated by a MAC part 14. <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to line failure detection and data transfer, and more particularly to a communication path switching device for detecting line failure and data transfer on a physical layer in Ethernet (registered trademark). .

[0002]

2. Description of the Related Art A LAN (Local Area Network) is a limited network such as a premises or a building.
In recent years, there are devices such as routers that connect LANs to each other, and it is possible to configure a network in a wide range.

As such a conventional technique, Japanese Patent Laid-Open No. 2001-2001
Japanese Patent Laid-Open No. 186140 discloses a technology of an ATM (Asynchronous Transfer Mode) exchange having a LAN interface function.

FIG. 15 shows an LA used in the prior art.
3 is a block diagram showing a configuration of an N adapter unit 50. FIG. LA
The N adapter unit 50 includes a physical interface unit 52-1.
To 52-4, the processing unit 54, and the address table 56.
, Memory 58, and cell switch interface unit 6
0 and a control interface unit 62.

The processing unit 54 controls various processes performed by the LAN adapter unit 50, such as a layer 2 address resolution function and an address learning function. The physical interface units 52-1 to 52-4 are physical interfaces of Ethernet (registered trademark), for example, and 10BASE-
T / 100BASE-TX can be automatically switched.

The address table 56 constitutes a layer 2 or layer 3 address management table. The memory 58 stores the microprogram of the processing unit 54 and necessary data. The cell switch interface unit 60 is an interface between the processing unit 54 and the cell switch. In addition, the control interface unit 62
Is an interface between the LAN adapter unit 50 and the control unit of the ATM exchange.

The LAN adapter section 50 transmits and receives a unicast frame including at least a destination MAC address and a source MAC address.

[0008]

However, in the above-mentioned conventional technique, the destination MAC address of the packet is referred to from the address management table without detecting the failure in the layer 1 (physical layer) of the received packet, and the destination cell is referred to. Since the address is determined, when a failure is detected in the layer 2 (data link layer), there is a problem that the failure is not detected in time and the cell with the failure is transmitted.

Further, when a failure occurs in the MAC address itself, the failed MAC address is referred to from the address table before the failure, which causes a problem of erroneous switching.

The present invention has been made in view of the above,
An object of the present invention is to detect a fault in layer 1, to bypass the destination of the packet before the faulty packet arrives, and to obtain a communication route switching device in which the fault detection time of the faulty route is short.

[0011]

In order to achieve the above object, a communication path switching device according to the present invention performs physical layer communication control and detects a physical layer fault in the physical layer. And a physical layer failure transfer means for notifying the detected failure to the data of the opposite communication when the failure detected by the physical layer failure detection means is a failure in the communication direction from the physical layer to the MAC layer. , MAC
A plurality of ports provided with MAC means for controlling communication of layers, a port state storage means for storing states of the plurality of ports, and an address storage means for storing communication paths of the plurality of ports by port addresses. , A plurality of ports are connected, and a layer 2 switching unit for performing layer 2 switching is provided.

According to the present invention, the physical layer failure detecting means detects a failure in the physical layer in the physical layer, and when the detected failure is a failure in the communication direction from the physical layer to the MAC layer, the physical layer failure is detected. The transfer means also notifies the data of the opposite communication of the detected failure.

A communication path switching device according to the next invention is
In the above invention, the physical layer failure transfer means is
When the physical layer failure detection means detects a failure,
The port state of the port state storage means is changed, and the layer 2 switching means stores the address so as to transmit the packet by bypassing the faulty port based on the changed port state of the port state storage means. It is characterized in that the route information regarding a port in which a failure is detected is changed and the packet is transmitted.

According to the present invention, the physical layer failure transfer means changes the port status of the port status storage means when the physical layer failure detection means detects a failure, and the layer 2 switching means changes the port status. Based on the port state of the state storage means, the route information relating to the port in which the failure is detected in the address storage means is changed so that the packet is transmitted around the faulty port, and the packet is transmitted. .

A communication path switching device according to the next invention is
In the above invention, the MAC unit inserts physical layer failure information for notifying that the physical layer failure detection unit has detected a failure into a data area in a MAC packet.

According to the present invention, the MAC means is a MAC
Physical layer failure information for notifying that the physical layer failure detection means has detected a failure is inserted in the data area in the packet.

A communication path switching device according to the next invention is
In the above invention, the MAC unit acquires the source address and the destination address of the MAC packet in which the physical layer failure information is inserted from the port state storage unit and the address storage unit.

According to the present invention, the MAC means acquires the source address and the destination address of the MAC packet in which the physical layer failure information is inserted from the port state storage means and the address storage means.

A communication path switching device according to the next invention is
In the above invention, the physical layer failure transfer means is
When notifying the failure detection to the other node, the information notifying the failure is inserted into a time slot different from the packet sequence transmitted from the MAC means.

According to the present invention, the physical layer failure transfer means inserts the failure notification information into a time slot different from the packet sequence transmitted from the MAC means when notifying the failure detection to another node. ing.

[0021]

BEST MODE FOR CARRYING OUT THE INVENTION Preferred embodiments of a communication path switching device according to the present invention will be described in detail below with reference to the accompanying drawings.

FIG. 1 is a block diagram showing the configuration of the communication path switching apparatus of this embodiment. The communication path switching device of the present embodiment includes a plurality of (n in this case) ports 10 for transmitting and receiving Ethernet (registered trademark), a table unit 20 for storing MAC addresses and port states, and a layer 2 switch 30. I have it.

The port 10 has a PHY unit 11 and LF / RF.
The detection unit 12, the RF transfer unit 13, and the MAC unit 14 are provided.

The PHY unit 11 is connected to another node via a transmission line, and is in layer 2 (in this case, the MAC unit 14).
And an interface function with another node, transmitting the data input from the RF transfer unit 13 to the other node (uplink communication), and receiving the data from the other node (downlink communication), the LF / RF detection unit 12 Output to. Also, it detects a failure (local fault) in downlink communication.

The LF / RF detector 12 changes the port state table 21 when a local fault or a remote fault is detected.

The RF transfer unit 13 inserts the data for notifying the remote fault into the opposite data in order to notify the other node on the downstream side of the local fault detected by the PHY unit 11.

The MAC unit 14 has a layer 2 interface function, and when the PHY unit 11 detects a local fault, an RF packet for notifying another port that a local fault has occurred in the port 10. Is generated and output to the layer 2 switch 30.

The table unit 20 is a port state table 2
1 and an address table 22. As shown in FIG. 8, the port status table 21 includes the layer 2 switch 3
Port number P which is all ports connected to 0
The state is stored for each of # 1 to P # n. For example,
FIG. 8A shows a case where all the ports are normal, and "OK" indicating that the port is available is stored corresponding to the port number. The address table 22 stores the route status of the port 10 of this device as shown in FIG. For example, FIG. 9A shows a route state in which the port 10P # 1 and the ports 10P # 2 to P # n are connected in a one-to-n-1 manner.

The layer 2 switch 30 includes a plurality of ports 1
0 (n in this case) is connected, the packet transmitted from the port is analyzed, the port state of the transmission destination is confirmed, and the port is switched.

FIG. 2 shows a data format between the PHY unit 11 and the MAC unit 14. The data format between the PHY section 11 and the MAC section 14 is an encapsulated data format surrounded by a start flag / S / and a terminate / T /. Between the start flag / S / and the terminate / T /, there is one or more data area data / D /, and the data / D / accommodates the MAC frame shown in FIG. 3 to form a packet. Idle / I / is inserted between these packets.

FIG. 3 shows a MAC frame format between the MAC unit 14 and the layer 2 switch 30. MA between the MAC unit 14 and the layer 2 switch 30
The C frame format is a 7-byte preamble PR used for synchronization establishment, a 1-byte frame start delimiter SFD, a 6-byte destination address DA indicating a packet destination address, and a 6-byte transmission indicating a packet source address. Original address SA,
A packet length / type Ln of 2 bytes indicating the length and type of the packet, transmission data DATA of a variable length of 46 to 1500 bytes, and a frame check FCS of 4 bytes used for detecting a data error of the frame.
Composed of.

Referring to the flow charts of FIGS. 4 to 7,
The operation of the communication path switching device of this embodiment will be described. First, port 10P # 1 and ports 10P # 2-P # n are 1
The operation of the downlink communication in the normal state when the pair n-1 is connected will be described with an example focusing on the port 10P # 3.

The PHY unit 11 of the port 10P # 3 receives the data of the data format shown in FIG. 2 (data at the normal time) from another node. When data is received, PHY
The unit 11 detects a failure. The failure is detected as a bit error, for example, when the decoding process is a code that violates the coding rule. When connected by an optical cable, it can be detected even if the signal is broken. If there is no obstacle, P
The HY unit 11 outputs the received data as it is to the LF / RF detection unit 12.

When data is input from the PHY unit 11,
The LF / RF detection unit 12 detects whether or not the data input from the PHY unit 11 includes information notifying that a remote fault has occurred (step S1 in FIG. 4).
00). If a remote fault has occurred, see Figure 1.
As shown in 3, since the remote fault / RF / is inserted immediately after the termination / T / of the data format between the PHY unit 11 and the MAC unit 14, this remote fault / RF / is included in the data. Is detected. In this case, since the data input from the PHY unit 11 is idle / I / immediately after termination / T / as shown in FIG. 2, the remote fault is not detected.

When the remote fault is not detected, the LF / RF detection unit 12 detects whether or not the data contains information notifying that a local fault has occurred (step S120 in FIG. 4). When a local fault has occurred, as shown in FIG. 11, since the local fault / LF / is inserted in the data, it is detected whether or not this local fault / LF / is included in the data. . In this case, the data input from the PHY unit 11 does not include the local fault / LF / as shown in FIG. Therefore, no local fault is detected. When the detection of the remote fault and the local fault is completed, the LF / RF detection unit 12 outputs the data to the MAC unit 14.

When data is input from the LF / RF detection unit 12, the MAC unit 14 detects whether or not the input data includes information for notifying that a local fault has occurred (see FIG. 6 steps S300). Similar to the LF / RF detection unit 12, the detection is performed depending on whether or not the data includes the local fault / LF /. In normal operation, since the data is the data shown in FIG. 2, the local fault / LF / is detected. Not done.

When the local fault / LF / is not detected in the data, that is, when the data is normal,
The MAC unit 12 creates a normal MAC packet as shown in FIG. 3 and transmits the data to the layer 2 switch 30 (step S310 in FIG. 6). In this case, the MAC address of port 10P # 1 is inserted into the destination address DA, and the MAC address of port 10P # 3 is inserted into the source address SA.

Upon receiving the data from the MAC unit 14, the layer 2 switch 30 refers to the port state table 21 and confirms the port state indicated by the destination address DA and the source address SA in the MAC frame (step S400 in FIG. 7). ). FIG. 8A shows the contents of the port status table 21 at the normal time. In the port status table 21 shown in FIG. 8A, the port 10P # 1 and the port 10P are
Since "OK" is displayed for all the ports # 3, it is confirmed that no fault has occurred in the ports.

When the port status is checked by referring to the port status table 21, the layer 2 switch 30 checks the communication path by referring to the address table 22 (step S410 in FIG. 7). As shown in FIG. 9A, the address table 22 in the normal state shows that the port 10P # 3 is the port 10
It is connected to P # 1.

The layer 2 switch 30 determines from the result of referring to the port status table 21 whether or not a remote fault has occurred in the destination port 10 (step S420 in FIG. 7). This is because if a remote fault has occurred in the destination port 10, even if a packet is transmitted, that port 10 cannot process it. When the remote fault has not occurred in the destination port 10, packet transmission processing is performed according to the normal protocol (step S430 in FIG. 7).

Here, as shown in FIG. 10, the port 10P
It is assumed that a failure has occurred (a local fault has occurred) in the receiving side of the # 3 PHY unit 11, the transmission path through which the signal received by the PHY unit 11 is transmitted, or another node. PHY
When a failure occurs on the receiving side of the unit 11, the transmission path on which the signal received by the PHY unit 11 is transmitted, or another node,
The PHY unit 11 detects the failure, and as shown in FIG. 11, local failure / LF /
Convert to. The PHY unit 11 converts the converted data to LF /
Output to the RF detection unit 12.

When data is input from the PHY unit 11,
The LF / RF detection unit 12 detects whether or not the data input from the PHY unit 11 includes information notifying that a remote fault has occurred (step S1 in FIG. 4).
00). The data input from the PHY unit 11 is shown in FIG.
As shown in FIG. 3, the remote fault is not detected because the termination / T / and the remote fault / RF / are not included.

When the remote fault is not detected, the LF / RF detection unit 12 detects whether or not the data includes information notifying that a local fault has occurred (step S120 in FIG. 4). As shown in FIG. 11, the data input from the PHY unit 11 includes the local fault / LF /. Therefore, LF
The / RF detection unit 12 detects this local fault / LF /. When a local fault / LF / is detected, L
The F / RF detection unit 12 notifies the RF transfer unit 13 that a local fault has occurred (step S14 in FIG. 4).
0).

When the RF transfer unit 13 is notified of the occurrence of the local fault, the LF / RF detection unit 12 changes the port state of the port 10 in the port state table 21 and performs the port use prohibition process (step S15 in FIG. 4).
0). Specifically, as shown in FIG. 8B, the normal port status table 21 shown in FIG. 8A shows the port status of the port 10P # 3 in which the local fault has occurred as “OK”. To "LF". As a result, referring to the port status table 21 thereafter, the port 10P
# 3 becomes able to recognize that a local fault has occurred. If you change the port status table 21,
The LF / RF detection unit 12 outputs the data to the MAC unit 14.

When data is input from the LF / RF detection unit 12, the MAC unit 14 detects whether or not the input data includes information for notifying that a local fault has occurred (see FIG. 6 steps S300). LF /
Since the data input from the RF detection unit 12 includes the local fault / LF / as shown in FIG. 11, the MAC unit 14 detects this local fault / LF /. When the local fault / LF / is detected in the data input from the LF / RF detection unit 12, the MAC unit 14 creates the RF packet shown in FIG. 12 (step S320 in FIG. 6).

In the data input from the LF / RF detection unit 12, the local fault / LF / is inserted as shown in FIG. 11, and the destination address DA and the source address SA shown in FIG. 3 are all overwritten. Information is lost. The MAC unit 14 searches the port state table 21 changed by the LF / RF detection unit 12 for a port in which a local fault has occurred. As shown in FIG. 8B, since the port status table 21 stores “LF” indicating a local fault in the port 10P # 3, the port 10P # 3 is detected. Port 10P # 3
When the MAC address is detected, the MAC unit 14 acquires the address of the relevant transfer destination from the address table 22, and uses this address as the source address SA of the RF packet for the source address of the signal in which the local fault / LF / has occurred. insert. That is, the address of the port 10P # 3 is inserted into the source address SA of the RF packet. Further, since it is known from the address table 22 that the port connected to the port 10P # 3 is the port 10P # 1, M
The AC unit 14 inserts the address of the port 10P # 1 into the destination address DA. Further, the transmission data DA is normally transmitted between the packet length / type Ln and the frame check FCS.
A remote fault / RF / of 7 bytes is inserted in the portion where the TA is inserted, and the RF packet shown in FIG. 12 is created. The created RF packet is transmitted to the layer 2 switch 30.

Upon receiving the MAC packet from the port 10P # 3, the layer 2 switch 30 refers to the port state table 21 and refers to the source address SA in the MAC frame.
Then, the port state indicated by the destination address DA is confirmed (step S400 in FIG. 7). The destination address DA of the MAC packet is the port 10P # 1, and the changed port state table 21 can use the port 10P # 1 as shown in FIG.
0 determines that the port 10P # 1 is available. When confirming the port state of the destination address DA, the layer 2 switch 30 refers to the address table 22 shown in FIG. 9B and recognizes that the transmission route is from the port 10P # 3 to the port 10P # 1 (FIG. 7). Step S41
0). Destination, that is, destination address D of MAC packet
It is determined whether or not the port 10P # 1 which is the port A is a remote fault (step S420 in FIG. 7). Since the port 10P # 1 is available, the layer 2 switch 30 performs packet transmission processing (step S in FIG. 7).
430).

The layer 2 switch 30 uses the port 10P #
When transmitting a MAC packet from port 1 to port 10P # 3, the layer 2 switch 30 uses the port state table 2
1, the port state indicated by the source address SA and the destination address DA in the MAC frame is confirmed (step S400 in FIG. 7). Since the port state table 21 stores “LF” indicating that the port state of the port 10P # 3 is a local fault as shown in FIG. 8B, the layer 2 switch 30 sets the port 10P # 3 to the port 10P #.
Determine that 3 is a local fault. When the port status is confirmed by referring to the port status table 21, the layer 2 switch 30 recognizes that the transmission route is from the port 10P # 1 to the port 10P # 3 by referring to the address table 22 (step S410 in FIG. 7). .

The layer 2 switch 30 determines whether or not the port state of the destination port 10P # 3 is a remote fault (step S420 in FIG. 7). Port 1
The port state of 0P # 3 is a local fault, that is, a failure on the receiving side, and it is determined that transmission processing is possible because no failure has occurred on the transmitting side, and packet transmission processing is performed (step S430 in FIG. 7). .

On the other hand, when receiving the notification from the LF / RF detection unit 12, the RF transfer unit 13 determines whether the notification from the LF / RF detection unit 12 is a local fault (step S200 in FIG. 5). In this case, the LF / RF detection unit 12 has detected a local fault. Therefore, the RF transfer unit 13 uses the opposite data (MAC unit 14
13 to the PHY unit 11), a remote fault / RF / for notifying a remote fault is inserted immediately after the termination / T / as shown in FIG. 13 (step S220 in FIG. 5). That is, the data length of the opposing data is longer than that of the normal data format shown in FIG. 2 by the inserted remote fault / RF /. The data with remote fault / RF / inserted is
It is output to the PHY unit 11. The PHY unit 11 transmits the data shown in FIG. 13 to another node.

Next, the packet transmitted from the port 10P # 1 passes through the layer 2 switch 30 to the port 10P.
The operation of port 10P # 3 when # 3 receives (uplink communication) will be described. The operation until the port 10P # 1 transmits data to the layer 2 switch 30 is the same as the normal operation of the downlink communication described above, and thus the description thereof is omitted here.

Upon receiving the data from the layer 2 switch 30, the MAC unit 14 generates the data in the format shown in FIG. 2 and outputs it to the RF transfer unit 13. Here, the received packet (see FIG. 3) is inserted into the data / D / in FIG.

When data is input from the MAC unit 14,
The RF transfer unit 13 determines whether or not the data includes information on a local fault (step S2 in FIG. 5).
00). In this case, as shown in FIG. 2, the local fault / LF / is not included in the data. Therefore, the RF transfer unit 13 does not detect the local fault / LF /. When the local fault / LF / is not detected, the RF transfer unit 13 outputs the data as it is to the PHY unit 11 (step S210 in FIG. 5).

When data is input from the RF transfer unit 13, the PHY unit 11 outputs the data to another node.

Next, as shown in FIG.
The operation when a failure occurs in the transmission path from the PHY unit 11 of P # 3 or the port 10P # 3 to another node will be described by focusing on the port 10P # 3. This fault is detected by the reception side of the PHY unit 11 of another node (not shown) (local fault), and the LF / RF detection unit 12 notifies the RF transfer unit 13 of the fault, so that the return remote Port 10P via PHY unit 11 as a fault
Sent to # 3.

The PHY unit 11 of the other node transmits the data shown in FIG. 13, and the PHY unit 11 of the port 10P # 3 receives the data. Since no failure is detected on the receiving side, the PHY unit 11 outputs the received data to the LF / RF detection unit 12.

When data is input from the PHY unit 11,
The LF / RF detection unit 12 detects whether or not the data input from the PHY unit 11 includes information notifying that a remote fault has occurred (step S1 in FIG. 4).
00). The data input from the PHY unit 11 is shown in FIG.
As shown in, the remote fault / RF / is inserted after the termination / T /. Therefore, LF /
The RF detection unit 12 detects this remote fault / RF /. LF is detected when remote fault / RF / is detected.
The / RF detection unit 12 changes the port status for the own port in the port status table 21 and performs the port use prohibition process (step S130 in FIG. 4). Specifically, FIG. 8 (c)
As shown in, the port status of port 10P # 3 is changed to "R
Change to "F".

When the port status table 21 is changed,
The LF / RF detection unit 12 detects whether or not the data includes information for notifying that a local fault has occurred (step S120 in FIG. 4). Since the data input from the PHY unit 11 does not include the local fault / LF / as shown in FIG. 13, the local fault is not detected. When the detection of the local fault is completed, the data is output to the MAC unit 14.

When data is input from the LF / RF detection unit 12, the MAC unit 14 detects whether or not the input data includes information for notifying that a local fault has occurred (see FIG. 6 steps S300). The input data is the local fault /
Since the LF / is not included, the MAC unit 14 transmits the packet to the layer 2 switch 30 (step S in FIG. 6).
310).

Here, from port 10P # 1 to port 10
It is assumed that the packet for P # 3 is transmitted to the layer 2 switch 30. The layer 2 switch 30 refers to the port state table 21 and confirms the port state indicated by the source address SA and the destination address DA in the MAC frame (step S400 in FIG. 7). Port status table 2
1, the port status of the port 10P # 3 is "RF" as shown in FIG.
Determines that port 10P # 3 has a local fault. When the port status is confirmed by referring to the port status table 21, the layer 2 switch 30 recognizes that the transmission route is from the port 10P # 1 to the port 10P # 3 by referring to the address table 22 (step S in FIG. 7).
410).

The layer 2 switch 30 determines whether or not the port state of the destination port 10P # 3 is a remote fault (step S420 in FIG. 7). Port 1
The port status of 0P # 3 is a remote fault, and even if data is transmitted, it will be lost due to the failure. for that reason,
The layer 2 switch 30 selects a port other than the port of the port 10P # 3 and changes the address table 22 (step S440 in FIG. 7). Specifically, as shown in FIG. 9C in the address table 22 in the normal state shown in FIG. 9A, the transmission path of the faulty port 10P # 3 is changed to another port (in this case, p). Since the address table 22 has been changed, port 10P #
No transmission is made to port 3, and the route is detoured to port 10P # p.

As described above, in this embodiment, the PHY unit 1
1 detects a fault in the physical layer, and the LF / RF detection unit 12
Changes the port status table 21 referred to by the layer 2 switch 30 based on the failure occurrence status. When the fault is a local fault, the RF transfer unit 13 creates an RF packet in order to notify the fault in the upstream communication direction as well, and terminates / Ts the data generated by the MAC unit 14.
Insert remote fault / RF / between / and idle / I / and output. This reduces the physical layer failure detection and port switching time when a failure occurs,
It is possible to notify another node of the occurrence of a failure without affecting the packet generated by the AC unit 14.

When the local fault occurs, the position of the local fault / LF / inserted by the PHY unit 11 may be anywhere as long as it is the start flag / S / or later.

[0064]

As described above, according to the present invention, the physical layer failure detection means detects a failure in the physical layer at the physical layer, and the detected failure is in the communication direction from the physical layer to the MAC layer. In the case of a failure, the physical layer failure transfer means also notifies the data of the opposite communication of the detected failure, so that the failure detection time can be shortened.

According to the next invention, the physical layer failure transfer means changes the port status of the port status storage means when the physical layer failure detection means detects a failure, and the layer 2 switching means is changed. Based on the port status of the port status storage means, the route information regarding the port in which the failure is detected in the address storage means is changed so that the packet is sent around the failed port, and the packet is sent. Therefore, the failure detection time and the port switching time can be shortened as compared with the case where the failure detection and the port detection of the failure are performed only in the MAC layer.

According to the next invention, the MAC means is an MA.
Since the physical layer fault information for notifying that the physical layer fault detecting means has detected the fault is inserted into the data area in the C packet, other information is connected to the port in Ethernet (registered trademark). The node can also be notified of the failure occurrence.

According to the next invention, the MAC means acquires the source address and the destination address of the MAC packet in which the physical layer failure information is inserted from the port state storage means and the address storage means.
Even if the AC address is lost, the packet can be sent to the original destination address.

According to the next invention, the physical layer failure transfer means inserts the failure notification information into the time slot different from the packet sequence transmitted from the MAC means when notifying the failure detection to the other node. Because
Faults can be forwarded without affecting legitimate packets.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of an embodiment of a communication path switching device according to the present invention.

FIG. 2 is a diagram showing a data format in a physical layer.

FIG. 3 is a diagram showing a data format in a MAC layer.

FIG. 4 is a flowchart for explaining the operation of the LF / RF detection unit.

FIG. 5 is a flowchart for explaining the operation of the RF transfer unit.

FIG. 6 is a flowchart for explaining the operation of the MAC unit.

FIG. 7 is a flowchart for explaining the operation of the layer 2 switch.

FIG. 8 is a diagram showing a port status table.

FIG. 9 is a diagram showing an address table.

FIG. 10 is a block diagram showing a configuration of a communication path switching device when a failure occurs on the receiving side.

FIG. 11 is a diagram showing a data format in a physical layer when a local fault occurs.

FIG. 12 is a diagram showing a format of an RF packet for notifying a remote fault.

FIG. 13 is a diagram showing a data format in a physical layer for notifying a remote fault.

FIG. 14 is a block diagram showing the configuration of a communication path switching device when a failure occurs on the transmission side.

FIG. 15 is a block diagram showing a configuration of a conventional LAN adapter unit.

[Explanation of symbols]

10 ports, 11 PHY section, 12 LF / RF detection section, 13 RF transfer section, 14 MAC section, 20 table section, 21 port status table, 22, 56 address table, 30 layer 2 switch, 50 LAN adapter section, 52- 1, 52-2, 52-3, 52-4 P
HY, 54 processing unit, 60 cell switch interface unit, 62 control interface.

Claims (5)

[Claims] 1. A physical layer failure detecting means for controlling a physical layer communication and detecting a failure in the physical layer in the physical layer, and a failure detected by the physical layer failure detecting means from the physical layer to the MAC layer. In the case of a failure in the communication direction, a physical layer failure transfer means for notifying the detected failure to the data of the opposite communication, a MAC means for controlling the communication of the MAC layer, and a plurality of ports, A port state storage unit that stores the state of the port; an address storage unit that stores the communication paths of the plurality of ports by port addresses; and a layer 2 switching unit that connects the plurality of ports and performs layer 2 switching A communication path switching device comprising: 2. The physical layer failure transfer means changes the port status of the port status storage means when the physical layer failure detection means detects a failure, and the layer 2 switching means changes the port status. Changing the route information related to the port in which the failure is detected in the address storage means based on the port state of the port state storage means so as to transmit the packet by bypassing the failed port, and transmitting the packet. Claim 1 characterized by
The communication path switching device described in. 3. The MAC means inserts physical layer failure information for notifying that the physical layer failure detection means has detected a failure into a data area in a MAC packet. Alternatively, the communication path switching device according to item 2. 4. The MAC means acquires a source address and a destination address of a MAC packet in which the physical layer failure information is inserted from the port state storage means and the address storage means.
The communication path switching device described in. 5. The physical layer failure transfer means, when notifying the failure detection to another node, inserts the failure notification information into a time slot different from the packet sequence transmitted from the MAC means. The communication path switching device according to claim 1.