JP2001053771A - Communication control method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To realize just access to a network by preferentially allocating a carrier for transmitting a packet generated in a self-node. SOLUTION: A reception packet formed of a light signal is demultiplexed into three wavelengths λ1 to λ3 by a demultiplexer(DEMUX) 151. The light signals of the respective wavelengths are converted into electric signals by an optic/electric converter (O/E) 152. Data packets converted into the electric signals are judged to be the packets generated in a node or not. When they are decided to be the packets generated in the node, it is decided whether an idle queue buffer 101 only for the node exists or not. When the idle buffer exists, the packet is stored in the queue buffer 101 only for the node. The stored packets are sequentially converted into the light signals of the wavelength λ1 by an electric/optic converter (E/O) 154 and are multiplexed by a multiplexer(MUX) 155.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a communication control method in a node of an ultra-high-speed and large-capacity communication network such as a photonic network.

[0002]

2. Description of the Related Art In recent years, a ring-shaped network configuration has attracted attention as a network configuration method in a high-speed data communication system. FIG. 14 shows a method using a spatial reuse protocol (SRP) among protocols for configuring a ring network.

In this method, one ring is provided in each of different propagation directions, and data transmission between the nodes is performed in accordance with the rules of shortest route delivery and destination removal. The outline of the operation is as follows. . Here, nodes 1 to 8
Consider a ring network consisting of eight nodes
Each of the nodes 1 to 8 is connected by a pair of rings having different propagation directions, and these are called a clockwise ring 141 and a counterclockwise ring 142, respectively. When transmitting data from the node 1 to the node 4, the clockwise ring 141 is selected according to the shortest route distribution rule, and the nodes 1 → 2 → 3
→ Data packets are transferred in the order of 4. Further, the data packet is processed in accordance with the rule of destination removal, and is therefore removed from the clockwise ring 141 at the destination node 4.

On the other hand, when data is transmitted from the node 1 to the node 6, the counterclockwise ring 142 is selected, and data packets are transferred in the order of the nodes 1 → 8 → 7 → 6. Removed from ring 142.

As described above, in a ring network having a double ring configuration using the SRP, a data packet does not go around the ring unnecessarily, is delivered to the destination node via the shortest path, and is then removed from the ring at the destination node. Therefore, there is an advantage that the band of the ring can be used efficiently.

FIGS. 15 and 16 show a typical configuration of a queue buffer and a packet processing procedure in such a ring network node. The configuration shown here is 3
This is an example of a node that constitutes a ring network using optical signals that have been heavily wavelength-multiplexed as carriers, and shows the configuration of node 1 as a representative. When a data packet composed of an optical signal is received at the node 1 (step 1).
601) First, a received packet composed of the optical signal is
By the demultiplexer (DEMUX) 151, three wavelengths λ1
To λ3. The optical signal of each wavelength is converted into an electric signal by an optical / electrical converter (O / E) 152. It is determined whether the data packet converted into the electric signal is a packet to be received by the present node (step 1602), and when the destination is the present node 1,
A receiving process is performed (step 1603).

However, if the destination is another node, it is determined whether or not there is a free space in the passage queue buffer 153 (step 1604). Stored in buffer 153 (step 1605), FIFO (Fast In Fast out)
Are sequentially converted into optical signals by the electrical / optical converter 154 (step 1606), and the multiplexer (M
(UX) 155, and multiplexed (wavelength multiplexed), and transmitted toward downstream nodes (step 160).
7). However, if there is no free space in the passage queue buffer 153, the packet is discarded (step 16).
08).

[0008] In this way, the received data packets are classified into packets to be received by the node 1 and packets to be transmitted to further downstream nodes, and then received by the node. , Sent to downstream nodes. In this case, a packet that arrives at the node 1 from the network outside the ring and is generated on the ring at the node 1 is also stored in the passage queue buffer 153, and transmitted to the downstream node through the same procedure. Is done.

[0009] However, when a ring network is constituted by nodes including such queue buffers, there is an essential problem that it is difficult to realize fair access from the nodes to the ring. For example, this node 1
Consider a case where a large amount of data packets are transmitted for a long time from an upstream node to a downstream node. In such a situation, the queue buffer 153 of this node 1 is occupied by data packets going from upstream to downstream for a long time. As a result, even if an attempt is made to transmit a data packet from the node 1 to the ring, the queue buffer 153 is full, so that the data packet generated at the node 1 is discarded and cannot be transmitted to the downstream node on the ring. That situation will continue for a long time.

The problem caused by the occupation of the queue buffer 153 by the data packet from the upstream node is shown in FIG.
This will be described in more detail with reference to FIG. Here, a situation in which traffic from the computer 2 connected to the node 2 to the computer 7 connected to the node 7 and traffic from the computer 6 connected to the node 6 to the computer 1 connected to the node 1 exist simultaneously. think of. In the ring network to which the SRP is applied, these traffics are respectively distributed according to the route of 2 → 1 → 8 → 7 and 6 → 7 → 8 → 1. In a situation where both of these traffics occupy the ring bandwidth for a long time, the queue buffer of the node 8 is a queue buffer for the counterclockwise ring,
Both the clock buffers for the clockwise ring are occupied by the data packets of the traffic. As a result, as described above, the computer 8 connected to the node 8 cannot send a data packet to either the clockwise or counterclockwise ring as long as the queue buffer of the node 8 is occupied. That is, the situation that the ring cannot be accessed continues for a long time. As described above, in the optical ring network using the SRP, there is a problem of access unfairness that a node that cannot access the ring occurs depending on a traffic pattern flowing on the ring.

[0011]

As described above, as described above,
In a conventional ring network composed of frequency-multiplexed carriers, it is difficult to maintain fair access to the ring, such as some nodes that cannot send data packets to the ring for a long time depending on the traffic on the ring. The essential problem exists. An object of the present invention is to provide a communication control method capable of realizing fair access to a network in a node such as a ring network including frequency-multiplexed carriers.

[0012]

In order to achieve the above object, the present invention provides a communication control method in a node of a communication network in which carriers are frequency-multiplexed, wherein a carrier for transmitting a packet generated in the own node is prioritized. It is characterized by the fact that it is allocated on a regular basis.

In a communication control method in a node of a communication network in which a carrier is frequency-multiplexed, it is determined whether a received packet is a packet generated by the own node or a packet passing through the own node, and generated by the own node. If the packet has passed, the packet is stored in the queue buffer dedicated to the own node. If the packet passes through the node, the packet is stored in the passing queue buffer. The packet stored in each queue buffer is transferred to the carrier corresponding to each queue buffer. Is transmitted.

In a communication control method in a node of a communication network in which a carrier is frequency-multiplexed, it is determined whether a received packet is a packet generated in the own node or a packet passing through the own node. If it is a packet that has passed, it is stored in the queue buffer dedicated to its own node, if it is a packet that passes through its own node, it is stored in a queue buffer for passing, and the packet stored in the queue buffer dedicated to its own node is stored in the queue buffer. The packet transmitted on the corresponding first carrier and stored in the passage queue buffer determines whether or not the first carrier is idle, and if it is idle, transmits on the first carrier. If not empty, transmission is performed on another carrier corresponding to the passage queue buffer. Further, the processing for determining whether or not the first carrier is idle is performed at a predetermined frequency.

In a communication control method in a node of a communication network in which a carrier is frequency-multiplexed, it is determined whether a received packet is a packet generated in the own node or a packet passing through the own node, and the generated packet is generated in the own node. If the priority packet is stored in a priority area of a specific queue buffer among a plurality of queue buffers, and if the priority area is not empty, the packet is stored in a non-priority area of another queue buffer and passed through the own node. If the packet is to be stored in the priority area of the specific queue buffer, it is determined whether the packet is to be stored in the priority area of the specific queue buffer. Store in the priority area of the specific queue buffer, and if the priority area is not empty, the priority of the other queue buffer is If the packet does not need to be stored in the priority area of the specific queue buffer, the packet is stored in the non-priority area of the other queue buffer, and the packet stored in each queue buffer is stored in each queue buffer. The transmission is performed on a corresponding carrier.

According to such a configuration, in a node constituting a network composed of frequency-multiplexed carriers, a packet generated in the own node and a packet passing through the own node are processed separately. ,
The problem of unfair access to a ring network or the like can be solved.

[0017]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described in detail based on embodiments shown in the drawings.

(First Embodiment) FIG. 1 is a diagram showing one node configuration of a ring network configuration to which the communication control method of the present invention is applied, and FIG. 2 is a diagram showing a flowchart of a packet processing procedure.

Here, as in the example shown in FIG. 15, a case will be described as an example where the input signal is triple-multiplexed. The difference from the configuration of the conventional queue buffer shown in FIG. 15 is that a queue buffer 101 that can use data packets generated in the node 1 preferentially is provided independently of a passage queue buffer. When transmitting the data packet generated in the above, the carrier of the wavelength λ1 can be preferentially occupied. The optical signal propagating through the ring has three wavelengths, for example, 155.33 nm (193 OTH).
It is multiplexed from z) to the longer wavelength side at 100 GHz intervals. The same or corresponding parts as those in FIG. 15 are denoted by the same reference numerals.

According to such a node configuration, a data packet passing through the present node 1 from the upstream node to the downstream node is stored in the passage queue buffer 153, and is sequentially converted into a wavelength-multiplexed signal. Sent to downstream nodes.

On the other hand, the data packet generated in the node 1 is stored in the unique queue buffer 101 independent of the passage queue buffer 153. Then, the signal is transmitted to the downstream node using one wavelength channel (λ1) of the wavelength-multiplexed optical signal.

In FIG. 2, when a data packet composed of an optical signal is received at the node 1 (step 20).
1) First, a received packet composed of the optical signal is divided into three wavelengths λ1 to λ by a demultiplexer (DEMUX) 151.
It is split into three. The optical signal of each wavelength is converted into an electric signal by an optical / electrical converter (O / E) 152.
It is determined whether or not the data packet converted into an electric signal is a packet generated by the node 1 (step 202). If the data packet is not a packet generated by the node, a packet to be received next by the node is determined. It is determined whether or not there is (step 203). As a result, the destination is this node 1
, A reception process is performed (step 20).
8).

However, if the destination is another node, it is determined whether or not there is a free space in the passage queue buffer 153 (step 204). If there is a free space, the queue buffer 153 is temporarily stored in the queue buffer 153. Are stored (step 205), and the wavelengths λ1 and λ are sequentially converted by the electrical / optical converter 154 in accordance with the principle of FIFO (Fast In Fast out).
After being converted into two optical signals (step 206), they are multiplexed (wavelength multiplexed) by a multiplexer (MUX) 155 and transmitted to a downstream node (step 2).
07). However, if there is no free space in the passage queue buffer 153, the packet is discarded (step 2).
09).

On the other hand, if it is determined in step 202 that the packet is generated by the node, it is determined whether or not the queue buffer 101 dedicated to the node has a free space (step 210). If there is space,
The packet is stored in the queue buffer 101 dedicated to the node (step 211). The stored packet is
After being sequentially converted into an optical signal of wavelength λ1 by the electrical / optical converter 154 (step 212), the multiplexer (MUX)
The signals are multiplexed (wavelength multiplexed) by 155 and sent to downstream nodes (step 207). However, if there is no free space in the queue buffer 101, the packet is discarded (step 213).

As described above, the carrier (wavelength λ1) for transmitting the data packet generated in the node is preferentially assigned, so that a large amount of data packet flows from the upstream node for a long time. However, a situation does not occur in which the queue buffer 101 for storing the data packet generated in the node 1 continues to be occupied.

Further, since the wavelength channel of λ1 is preferentially reserved for transmitting the data packet generated in the node 1, even if a large number of passing packets exist, all the bands of the transmission signal are transmitted. , It is not occupied for sending out passing packets.

Therefore, according to the communication control method of the present embodiment, since a queue buffer for storing data packets generated in each node and a part of the band of the transmission side channel are secured, Since data packets can be sent out on the ring network even in a pattern situation, fair access to the ring network can be realized while taking advantage of the SRP.

(Second Embodiment) FIG. 3 is a configuration diagram of a node showing a second embodiment of the present invention, and FIG. 4 is a diagram showing a flowchart of a packet processing procedure. The nodes shown here are nodes that constitute a ring network using triple wavelength multiplexed optical signals as carriers.
The optical signal propagating through the ring has three wavelengths, for example, 155
It is multiplexed from 3.33 nm (193.OTHz) to the longer wavelength side at 100 GHz intervals.

The difference between the second embodiment and the first embodiment is that the wavelength channel (λ
If there is a vacancy in 1), it is also used for transmitting packets for passage.

The optical signal received by the node 1 is a DEM
The signal is demultiplexed by the UX 151 and classified into packets received by the node 1 and packets transmitted to the downstream nodes for each wavelength channel. The packet transmitted to the downstream node is stored in the passage queue buffer 153 in the figure.

On the other hand, the packet generated at the node 1 is
A queue buffer 10 different from the queue buffer 153
1 is stored. If the queue buffer 101 is occupied by a packet generated by the node 1,
These packets are transmitted to the downstream node by the channel of the wavelength λ1. Queue buffer 153 for passage
Are transmitted to the downstream nodes by the wavelength channels of λ2 and λ3 in principle.

The above operation is the same as that of the first embodiment. However, in this embodiment, if there is a free space in the queue buffer 101 of the wavelength λ1, in order to effectively utilize the band of the ring, The mechanism is such that a passing packet can be transmitted using the channel of the wavelength λ1. The pass-through packet is sent out to the downstream node through the wavelength channels of λ2 and λ3 in principle. However, for example, an opportunity is provided to search the state of the channel of wavelength λ1 at a frequency determined by the weighted cyclic method, and if the channel of wavelength λ1 is free, the channel of wavelength λ1 is transmitted to the downstream node. You. This is to prevent the band of the channel of the wavelength λ1 from being wasted when the number of data packets generated in the node 1 is very small.

The details of the operation will be described with reference to the flowchart of FIG. In FIG. 4, when a data packet composed of an optical signal is received at the node 1 (step 40).
1) First, a received packet composed of the optical signal is divided into three wavelengths λ1 to λ by a demultiplexer (DEMUX) 151.
It is split into three. The optical signal of each wavelength is converted into an electric signal by an optical / electrical converter (O / E) 152.
It is determined whether or not the data packet converted into the electric signal is a packet generated in the node 1 (step 402). If the data packet is not a packet generated in the node, a packet to be received next by the node is determined. It is determined whether there is (step 403). As a result, the destination is this node 1
, A reception process is performed (step 41).
0).

However, if the destination is another node, it is determined whether there is a free space in the passage queue buffer 153 (step 404). If there is a free space, the queue buffer 153 is temporarily stored in the queue buffer 153. It is stored (step 405).

Next, the queue buffer 10 dedicated to this node
It is determined whether or not there is a free space in step 1 (step 406). If there is a free space, the packets stored in the passage queue buffer 153 are sequentially transmitted in accordance with the principle of FIFO (Fast In Fast out). The wavelength λ1,
After being converted into an optical signal of either λ2 or λ3 (step 409), it is multiplexed (wavelength multiplexed) by a multiplexer (MUX) 155 and transmitted to a downstream node (step 408). . However, if there is no free space in the queue buffer 101 dedicated to this node (when λ1 is occupied exclusively for this node), the packet stored in the queue buffer 153 for passing is transmitted to the electrical / optical converter 154.
After being converted into an optical signal having the wavelength λ1 or λ2 (step 407), the optical signal is multiplexed (wavelength multiplexed) by the multiplexer (MUX) 155 and transmitted to the downstream node (step 408).

Here, the processing of step 406 is executed at a predetermined frequency by, for example, a weighted cyclic method.
Therefore, the process of step 406 is not performed every time a passing packet is received. When the opportunity to execute the process of step 406 is given at the timing determined by the weighted cyclic method, if there is a free space in the queue buffer 101 dedicated to this node, even if the packet is a pass-through packet, the channel of the wavelength λ1 is used. Sent. on the other hand,
In step 404, the queue buffer 15 for passing
If there is no free space in No. 3, the packet is discarded (step 411).

On the other hand, if it is determined in step 402 that the packet has occurred in the present node, it is determined whether or not the queue buffer 101 dedicated to the present node has a free space (step 412). If there is space,
The packet is stored in the queue buffer 101 dedicated to this node (step 413). The stored packet is
After being sequentially converted into an optical signal of wavelength λ1 by the electrical / optical converter 154 (step 414), the multiplexer (MUX)
The signals are multiplexed (wavelength multiplexed) by 155 and sent to downstream nodes (step 408). However, if there is no free space in the queue buffer 101, the packet is discarded (step 415).

With such a node configuration,
Approximately 1/3 of the bandwidth of the transmitting ring can be reserved for transmitting packets generated by this node 1, so that data packets can be transmitted from this node 1 onto the ring regardless of the traffic pattern on the ring. And fairness of access can be realized.

(Third Embodiment) FIG. 5 is a node configuration diagram showing a third embodiment of the present invention, and FIGS. 6 to 8 are flowcharts showing a packet processing procedure. Here, FIG.
The node shown in FIG. 1 is applied to the node 4 of the triple-wavelength-multiplexed optical ring network as shown in FIG. The node 4 of the present embodiment includes separate queue buffers 156-1, 156-2, and 156-3 each having a priority area A and a non-priority area B for each wavelength channel. , The queue buffer 156-1 of the channel of the wavelength λ1 can be used preferentially, while the queue buffer 156-2 of the other wavelength channel can be used preferentially.
When there is a free space in 156-3, these non-priority areas B can be used. With this configuration, a band corresponding to the priority area A of the queue buffer 156-1 of the channel of the wavelength λ1 is secured for packets generated in the node 4, and at the same time, the band of the entire ring is effectively used. Becomes possible. In FIG. 5, the priority area A is indicated by vertical, horizontal, and oblique hatching.

In the present embodiment, in the ring network configuration of FIG.
Queue buffers 156-1 to 156 for 1, λ2 and λ3
3 priority area can be used, and the other nodes
The queue buffer 15 shown in FIG.
It is assumed that allocation using the non-priority areas B of 6-1 to 156-3 is performed. This is because the nodes 4, 6, and 8 in FIG.
This is because there is a high possibility that a large amount of data packets will flow into the ring network 904 via these nodes 4, 6, and 8, as compared with other nodes.

Hereinafter, the processing procedure of the packet will be described in detail with reference to the flowcharts of FIGS. In FIG. 6, when a data packet composed of an optical signal is received at the node 4 (step 601), the received packet composed of the optical signal is first divided into three wavelengths λ1 to λ3 by a demultiplexer (DEMUX) 151. It is split. And
The optical signal of each wavelength is converted into an electrical signal by an optical / electrical converter (O / E) 152. It is determined whether or not the data packet converted into the electric signal is a packet generated in the present node 4 (step 602). It is determined whether the packet is a packet (step 60).
3). As a result, if the destination is the present node 4, a reception process is performed (step 607).

However, when the destination is another node, the processing of the packet passing through the node 4 is performed (step 604), and thereafter, the electrical / optical converter 154 is sequentially performed.
Is converted into an optical signal of one of the wavelengths λ1, λ2, λ3 (step 605), multiplexed (wavelength multiplexed) by the multiplexer (MUX) 155, and transmitted to the downstream node (step 605). 606).

On the other hand, if it is determined in step 602 that the packet is generated by the node 4, the packet generated by the node 4 is processed (step 608), and then the electrical / optical converters 154 are sequentially processed. By the wavelength λ
1 (step 609), and the multiplexer (M
(UX) 155, and multiplexed (wavelength multiplexed), and transmitted toward a downstream node (step 606).

FIG. 7 is a flowchart showing the details of step 608. First, it is determined whether or not there is a free space in the priority area A of the queue buffer 156-1 corresponding to channel 1 (the wavelength channel of λ1) ( Step 701) If there is a free space, the packet generated in the own node 4 is stored in the priority area A of the queue buffer 156-1 (Step 702). The stored packet is converted into an optical signal of λ1 in step 609 of FIG.

However, if there is no free space in the priority area A of the queue buffer 156-1 corresponding to channel 1, then it is determined whether there is free space in the non-priority area B of the queue buffer 156-2 corresponding to channel 2. (Step 7
03), if there is a vacancy, the queue buffer 156-2
The packet generated in the own node is stored in the non-priority area B (step 704). The stored packet is converted into an optical signal of λ1 in step 609 of FIG.

If there is no free space in the non-priority area B of the queue buffer 156-2 corresponding to the channel 2, then it is determined whether there is a free space in the non-priority area B of the queue buffer 156-3 corresponding to the channel 3. (Step 7)
05), if there is free space, the corresponding queue buffer 156-3
The packet generated in the own node 4 is stored in the non-priority area B (step 706). The stored packet is converted into an optical signal of λ1 in step 609 of FIG. Queue buffer 156 corresponding to channel 3
If there is no free space in the non-priority area B of -3, the packet is discarded (step 707).

Therefore, in this embodiment, if the channel 1 of the wavelength λ1 is vacant, the packet generated at the own node is transmitted using this channel 1 with priority. However, when channel 1 of wavelength λ1 is not vacant, transmission is performed using non-priority area B of queue buffer 156-2 of channel 2 or non-priority area B of queue buffer 156-3 of channel 3.

On the other hand, if the packet passes through the node 4, as shown in FIG.
It is determined whether the packet is to be stored in the priority area A of 6-1 to 156-3 (step 801). The packets to be stored in the priority area A of the queue buffers 156-1 to 156-3 are the packets transmitted from the nodes 4, 6, and 8 in FIG. It is.

If the packet is transmitted from the nodes 4, 6 and 8, all the queue buffers 156-1 to 15-15
It is determined whether there is an empty space in the priority area A of 6-3 (step 802).
The packet is stored in the priority area A of 6-1 to 156-3 (step 803). However, if there is no free space, it is determined whether or not there is free space in the non-priority area B of the queue buffers 156-1 to 156-3 (step 804). The packet is stored in the third non-priority area B (step 805). If there is no free space, the packet is discarded (step 80).
6).

If it is determined in step 801 that the packet is not a packet to be stored in the priority area (that is, if it is a packet transmitted from a node other than the nodes 4, 6, and 8 in FIG. 9), the queue buffers 156-1 to 156 It is determined whether or not there is a free space in the non-priority area B of -3 (step 80)
7) If there is a space, the queue buffer 156-1
The packet is stored in the non-priority area B of 156-3 (step 808). If there is no space, the packet is discarded (step 809).

The packets stored in the queue buffers 156-1 to 156-3 are converted into optical signals of any of the wavelengths λ1, λ2, λ3 in step 605 of FIG.

FIG. 10 shows the throughput of each node in such a ring network.
11 shows the throughput of the nodes 7 and 5 in FIG. 10 individually, FIG. 12 shows the throughput of the nodes 8 and 6 in FIG. 10, and FIG.
7 individually shows the throughput of the node 4 at 0.

In this example, first, from node 7 to node 3
Normalization time 1 in the presence of traffic to
At 0, traffic from node 5 to node 1 is occurring. At this time, both node 7 and node 5
Since the node has the conventional configuration, the throughput of the traffic from the node 7 which originally occupies the band decreases with the generation of the traffic from the node 5, and both nodes share the band of the ring.

Further, under such a situation, at the standardized time 22, traffic from the node 8 to the node 4 occurs. In the node of the conventional configuration, since the queue buffer of the node 8 is occupied by the traffic from the nodes 7 and 5, it is necessary to wait a long time before transmitting the data packet from the node 8 onto the ring. In the embodiment, since the node having the queue buffer configuration of the present invention is adopted as the node 8, a band corresponding to the priority area in the queue buffer of the channel of the wavelength λ1 is secured, and access fairness is secured. Understand.

The figure further shows the change in throughput when traffic from node 6 to node 2 and traffic from node 4 to node 8 occur at normalization time 32 and normalization time 42. I have. Since both the nodes 6 and 4 use the node configuration of the present invention, it can be seen that access fairness is ensured regardless of the traffic pattern. In addition, by adopting the node configuration shown in FIG. 5, the bandwidth for the nodes 8, 6, and 4 can be secured throughout the ring, so that even if new traffic occurs in the ring, the effect on the throughput is affected. Is decreasing.

In this embodiment, the node configuration of the present invention is changed to a part (node 4,
6, 8), but using the appropriate number of wavelength multiplexing,
If the node configuration of the present invention is applied to all the nodes constituting the ring, it goes without saying that the fairness of access can be maintained over the entire ring.

[0057]

As described above, according to the communication control method of the present invention, since the packet generated at the own node and the packet passing through the own node are processed separately, Packet can be sent out on the network under the
While making use of the merits of P, fair access to the network can be realized, and a network satisfying such fair access can be easily constructed.

[Brief description of the drawings]

FIG. 1 is a diagram showing a first embodiment of a configuration of a queue buffer in a node for realizing a communication control method of the present invention.

FIG. 2 is a flowchart showing a packet processing procedure in the configuration of FIG. 1;

FIG. 3 is a diagram showing a second embodiment of a configuration of a queue buffer in a node for realizing the communication control method of the present invention.

FIG. 4 is a flowchart illustrating a packet processing procedure in the configuration of FIG. 3;

FIG. 5 is a diagram showing a third embodiment of a configuration of a queue buffer in a node for realizing the communication control method of the present invention.

FIG. 6 is a flowchart showing a packet processing procedure in the configuration of FIG. 5;

FIG. 7 is a flowchart showing in detail a procedure of processing a packet generated in the own node in the configuration of FIG. 5;

FIG. 8 is a flowchart showing in detail a procedure of processing a packet passing through the own node in the configuration of FIG. 5;

FIG. 9 is a diagram schematically illustrating a configuration of an optical ring network to which a node according to a second embodiment of the present invention is applied.

FIG. 10 is a diagram illustrating a change in throughput according to the second embodiment of the present invention.

11 is a diagram individually showing a change in throughput of nodes 7 and 5 in FIG. 10;

FIG. 12 is a diagram individually showing a change in throughput of nodes 8 and 6 in FIG. 10;

FIG. 13 is a diagram individually illustrating a change in the throughput of the node 4 in FIG. 10;

FIG. 14 is a diagram schematically showing a configuration of a conventional ring network.

FIG. 15 is a diagram schematically showing a configuration of a queue buffer in a conventional node.

FIG. 16 is a flowchart of a packet processing procedure in a conventional communication control method.

FIG. 17 is a diagram showing an example of a traffic pattern that causes access unfairness in a conventional ring network.

[Explanation of symbols]

101: queue buffer dedicated to own node, 151: duplexer, 152: optical / electrical converter, 153: queue buffer, 154: electrical / optical converter, 156-1 to 156-3
... the queue buffer.

──────────────────────────────────────────────────の Continued on the front page (51) Int.Cl. ⁷ Identification FI FI Theme Court ゛ (Reference) H04L 12/56 (72) Inventor Takuo Hirono 2-3-1 Otemachi, Chiyoda-ku, Tokyo Nippon Telegraph and Telephone Telephone Co., Ltd. F-term (reference) 5K002 AA05 BA05 DA02 DA05 DA11 FA01 5K022 AA09 AA31 5K030 GA02 HA08 HB28 HC14 JL03 KX11 KX29 LA03 LA17 5K031 AA01 AA05 CA15 CB01 CC03 DA12 DA19 9A001 BB04 CC07 DD10 JJ18 KK56

Claims (6)

[Claims] 1. A communication control method in a node of a communication network in which carriers are frequency-multiplexed, wherein a carrier for transmitting a packet generated in the own node is preferentially assigned. 2. A communication control method in a node of a communication network in which carriers are frequency-multiplexed, wherein it is determined whether a received packet is a packet generated by the own node or a packet passing through the own node, and generated by the own node. If it is a packet that has passed, it is stored in a queue buffer dedicated to its own node. If it is a packet that passes through its node, it is stored in a queue buffer for passing. The packets stored in each queue buffer are stored in the carrier corresponding to each queue buffer. A communication control method, characterized by transmitting by: 3. A communication control method in a node of a communication network in which carriers are frequency-multiplexed, wherein it is determined whether a received packet is a packet generated by the own node or a packet passing through the own node, and generated by the own node. If it is a packet that has passed, it is stored in the queue buffer dedicated to its own node, if it is a packet that passes through its own node, it is stored in a queue buffer for passing.
The packet transmitted on the first carrier corresponding to the queue buffer and stored in the passage queue buffer determines whether or not the first carrier is idle. A communication control method, wherein the transmission is performed on a carrier and, if not empty, the transmission is performed on another carrier corresponding to the passage queue buffer. 4. The communication control method according to claim 3, wherein the process of determining whether the first carrier is idle is performed at a predetermined frequency. 5. A communication control method in a node of a communication network in which a carrier is frequency-multiplexed, wherein it is determined whether a received packet is a packet generated by the own node or a packet passing through the own node, and generated by the own node. If the priority packet is not empty, the packet is stored in a non-priority area of another queue buffer and passed through the own node if the priority area is not empty. If the packet is to be stored in the priority area of the specific queue buffer, it is determined whether the packet is to be stored in the priority area of the specific queue buffer. Store in the priority area of the specific queue buffer, and if the priority area is not empty, a non-priority area of another queue buffer If the packet does not need to be stored in the priority area of the specific queue buffer, it is stored in the non-priority area of the other queue buffer, and the packet stored in each queue buffer corresponds to each queue buffer. A communication control method characterized by transmitting by a carrier. 6. The communication control method according to claim 1, wherein the carrier is an optical signal.