JP2004297178A - Network traffic control method and apparatus thereof - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a network traffic control method and an apparatus thereof capable of preventing excess wavelength from being used for congestion avoidance and reducing the hardware amount in a hybrid node. <P>SOLUTION: An optical label path and an electric label path passing through an optical node and an electric node configuring the hybrid node existing on a path from an ingress node to an egress node are set, the electric node of the hybrid node aggregates packets, and on the occurrence of congestion in the electric node of the hybrid node, the electric label path causing the congestion is particularized and a new optical label path and a new electric label path are set and selected so that the electric label path causing the congestion bypasses the electric node of the hybrid node having caused the congestion. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a network traffic control method and a device thereof, and more particularly, to a network traffic control method and a device thereof in a network system in which routing and signaling are performed autonomously by nodes.
[0002]
[Prior art]
2. Description of the Related Art Conventionally, the following standardization organizations have proposed standardization of automatic line setting in an optical network.
[0003]
1. ASON standardized by ITU-T (such as ITU-T recommendation G.8080)
2. GMPLS signaling standardized by IETF
3. UNI signaling standardized by OIF
As shown in the network architecture of FIG. 1, the standardization of these automatic line settings is provided with an optical interface such as WDM (Wavelength Division Multiplexing) to convert the wavelength of an input port and the wavelength of an output port into different wavelengths. OXC (Optical Cross Connect: Optical Cross Connect), SONET (Synchronous Optical Network), SDH (Synchronous Digital Hierarchy (SDH), ADM (Add-Drop Transport Equipment), etc. Equipped with a base control plane, on which is currently done on the Internet Performed so that autonomous routing, in which using the signaling to set the line, such as wavelength and SONET / SDH.
[0004]
As shown in FIG. 1, in a framework for controlling a transport device such as an OXC using an IP-based control plane, various organizations are currently standardizing the same. As a feature of such a network, since various transmission media (fiber, wavelength, SONET / SDH, MPLS, etc.) can be controlled by a single control plane, control is conventionally performed separately for each layer of the transport plane, As a result, it is possible to solve the inefficient control.
[0005]
For example, if a control layer for setting a wavelength path and a layer for controlling an MPLS (Multiprotocol Label Switching) path are separated, it is unknown from the MPLS layer what topology the wavelength layer has, and vice versa. That's right. Therefore, in some cases, an optical label path is set every time an MPLS label path is set.
[0006]
On the other hand, when both control planes are a single layer, no matter how many MPLS label paths are set, if the MPLS label paths with the same destination are accommodated in the same optical label path, the resources of the optical label path can be reduced. become. Conventionally, such a path setting has been manually set using a network management system.
[0007]
That is, in the prior art, there was no cooperation between the layers of the transport plane, so that inefficient resource management was performed. Also, even if the layers were managed in an integrated manner by the network management system, the paths had to be set manually. In order to solve such problems, each research institution has proposed a label path management method (for example, see Non-Patent Documents 1, 2, and 3).
[0008]
[Non-patent document 1]
S. Okamoto, K .; Shimano, and A .; Sahara, "GMPLS based IP over Photonic Networks Complicating HIKARI (Photonic MPLS) Routers", PS + COIN2002, COIN. WeA2, July 2002.
[0009]
[Non-patent document 2]
Shiomoto et al., "Dynamic Multilayer Traffic Engineering in MPLAmbdaS Networks", IEICE PS2002-10, April 2002.
[0010]
[Non-Patent Document 3]
Kano et al., “Proposal of L1 / L2 Cooperative Control Realization Method Using GMPLS”, IEICE, No. SB-13-2 pp. 576-577, Sept. 2001
[0011]
[Problems to be solved by the invention]
Any of the methods described in Non-Patent Documents 1, 2, and 3 is intended to avoid the traffic congestion of the electric switch by the shortcut of the optical label path by combining the optical switch such as OXC and the electric switch. In particular, Non-Patent Document 2 describes a detailed control algorithm, but has the following problems.
(1) When congestion occurs in an electric node, a new optical label path is set end-to-end to avoid congestion, so that extra optical resources are used.
(2) Since the traffic flow of all the electric label paths must be measured at each electric node, the amount of hardware for that purpose becomes enormous.
(3) When the optical resources (the number of wavelengths) are large, there is a possibility that the traffic will be forced through an electric node, and therefore, the service quality of the traffic may be degraded.
[0012]
The present invention has been made in view of the above points, it is possible to prevent the use of an extra wavelength to avoid congestion, can reduce the amount of hardware in the hybrid node, and, when there is a lot of optical resources It is an object of the present invention to provide a network traffic control method and device capable of preventing deterioration of traffic service quality.
[0013]
[Means for Solving the Problems]
According to the present invention, when an ingress node and an egress node for which a path is to be set are specified, an optical node and an electric node constituting a hybrid node existing on a route from the ingress node to the egress node are designated. Set the optical label path and electrical label path that pass through the node,
Aggregation of packets in the electric node of the hybrid node, when congestion occurs in the electric node of the hybrid node, to identify the electric label path causing the congestion,
By setting and switching a new optical label path and a new electric label path so that the electric label path causing the congestion bypasses the electric node of the congested hybrid node,
A new optical label path that bypasses only the electric node of the hybrid node in which congestion has occurred can be set, and use of an extra wavelength for avoiding congestion can be prevented.
[0014]
The invention according to claims 2 and 5 is characterized in that, when congestion occurs in an electric node of a hybrid node, an input node terminating an electric label path passing through the congested output port among output ports of the electric node. Query the traffic flow rate of the electrical label path,
By identifying the electric label path causing the congestion from the response to the inquiry at the electric node of the hybrid node,
In the hybrid node, it is not necessary to measure the traffic flow of all the electric label paths, and the amount of hardware for measuring the traffic flow can be reduced.
[0015]
According to a third aspect of the present invention, when an ingress node and an egress node for which a path is to be set are designated, light is transmitted without passing through an electric node constituting a hybrid node existing on a path from the ingress node to the egress node. Set the optical label path and electric label path that pass only through the node,
When an ingress node and an egress node that intend to set a new path in a state where the optical resources are depleted are specified, the hybrid path close to the hybrid node near the egress node from the ingress node of the new path is shared with the existing path,
By separating the existing path into two at the electric node of the hybrid node and setting and switching a new optical label path and a new electric label path to be connected to the egress node of the existing path and the egress node of the new path, ,
When there is a lot of optical resources, an optical label path can be set up without passing through an electric node, thereby preventing the degradation of traffic service quality.
[0016]
BEST MODE FOR CARRYING OUT THE INVENTION
FIG. 2 shows a configuration diagram of a basic network to which the network traffic control method of the present invention is applied. In the figure, e1 to e7 in the data plane indicate electric nodes that perform switching for each packet of the electric signal, o1 to o5 indicate optical nodes that perform switching for each wavelength of the optical signal, and h1 to h4 indicate electric nodes. And a hybrid node including an optical node. User devices y1 to y7 are connected to the electric nodes e1 to e7.
[0017]
The control nodes c1 to c8 forming the control plane are connected to the electric nodes e1 and e2, the hybrid nodes h1 to h4, the electric node e7, and the optical node o5 forming the data plane, respectively. Is equipped with routing and signaling protocols. Further, a network management system NMS is connected to the control node c4.
[0018]
Here, the incoming node is an electric node, but is assumed to be equipped with an optical interface such as WDM. The relay node may be an optical node only or a hybrid node of an optical switch and an electric switch. A hybrid node may combine OXCs and electrical switches from different manufacturers. In practice, such a hybrid node is installed at a boundary point between a metro network in a city and a long-haul long-haul network.
[0019]
FIG. 3 shows a block diagram of one embodiment of an electrical node. In FIG. 1, an optical / electrical conversion / separation unit 10 separates a WDM signal supplied from an optical node for each wavelength, and converts each wavelength into an electric signal. Input IF 12 in input interface (IF) group 11 ₁ To 12n receive the electric signal supplied from the user device or the optical / electrical conversion / separation unit 10 and supply the same to the switch 13.
[0020]
The switch 13 is connected to each input IF 12 ₁ 12n is switched under the control of the node control unit 14, and the output IF 16 in the output interface (IF) group 15 is switched. ₁ ~ 16n. Output IF16 ₁ To 16n receive the electric signal supplied from the switch 13 and supply it to the user apparatus or the electric / optical conversion / multiplexing unit 17. The electrical / optical conversion / multiplexing unit 17 converts the electrical signals from the plurality of output IFs into optical signals of different wavelengths, multiplexes the signals, and transmits the multiplexed WDM signal to the optical node.
[0021]
The node control unit 14 is connected to a control node of the control plane via an external control interface (IF) 18. The node control unit 14 has a resource management unit 20, a routing unit 21, and a signaling unit 22 in addition to controlling the switching of the switch 13. The resource management unit 20 grasps the usage status and the congestion status of the node resources from the measured values such as the packet flow rate observed at the input IF and the output IF. The routing unit 21 automatically grasps the topology of the network. The signaling unit 22 sets a label path on the grasped topology. In this case, wavelengths, SONET / SDH channel numbers, MPLS labels, and the like are handled as labels.
[0022]
FIG. 4 is a block diagram showing an embodiment of the input / output IF unit of the electric node connected to the user equipment. The input / output IF unit includes an input IF 30 (for example, 12 ₁ And an output IF 35 (for example, 16 ₁ ). The packet processing unit 31 of the input IF 30 adds a GMPLS (Generalized MPLS) label managed by the label management unit 32 to a packet supplied as an electric signal from the user device, generates a GMPLS packet, and supplies the packet to the packet buffer unit 33. . The packet buffer 33 buffers the GMPLS packet supplied from the packet processor 31 and supplies the buffered GMPLS packet to the switch 13 under the control of the node controller 14. The traffic measuring unit 34 measures the traffic flow rate of the GMPLS packet output from the packet buffer unit 33 and the queue length of the packet buffer unit 33 and notifies the resource managing unit 20 in the node control unit 14 of the measurement.
[0023]
The packet buffer unit 36 of the output IF 35 buffers the GMPLS packet supplied from the switch 13 and supplies the buffered GMPLS packet to the packet processing unit 38 under the control of the node control unit 14. The traffic measurement unit 37 measures the traffic flow rate of the GMPLS packet output from the packet buffer unit 36 and the queue length of the packet buffer unit 36 and notifies the resource management unit 20 in the node control unit 14 of the measurement.
[0024]
The packet processing unit 38 extracts the label of the GMPLS packet, specifies the destination user device with reference to the management table of the label management unit 39 using this label, and sends the packet with the label removed to the destination user device.
[0025]
FIG. 5 shows a block diagram of one embodiment of the optical node. In the figure, an input interface (IF) 42 in an input interface (IF) group 41 is provided. ₁ To 42 m receive a WDM signal supplied from an adjacent optical node, separate the WDM signal for each wavelength, and supply it to the optical switch 43.
[0026]
The optical switch 43 is connected to each input IF 42 ₁ .. 42m under the control of the node control unit 44, and the output IF 46 in the output interface (IF) group 45. ₁ ~ 46m. Output IF 46 ₁ .About.46 m multiplexes the optical signals for each wavelength supplied from the optical switch 43 and outputs them to adjacent optical nodes as WDM signals.
[0027]
The node control unit 44 is connected to a control node of the control plane via an external control interface (IF) 48. The node control unit 44 has a resource management 50, a routing unit 51, and a signaling unit 52 in addition to controlling the switching of the optical switch 43. The resource management unit 50 grasps the usage status and the congestion status of the node resources. The routing unit 51 automatically grasps the topology of the network. The signaling unit 52 sets a label path on the grasped topology.
[0028]
FIG. 6 is a block diagram showing an embodiment of an input / output IF unit of an optical node connected to an adjacent optical node. The input / output IF unit includes an input IF 60 (for example, 42 ₁ And an output IF 65 (for example, 46 ₁ ). The optical amplifier / separation unit 61 of the input IF 60 amplifies the received WDM signal, separates it for each wavelength, and supplies it to the optical level monitoring unit 62. The optical level monitoring unit 62 performs monitoring control so that the optical level of each wavelength becomes a predetermined level, and supplies the optical switch 43.
[0029]
The transponder 66 of the output IF 65 converts the optical signal of each wavelength supplied from the optical switch 43 into an electric signal, and further converts the signal into an optical signal, thereby setting the optical level to a predetermined level to the optical amplifier / multiplexing / optical level monitoring unit 67. Supply. The optical amplifier / multiplexer / optical level monitor 67 amplifies the optical signal of each wavelength, multiplexes the optical signal, monitors and controls the optical level, and transmits it to an adjacent optical node.
[0030]
FIG. 7 shows a block diagram of one embodiment of the hybrid node. In the figure, the same parts as those in FIG. 3 or FIG. In FIG. 7, the optical / electrical conversion / separation unit 10 in the electric node unit 70 is connected to the output IF 46 of the optical node unit 72. ₁ Is separated for each wavelength, and each wavelength is converted into an electric signal. Input IF 12 in input interface (IF) group 11 ₁ To 12n receive the electric signal supplied from the user device or the optical / electrical conversion / separation unit 10 and supply the same to the switch 13.
[0031]
The switch 13 is connected to each input IF 12 ₁ 12n is switched under the control of the node control unit 14, and the output IF 16 in the output interface (IF) group 15 is switched. ₁ ~ 16n. Output IF16 ₁ To 16n receive the electric signal supplied from the switch 13 and supply it to the user apparatus or the electric / optical conversion / multiplexing unit 17. The electrical / optical conversion / multiplexing unit 17 converts the electrical signals from the plurality of output IFs into optical signals having different wavelengths, and multiplexes the converted optical signals into WDM signals to form an input IF 42 in the optical node unit 72. ₁ To supply. The electric node control unit 74 is connected to the hybrid node control unit 76, and has a resource management unit 20 that manages resources of the entire electric node unit.
[0032]
An input IF 42 constituting an input interface (IF) group 41 in the optical node unit 72 ₁ To 42 m receive a WDM signal supplied from the electrical / optical conversion / multiplexing unit 17 or an adjacent optical node, demultiplex the WDM signal for each wavelength, and supply it to the switch 43.
[0033]
The optical switch 43 is connected to each input IF 42 ₁ .. 42m under the control of the node control unit 44, and the output IF 46 in the output interface (IF) group 45. ₁ ~ 46m. Output IF 46 ₁ .About.46 m multiplexes the optical signal for each wavelength supplied from the optical switch 43 into a WDM signal and sends it to the optical / electrical conversion / separation unit 10 in the electric node unit 70 or an adjacent optical node. The optical node control unit 78 is connected to the hybrid node control unit 76, and has a resource management unit 50 that manages resources of the entire optical node unit.
[0034]
The hybrid node control unit 76 is connected to a control plane control node via an external control interface (IF) 79. The hybrid node control unit 76 includes a resource management unit 80, a routing unit 81, and a signaling unit 82. The resource management unit 80 grasps the usage status and the congestion status of the node resources. The routing unit 81 automatically grasps the topology of the network. The signaling unit 82 sets a label path on the grasped topology.
[0035]
FIG. 8 shows a flowchart of the path setting process in the first embodiment of the network traffic control method of the present invention. In this embodiment, the number of wavelengths is reduced as much as possible. It is assumed that the ingress node e1 and the egress node e7 are designated, and the ingress node e2 and the egress node e7 are designated in the basic network shown in FIG. By the way, this process may be executed under the control of the NMS in which the incoming node and the outgoing node are specified from the NMS of the control plane, or may be executed under the control of the incoming node where the path setting request has been issued.
[0036]
Step S1: SPF (Shortest Path Fast) calculation is performed to set an optical label path (including a wavelength number and the like) from the electric node e1 to the electric node e7.
[0037]
Step S2: If the passing node is a hybrid node, an optical label path is set so as to always pass through an electric switch in the hybrid node. This is shown in FIG. In the figure, the optical label path is indicated by a broken line.
[0038]
Step S3: An electric label path (including MPLS) is set on the path of the set optical label path.
[0039]
Step S4: If the passing node is a hybrid node, an electric label path is set so as to always pass through an electric switch. This is shown in FIG. In the figure, the electric label path is indicated by a thick solid line.
[0040]
Step S5: It is determined whether or not another incoming node and outgoing node are designated. If another path is designated, the process proceeds to step S5, and if no other path is designated, the process ends.
[0041]
Step S6: In order to set a path from the ingress node e2 to the egress node e7 different from the electric node e1, an SPF calculation is performed to calculate an optical label path from the electric node e2 to the electric node e7 in the same manner as in step S1.
[0042]
Step S7: Set an optical label path from the electric node e2 to the electric node e7. However, in the case where the optical label path from the ingress node to the outgoing node between the outgoing nodes h2, h1, h3, and o5 already exists to the electric outgoing node e7, which is the target outgoing node, the relay node has an electrical level at the electrical level. Aggregate to save optical label path resources. This is shown in FIG. In the figure, a new optical label path is indicated by a broken line.
[0043]
Step S8: An electric path is set from the electric node e2 to the electric node e7. This is shown in FIG. In the figure, the electric label path is indicated by a thick solid line.
[0044]
Thereafter, each time another incoming node and outgoing node are designated, the processing of steps S6 to S8 is repeated.
[0045]
FIG. 13 shows a flowchart of the congestion avoidance process in the first embodiment of the network traffic control method of the present invention. Here, it is assumed that congestion has occurred in the electric node e4 of the hybrid node h3.
[0046]
Step S10: The electric node e4 that has detected the congestion specifies an incoming node terminating the electric label path passing through the output IF (output port) where the congestion has occurred (in this case, e1, e2), It instructs the incoming node to notify the traffic flow of the electric label path. The occurrence of congestion is determined by the resource management unit 20 shown in FIG. ₁ The occurrence of congestion is confirmed from the smoothed value of the packet flow of ~ 16n.
[0047]
As a result, the electric nodes e1 and e2 corresponding to the congestion start measuring the traffic flow in the traffic measuring unit 34 when the instruction of the traffic flow notification is given, and after a certain period of time, the measured values are congested. Notify the existing electric node e4. This is shown in FIG.
[0048]
Step S11: When it is recognized that the traffic flow rate of the electric node e1 is 1 Gbit / s and the traffic flow rate of the electric node e2 is 5 Gbit / s based on the notification of the measurement values from the electric nodes e1 and e2, congestion occurs. The specified electric label path is a path from the electric node e2.
[0049]
Step S12: To avoid congestion, the congested electric node e4 is provided with a short path (detour) by an optical path so that the path from the electric node e2 does not pass through the congested electric node e4. The electric node e3 is instructed to set a new optical label path and a new electric label path toward the electric node e7 to switch the traffic.
[0050]
Step S13: The designated electric node e3 switches traffic after setting a new optical label path and a new electric label path. This is shown in FIG. In the figure, new optical label paths L1 and L2 and new electrical label path L3 are set.
[0051]
Step S14: After the traffic switching, the electric nodes e3 and e4 delete the optical label path and the electric label path that can be deleted.
[0052]
FIG. 16 shows a flowchart of the congestion end processing in the first embodiment of the network traffic control method of the present invention. Here, it is assumed that congestion has occurred at the electric node e4 of the hybrid node h3.
[0053]
Step S15: It is determined whether or not congestion has ended in the electric node e4 that has activated the switching control.
[0054]
Step S16: If it is determined that the congestion ends, the electric node e4 switches back the path. In that case, each output IF 16 ₁ The smoothed value of the packet flow rate in the range of ~ 16n is obtained, and the switchback is performed based on the smoothed value of the packet flow rate when it is confirmed that congestion does not occur again even if the switchback is performed. This is shown in FIG.
[0055]
The smoothed value P of the packet flow rate is expressed by equation (1) as a smoothed value P_old of the packet flow rate obtained last time, a current measured value (packet flow rate value) M, and a smoothing coefficient α (0 <α <1). .
[0056]
P = α × M + (1−α) × P_old (1)
Alternatively, a fixed time width may be defined, and a value obtained by dividing the sum of the measured values M during the preceding time period by the fixed time width by the fixed time width may be used as the smoothed value P of the packet flow rate. .
[0057]
The resource management unit 20 determines that congestion has occurred when the smoothed value P of the packet flow rate obtained by, for example, equation (1) exceeds a preset first threshold value a specified number of times, and the smoothed value P is set in advance. If the second threshold value (first threshold value> second threshold value) falls below a specified number of times, it is determined that congestion has ended.
[0058]
Note that the resource management unit 20 calculates the sum of the plurality of smoothed values P obtained within a predetermined time interval, and determines that the congestion occurs when the sum exceeds a third threshold (third threshold> first threshold). May fall below the fourth threshold value (third threshold value> fourth threshold value).
[0059]
Further, the resource management unit 20 obtains the time change rate ΔP of the smoothed value P of the packet flow obtained by the equation (1), and judges that the packet is congested when the time change rate ΔP exceeds a positive fifth threshold value (TH5). However, when the time rate of change ΔP falls below the negative sixth threshold value (−TH5), it may be determined that congestion has ended.
[0060]
When instructing the incoming node to notify the traffic flow in step S10, for example, a signaling message or a routing message is used.
[0061]
FIG. 18 shows a format of a signaling message using RSVP-TE. In the figure, following the length, which is the length of the signaling object, the Class-Num, which is the type of the object, and the C-Type, which is the detailed type of the object, an interface number, a label path identifier, the number of transmission bytes, the number of packet losses, and the bandwidth Etc. are set.
[0062]
FIG. 19 shows a format of a routing message using OSPF Opaque LSA. In the figure, LS age, Option, Opaque Type, Opaque ID, Advertising Router, LS Sequence Number, LS checksum, Length, followed by interface number, label path identifier, number of transmission bytes, number of packet loss, number of packet loss, etc.
[0063]
A case where the number of wavelengths can be used abundantly in a network will be described. In this case, a direct optical label path and an electric label path that do not pass through an electric node in the hybrid node, which is a relay node, are set between the end (ingress node) and the end (egress node), and high-quality data transfer is performed. In this case, the path setting process is performed under the control of the NMS, with the input node and the output node specified by the NMS on the control plane, but may be executed under the control of each input node.
[0064]
FIG. 20 is a flowchart of a path setting process when the number of wavelengths is depleted in the second embodiment of the network traffic control method of the present invention. Here, as shown in FIG. 21, there is an existing path from the ingress node e1 to the egress node e7 indicated by a thick solid line, and the number of wavelengths is depleted between the electrical node e1 and the optical node o1, and the ingress node e1 is in another state. It is assumed that the path setting of the output node e6 has been designated.
[0065]
Step S20: If the number of wavelengths is depleted between the electric node e1 and the optical node o1, for example, even if an attempt is made to set a path from the ingress node e1 to the egress node e6, the wavelength resources of the output port of the ingress node e1 are exhausted. , The optical label path cannot be set. In this case, the ingress node e1 is an existing path (a path from the ingress node e1 to the egress node e7) from the own device (e1) through the hybrid node h3 closest to the egress node e6, and is an outgoing path from the ingress node e1. Select the wavelength at the hybrid node h3 closest to the node e6, the optical label path has already been set from the ingress node e1 to the hybrid node h3, and the optical label path is free from the hybrid node h3 to the egress node e6. Then, the end of the optical label path and the end of the electric label path are instructed to the hybrid node h3 closest to the output node e6 in the existing path of the selected wavelength.
[0066]
Step S21: In the hybrid node h3, each switch is set as shown in FIG. 22, and switching is performed from a path close to the output port of the optical switch, thereby realizing instantaneous interruption switching.
[0067]
In step S21-1 (corresponding to the circled numeral 1 in FIG. 22), switching of the optical label path toward the optical nodes o4 and o5 is performed.
[0068]
In step S21-2 (corresponding to the number 2 with a circle), the electrical label path is separated and switched.
[0069]
Step S21-3 (corresponding to the number 3 with a circle) Switching of the optical label path and the electric label path during communication is performed.
[0070]
Step S21-4 (corresponding to the number 4 with a circle) An electric label path between the ingress node e1 and the egress node e6 is set.
[0071]
By performing the processing as described above, it becomes possible to set a new path without changing the wavelength resource depleted portion. This is shown in FIG. In the figure, a broken line indicates a newly set path.
[0072]
By the way, in the first embodiment, when the output port of the switch 13 of the electric node (for example, e10) which is the input node is congested due to traffic concentration, the congested path shown by the broken line in FIG. As shown by the thick solid line, the data is output from the vacant output port of the switch 13 and passes from the path passing through the nodes e10, h10, and e11 at the time of congestion to the path passing through the new paths of the nodes e10, h11, h12, and e11. Switching can also be used to avoid congestion.
[0073]
Steps S10 to S11 correspond to the specifying means described in the claims or appended claims, steps S12 to S13 correspond to the path setting switching means, step S10 corresponds to the inquiry means, and the traffic measuring unit 37 corresponds to the measuring means. Step S12 corresponds to the switching instruction unit, and the asset management unit 20 corresponds to the asset management unit.
[0074]
(Supplementary Note 1) A network that controls traffic of a network including an electric node that performs switching for each packet of an electric signal, an optical node that performs switching for each wavelength of an optical signal, and a hybrid node including the electric node and the optical node. In the traffic control method,
When an ingress node and an egress node for which a path is to be set are designated, an optical label path and an electric label path passing through an optical node and an electric node constituting a hybrid node existing on the path from the ingress node to the egress node are set. And
Aggregation of packets is performed at the electric node of the hybrid node,
When congestion occurs in the electric node of the hybrid node, identify the electric label path causing the congestion,
A network traffic control method, wherein a new optical label path and a new electric label path are set and switched such that the congested electric label path bypasses the congested hybrid node electric node.
[0075]
(Supplementary note 2) In the network traffic control method according to supplementary note 1,
When congestion occurs in the electric node of the hybrid node, query the traffic flow of the electric label path to the input node terminating the electric label path passing through the congested output port among the output ports of the electric node,
A network traffic control method, wherein an electric label path causing the congestion is identified from a response to the inquiry at an electric node of the hybrid node.
[0076]
(Supplementary Note 3) A network that controls traffic of a network including an electric node that performs switching for each packet of an electric signal, an optical node that performs switching for each wavelength of an optical signal, and a hybrid node including the electric node and the optical node. In the traffic control method,
When an ingress node and an egress node for which a path is to be set are specified, an optical label path and an electric signal that pass only through an optical node without passing through an electric node constituting a hybrid node existing on the path from the ingress node to the egress node Set the label path,
When an ingress node and an egress node that intend to set a new path in a state where the optical resources are depleted are specified, the hybrid path close to the hybrid node near the egress node from the ingress node of the new path is shared with the existing path,
Separating the existing path into two at the electric node of the hybrid node and setting and switching a new optical label path and a new electric label path to be connected to the egress node of the existing path and the egress node of the new path. Characteristic network traffic control method.
[0077]
(Supplementary Note 4) In a hybrid node device of a network including an electric node that performs switching for each packet of an electric signal, an optical node that performs switching for each wavelength of an optical signal, and a hybrid node including the electric node and the optical node,
When congestion occurs in the electric node of its own device, specifying means for specifying the electric label path causing the congestion,
Path setting switching means for setting and switching a new optical label path and a new electric label path so that the electric label path causing the congestion bypasses the electric node causing the congestion.
A hybrid node device comprising:
[0078]
(Supplementary note 5) In the hybrid node device according to supplementary note 4,
The specifying means has an inquiry means for inquiring a traffic flow rate of the electric label path for an input node terminating an electric label path passing through an output port where congestion has occurred among output ports of the electric node,
A hybrid node device, wherein an electric label path causing the congestion is identified from a response to the inquiry at an electric node of the hybrid node.
[0079]
(Supplementary note 6) In the hybrid node device according to supplementary note 5,
The hybrid node device, wherein the inquiry means inquires of a traffic flow of the electric label path using a signaling message.
[0080]
(Supplementary note 7) In the hybrid node device according to supplementary note 5,
The hybrid node device, wherein the inquiry means inquires about a traffic flow of the electric label path using a routing message.
[0081]
(Supplementary Note 8) In an electric node device of a network including an electric node that performs switching for each packet of an electric signal, an optical node that performs switching for each wavelength of an optical signal, and a hybrid node including the electric node and the optical node,
The measuring means for starting the measurement of the traffic flow after receiving an inquiry about the traffic flow from the hybrid node device described in Appendix 5 and notifying the hybrid node device of the measured amount after a predetermined time.
An electric node device comprising:
[0082]
(Supplementary note 9) In the hybrid node device according to supplementary note 4,
The path setting switching means, after setting a new optical label path and a new electrical label path, switching instruction means for instructing a node in the preceding stage of its own device to switch traffic.
A hybrid node device comprising:
[0083]
(Supplementary Note 10) In the hybrid node device according to Supplementary Note 4,
When the congestion that has occurred in the electric node of its own device ends, a switchback unit that checks that the congestion does not occur even if the switchback is performed and switches back the path.
A hybrid node device comprising:
[0084]
(Supplementary note 11) In the hybrid node device according to supplementary note 4,
Measuring means for measuring the traffic flow rate of each of the plurality of output interfaces of the electrical node of the own device,
A hybrid node device that determines occurrence of congestion and termination of congestion based on a moving average of the measurement values obtained by the measurement unit.
[0085]
(Supplementary Note 12) In the hybrid node device according to Supplementary Note 4,
Measuring means for measuring the traffic flow rate of each of the plurality of output interfaces of the electrical node of the own device,
Asset management means for determining the occurrence of congestion and the end of congestion based on the smoothed value obtained by smoothing the measurement value obtained by the measurement means
A hybrid node device comprising:
[0086]
(Supplementary note 13) In the hybrid node device according to supplementary note 12,
The hybrid node device according to claim 1, wherein the smoothed value is obtained by dividing a total sum of measured values obtained by the measuring unit in a fixed time width by the fixed time width.
[0087]
(Supplementary note 14) In the hybrid node device according to supplementary note 12,
The asset management means determines that the congestion occurs when the smoothed value exceeds a preset first threshold by a specified number of times, and sets a second threshold smaller than the preset first threshold to the smoothed value P. A hybrid node device that determines that congestion has ended when the number of traffic drops below a predetermined number.
[0088]
(Supplementary Note 15) In the hybrid node device according to supplementary note 12,
The asset management means calculates a sum of a plurality of smoothed values obtained within a predetermined time interval, determines that the sum is congested when the sum exceeds a third threshold, and determines that the sum is less than a fourth threshold smaller than the third threshold. A hybrid node device that determines that congestion has ended when the communication is completed.
[0089]
(Supplementary note 16) In the hybrid node device according to supplementary note 12,
The asset management means obtains the time change rate ΔP of the smoothed value, determines that the time change rate is greater than the positive fifth threshold value, congestion, and determines that the time change rate value falls below the negative sixth threshold value. A hybrid node device that determines that congestion has ended.
[0090]
(Supplementary Note 17) In an electric node device of a network including an electric node that performs switching for each packet of an electric signal, an optical node that performs switching for each wavelength of an optical signal, and a hybrid node including the electric node and the optical node,
Measuring means for measuring the traffic flow rate of each of the plurality of output interfaces of the electrical node of the own device,
Asset management means for determining the occurrence of congestion and the end of congestion based on the smoothed value obtained by smoothing the measurement value obtained by the measurement means,
When the congestion occurs, switching means for switching the electric label path causing the congestion to another electric label path
A hybrid node device comprising:
[0091]
(Supplementary note 18) In the network traffic control method according to Supplementary note 1 or 2,
A network traffic control method, wherein control of the hybrid node and the incoming node is centrally controlled by a network management system.
[0092]
【The invention's effect】
As described above, according to the first and fourth aspects of the present invention, it is possible to set a new optical label path that bypasses only the electric node of the hybrid node in which congestion has occurred, and uses an extra wavelength to avoid congestion. Can be prevented.
[0093]
According to the second and fifth aspects of the present invention, it is not necessary for the hybrid node to measure the traffic flow of all the electric label paths, and the amount of hardware for measuring the traffic flow can be reduced.
[0094]
According to the third aspect of the present invention, when optical resources are large, an optical label path can be set up without passing through an electric node, thereby preventing deterioration of service quality of traffic.
[Brief description of the drawings]
FIG. 1 is a configuration diagram of a conventional network architecture.
FIG. 2 is a configuration diagram of a basic network to which the network traffic control method of the present invention is applied.
FIG. 3 is a block diagram of one embodiment of an electrical node.
FIG. 4 is a block diagram of an embodiment of an input / output IF unit of an electric node.
FIG. 5 is a block diagram of one embodiment of an optical node.
FIG. 6 is a block diagram of an embodiment of an input / output IF unit of the optical node.
FIG. 7 is a block diagram of one embodiment of a hybrid node.
FIG. 8 is a flowchart of a path setting process in the first embodiment of the network traffic control method of the present invention.
FIG. 9 is a diagram illustrating a path setting process.
FIG. 10 is a diagram illustrating a path setting process.
FIG. 11 is a diagram illustrating a path setting process.
FIG. 12 is a diagram illustrating a path setting process.
FIG. 13 is a flowchart of a congestion avoiding process in the first embodiment of the network traffic control method of the present invention.
FIG. 14 is a diagram illustrating a congestion avoidance process.
FIG. 15 is a diagram illustrating a congestion avoidance process.
FIG. 16 is a flowchart of a congestion termination process in the first embodiment of the network traffic control method of the present invention.
FIG. 17 is a diagram illustrating a congestion end process.
FIG. 18 is a diagram showing a format of a signaling message.
FIG. 19 is a diagram showing a format of a routing message.
FIG. 20 is a flowchart of a path setting process when the number of wavelengths is depleted in the second embodiment of the network traffic control method of the present invention.
FIG. 21 is a diagram illustrating a path setting process when the number of wavelengths is depleted.
FIG. 22 is a diagram illustrating a path setting process when the number of wavelengths is depleted.
FIG. 23 is a diagram illustrating a path setting process when the number of wavelengths is depleted.
FIG. 24 is a diagram for describing a modification of the congestion avoidance process.
[Explanation of symbols]
e1 to e7 Electric node
o1 to o5 Optical node
h1-h4 hybrid node
y1 to y7 User device
c1 to c8 control node
NMS Network Management System
10 Optical / electrical conversion / separation unit
11, 41 input interface group
12 ₁ ~ 12n, 42 ₁ ~ 42m input IF
13 switches
14,44 Node control unit
15, 45 output interface group
16 ₁ ~ 16n, 46 ₁ ~ 46m output IF
17 Electric / optical conversion / multiplexing unit
18, 48 External control interface
20,50 Resource Management Department
21, 51 routing unit
22, 52 Signaling unit
31, 38 Packet processing unit
32 Label Management Department
33, 36 Packet buffer unit
34,37 Traffic measurement unit
43 Optical Switch
61 Optical amplifier / separator
62 Optical level monitor
66 transponder
67 Optical Amplifier / Multiplexing / Optical Level Monitoring Unit
74 Electric node controller
76 Hybrid node control unit
78 Optical node controller

Claims (5)

In a network traffic control method for controlling traffic of a network composed of an electric node that performs switching for each packet of an electric signal, an optical node that performs switching for each wavelength of an optical signal, and a hybrid node including the electric node and the optical node ,
When an ingress node and an egress node for which a path is to be set are designated, an optical label path and an electric label path passing through an optical node and an electric node constituting a hybrid node existing on the path from the ingress node to the egress node are set. And
Aggregation of packets is performed at the electric node of the hybrid node,
When congestion occurs in the electric node of the hybrid node, identify the electric label path causing the congestion,
A network traffic control method, wherein a new optical label path and a new electric label path are set and switched such that the congested electric label path bypasses the congested hybrid node electric node. The network traffic control method according to claim 1,
When congestion occurs in the electric node of the hybrid node, query the traffic flow of the electric label path to the input node terminating the electric label path passing through the congested output port among the output ports of the electric node,
A network traffic control method, wherein an electric label path causing the congestion is identified from a response to the inquiry at an electric node of the hybrid node. In a network traffic control method for controlling traffic of a network composed of an electric node that performs switching for each packet of an electric signal, an optical node that performs switching for each wavelength of an optical signal, and a hybrid node including the electric node and the optical node ,
When an ingress node and an egress node for which a path is to be set are specified, an optical label path and an electric signal that pass only through an optical node without passing through an electric node constituting a hybrid node existing on the path from the ingress node to the egress node Set the label path,
When an ingress node and an egress node that intend to set a new path in a state where the optical resources are depleted are specified, the hybrid path close to the hybrid node near the egress node from the ingress node of the new path is shared with the existing path,
Separating the existing path into two at the electric node of the hybrid node and setting and switching a new optical label path and a new electric label path to be connected to the egress node of the existing path and the egress node of the new path. Characteristic network traffic control method. An electric node that performs switching for each packet of an electric signal, an optical node that performs switching for each wavelength of an optical signal, and a hybrid node device of a network including a hybrid node including the electric node and the optical node.
When congestion occurs in the electric node of its own device, specifying means for specifying the electric label path causing the congestion,
A hybrid node device comprising: a path setting switching unit that sets and switches a new optical label path and a new electric label path so that the congested electric label path bypasses the congested electric node. The hybrid node device according to claim 4,
The specifying means has an inquiry means for inquiring a traffic flow rate of the electric label path for an input node terminating an electric label path passing through an output port where congestion has occurred among output ports of the electric node,
A hybrid node device, wherein an electric label path causing the congestion is identified from a response to the inquiry at an electric node of the hybrid node.

Images